ARCHIVED - Transcript, Hearing 13 July 2011
This page has been archived on the Web
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.
Providing Content in Canada's Official Languages
Please note that the Official Languages Act requires that government publications be available in both official languages.
In order to meet some of the requirements under this Act, the Commission's transcripts will therefore be bilingual as to their covers, the listing of CRTC members and staff attending the hearings, and the table of contents.
However, the aforementioned publication is the recorded verbatim transcript and, as such, is transcribed in either of the official languages, depending on the language spoken by the participant at the hearing.
Volume 3, 13 July 2011
TRANSCRIPTION OF PROCEEDINGS BEFORE THE CANADIAN RADIO-TELEVISION AND TELECOMMUNICATIONS COMMISSION
Review of billing practices for wholesale residential high speed access services. Notice of Consultation CRTC 2011-77, 2011-77-1 and 2011-77-2
140 Promenade du Portage
13 July 2011
In order to meet the requirements of the Official Languages Act, transcripts of proceedings before the Commission will be bilingual as to their covers, the listing of the CRTC members and staff attending the public hearings, and the Table of Contents.
However, the aforementioned publication is the recorded verbatim transcript and, as such, is taped and transcribed in either of the official languages, depending on the language spoken by the participant at the public hearing.
Canadian Radio-television and Telecommunications Commission
Review of billing practices for wholesale residential high speed access services. Notice of Consultation CRTC 2011-77, 2011-77-1 and 2011-77-2
Konrad von FinckensteinChairperson
Crystal HulleyLegal Counsel
James WilsonLegal Counsel
Tom VilmansenHearing Manager and Manager, Costing Methods and Tariffs
140 Promenade du Portage
13 July 2011
- iv -
TABLE OF CONTENTS
PAGE / PARA
11. Primus Telecommunications Canada Inc., Distributel Communications Limited 511 / 3087
15. Shaw Cablesystems G.P. 600 / 3675
16. PIAC/CAC (Public Interest Advocacy Centre and the Consumers' Association of Canada639 / 3928
CONSUMER PANEL (EASTERN PROVINCES)
17. Kimberly Smith 671 / 4089
CONSUMER PANEL (PROVINCE OF QUEBEC)
18. Andrew Moore 683 / 4170
- v -
PAGE / PARA
Undertaking 553 / 3329
Undertaking 561 / 3385
Undertaking 596 / 3651
Undertaking 627 / 3840
Undertaking 628 / 3851
Undertaking 632 / 3885
Undertaking 693 / 4233
--- Upon resuming on Wednesday, July 13, 2011 at 0900
3067 THE SECRETARY: Order, please. A l'ordre, s'il vous plaît. On est prêt à commencer.
3068 LE PRÉSIDENT : Madame la Secrétaire, before we start, yesterday during my questioning of CNOC I referred to a meeting that we had with CNOC at the Commission, with the full Commission, where they made us a presentation on UBB and I said that you had a chart there where you suggested charge UBB here.
3069 People have asked for what chart I was talking about. So in the interest of full disclosure I am now introducing the whole thing called "CRTC Presentation January 26, 2011 by CNOC," and the chart that I was referring to is the third last page of this document.
3070 I am doing this purely for the interest of transparency because CNOC advised their thinking had progressed since then, but since I referred to the document I think it's only fair that I share it with everybody who is participating in this hearing.
3071 THE SECRETARY: And, sir --
3072 MR. TACIT: Mr. Chair --
3073 THE SECRETARY: -- some more copies are being printed right now and we will put them on the back table for everyone in a few minutes.
3074 THE CHAIRPERSON: We are not in the reply phase but I gather CNOC wants to say something.
3075 MR. TACIT: Just about this issue, Mr. Chair, I've conferred with a few of the people there and when we were talking about UBB we were talking about it in very generic terms. We were not proposing a particular methodology. We wanted it to be at an aggregated level at the ASPI but we had not proposed a specific methodology at that time.
3076 THE CHAIRPERSON: I didn't say that.
3077 MR. TACIT: Okay, fair enough.
3078 THE CHAIRPERSON: Nobody is putting anything in your mouth. I just wanted to recall our conversation where we talked about UBB generically and you basically took the same position as my colleague Commissioner Molnar that if you do apply it, that would be the right level to apply it.
3079 MR. TACIT: That would be the location, correct.
3080 THE CHAIRPERSON: Right. Okay, thank you.
3081 So now that we have that out of the way, let's go with today's proceeding.
3082 Madam Secretary.
3083 THE SECRETARY: We will now proceed with the presentations by Primus Telecommunications Canada Inc. and Distributel Communications Ltd., who will be appearing as a panel.
3084 We will hear each presentation, which will be followed by questions from the Panel.
3085 We will start with the presentation of Primus Telecommunications Canada Inc.
3086 Please introduce yourselves for the record, after which you will have 20 minutes for your presentation.
3087 MR. STEIN: Good morning, Mr. Chair and Commissioners. My name is Matthew Stein, and I am the Vice-President of Network Services, Primus Canada.
3088 Joining me today on the panel is Christopher Hickey, Regulatory Affairs for Primus.
3089 Let me begin by clarifying that l am not a lawyer, a regulatory expert or an economist. What I am is a network architect that has been designing, building and managing networks since long before the Internet had graphics.
3090 At Primus we provide retail Internet services to businesses and consumers, both through the use of incumbent wholesale Internet access services and over our own network. Our network, which is comprised of hundreds of DSLAMs stretched across five provinces, is built on leased facilities and services.
3091 In our network we meet new challenges due to growth every day. We go through the same process of capacity management and congestion relief as the rest of the industry. Put simply, we are challenged every day by the same rapid Internet data growth that the incumbents face in their networks.
3092 This experience is the foundation upon which Primus' two central messages in this proceeding are built.
3093 First, data transfer volume pricing has absolutely no place in wholesale billing arrangements.
3094 Second, peak traffic capacity billing methods -- peak capacity billing -- such as 95th percentile billing are the proper billing model for capacity-sensitive components of wholesale access services.
3095 Having said that, let me briefly answer the questions that you have asked us to address before I return to a couple in more detail.
3096 In your first question, you asked us to suggest which billing model should be adopted. We believe that the Commission should adopt a billing model that applies distinct rates to the access, traffic capacity sensitive aggregation and interface components of wholesale access services. This billing model should apply to all wholesale access providers and all wholesale access services: cable and DSL, residential and business, FTTN and legacy.
3097 With regard to your second question about the method for charging usage, the usage component should be measured and billed through the use of a peak-capacity billing method. Unlike data transfer volume pricing, this is an industry standard practice that links billing to the impact that competitors have on the available capacity in the network. This ensures that competitors are held responsible for managing that impact, while also allowing them to offer unique and differentiated retail service offerings.
3098 With respect to the impact of the two models, the measurement and billing based on peak capacity will result in the competitive ISPs bearing the costs of the impact they make on incumbent networks. This in itself will incent those ISPs to manage their traffic and result in competitors with higher peak traffic paying the incumbent more than competitors with lower peak traffic. Furthermore, it will invite innovation in how that traffic is managed and used by both ISPs and consumers alike.
3099 Your fourth question asked about symmetry with respect to the mark-up. Yes, as a matter of principle, the same mark-ups on cost-based rates should apply to all ILEC and cable carrier wholesale access services, with the only differences being the manner in which the mark-ups are applied to distinct usage-sensitive elements specific to each network architecture and the principle established in the speed-matching decision.
3100 It must be recognized though that these consistent, fair and reasonable mark-ups should be set at a level that allows a vibrantly competitive market to form on both service types. Also, the overall rates should decline over time to recognize the continual and dramatic decrease in cost that characterize these services.
3101 To your fifth question, there is no reason the mark-up should be the same for access and usage. These are different elements and do not need to be linked in this way. What should be the same is the insistence on these mark-ups being fair, reasonable and structured in a way that fosters a vibrantly competitive market.
3102 Lastly, the principles that result from this proceeding should apply to legacy services. There is absolutely no reason that legacy services should be treated differently.
3103 That said, I would be remiss if I did not point out the need to dispense with this idea that 5 Mbps and 6 Mbps DSL are somehow different because somebody chose to brand them that way. DSL is DSL. Whether we choose to now note that we're feeding the node with fibre where we never noted it before or not, it's the same.
3104 To understand why peak-capacity billing is superior to volume billing, it is first important to understand both models.
3105 That said, I will not go into a full review of the volume-pricing model as I believe that we are all familiar with it, but in summary it places a specific per GB rate on the volume of traffic that is passed.
3106 Peak-capacity billing simply means billing based on the amount of traffic to cross the network at the highest traffic moment of the billing period, which is a whole month -- in other words, the highest traffic in a month.
3107 The highest traffic moment is measured by continual monitoring of capacity usage at set intervals, typically five minutes or less. As applied in this manner, peak capacity is the highest amount of traffic that passes in any given second in a month.
3108 To be clear, billing according to peak-capacity usage also means that you pay for all capacity used under the peak, not just the moment of peak. You pay for the whole month as if you used that consistent amount at peak.
3109 For illustration, let's use the current capacity-based interconnection between Bell and a competitive ISP.
3110 To connect to Bell's wholesale access service, a competitor leases a pipe with a set maximum capacity from Bell, such as 100 million bits per second, 400 million bits per second or 1 billion bits per second -- we say 100 megabits, 400 megabits, 1 gigabit.
3111 If a competitor leases a pipe with a capacity of 1 gigabit per second, the maximum capacity the competitor can use at any time is 1 gigabit per second. That is also the maximum network impact that the competitor can have on Bell's network because you cannot pass more than 1 gigabit on a 1-gigabit link.
3112 In practice and as shown in the graph below, the amount of traffic that a competitor uses fluctuates throughout the day. Using the peak-capacity billing method, the billing is calculated by looking at the very highest capacity point reached overall and then billing for that amount for the entire month as if it was used at that peak flat.
3113 In practice, the use of peak-capacity billing provides numerous benefits relative to the volume-pricing model.
3114 First, peak-capacity billing makes competitors accountable for their actual network impact. Volume pricing does not. Networks are designed around peak traffic, or the highest amount of possible use at any given time, not the amount of data that passes across it. Congestion then occurs in the moments where there is insufficient capacity available in the network.
3115 In the context of competitive ISPs and incumbent networks, the volume of data transferred to and from the ISP is not an indicator of congestion. In fact, from a congestion standpoint, the amount of data transferred overall is wholly irrelevant because there's no guarantee that the volume in question was transferred at a time where there was any congestion at all.
3116 The right way to properly account for individual competitive ISP contribution to congestion is by charging for their peak, the highest amount, the highest impact they have on the network.
3117 Let's look at example to see how this works in practice with the peak-capacity billing model.
3118 In the following graph, the network capacity usage of three fictitious ISPs is represented. ISP A has a very high capacity-usage peak but very little use outside of its peak. ISP B has an obvious peak period but does have substantial use during the rest of the day. Lastly, ISP C has a perfect traffic distribution, which, while impossible to achieve, is useful for our example here today.
3119 In this example, we can see that ISP A with its extremely high peak is using a far higher proportion of the capacity available on the network during that time, and as such should pay more than the other ISPs.
3120 ISP B has filled in the valley quite a bit by moving traffic out of peak and has lowered their peak dramatically, and in doing so has lowered both their impact on the network and the costs they pay.
3121 ISP C, with its impossibly flat traffic distribution would be paying the least of the bunch because they are making the smallest dent.
3122 Please note, all three of these ISP networks have transferred the exact same number of gigabytes. Said differently, under the volume-billing model previously discussed, these ISPs would have all paid the same amount, despite making drastically different impacts upon the network.
3123 More important is that each ISP is separately accountable for the impact they have on the network.
3124 Peak-capacity billing also provides competitors with incentives to manage their impact on the network. Again, volume pricing does not.
3125 According to the model proposed by Bell a gigabyte is priced the same regardless of when it is passed, at 8:00 a.m., 4:00 p.m., midnight or 3:00 a.m. A volume-billing model therefore provides no incentive for competitors to manage their traffic in ways that will reduce capacity usage during the highest usage periods, which is the very issue that this is purported to solve.
3126 Essentially, a volume-billing model allows the ISP to abdicate its responsibility of managing traffic because it doesn't matter when it comes, how it arrives or what forms it takes. When efficiency is not rewarded, it will not be sought.
3127 In contrast, peak-capacity billing correctly places the responsibility to manage traffic back into the hands of competitors and incents them to manage that traffic accordingly.
3128 For example, in the graph above, ISP A may be very concerned with the large difference between their average-capacity usage and their peak-capacity usage on which they are billed.
3129 If ISP A wants to manage their capacity, they have many options. They could implement various economic or technical Internet traffic-management practices, seek customers with lower usage requirements or just even different usage requirements, different peaks for the day, or implement measures to incent their customers to shift usage from the peak period. In doing so, they will be able to lower that peak, which costs them less. Here, the incumbent also benefits by having more available capacity on its network.
3130 On the other hand, maybe ISP A is catering to a high-usage market that demands service offerings that result in such a capacity-usage pattern and can sustain premium prices. Maybe they cannot or do not want to alter their usage pattern. If this is the case, then ISP A will pay far more than ISP B. And why shouldn't they? They're the ones using that capacity. They should be held accountable for their business model and the incumbent benefits handsomely by charging the ISP for that high peak.
3131 Peak-capacity billing is an industry standard practice, whereas volume pricing is not. No other capacity-based data network components are billed on volume levels. It is telling that network equipment is never rated by how much data it can pass but by how much data it can pass in a second. Network links, devices, connections and systems are all rated by how much data they can pass in a single second.
3132 We've never heard a vendor trumpet the capabilities of their latest router by telling us it can pass a terabyte of data eventually. They say it can pass it in a second.
3133 Volume of data transferred is irrelevant to understand the capability or congestion state of a network unless you consider the amount of time that transfer took.
3134 Peak-capacity billing is currently used by Primus in all of its link, backhaul and transit arrangements. It is used in numerous ISP and carrier-to-carrier billing arrangements. At Primus, we buy Internet transit from many of the companies in this very room and we do so with them exclusively through a peak-capacity arrangement of one type or another.
3135 We do not believe that implementation issues should be an impediment to the use of peak-capacity billing. However, if somehow it is found to be, Primus supports an alternative approach that would recover the capacity component through the pipe size purchased by the competitor.
3136 Under this alternative model, the competitor would purchase a total maximum capacity at the network interface, for example, at amounts from 100 to 1000 Mbps, as required. The competitor's total impact on the network would be limited to the capacity it purchased.
3137 If for some reason the ILEC cannot figure out how to implement peak-capacity billing on these links, then just raise the price of these links to assume 100 percent of capacity is in use at all times. Just change it. If an ISP wants 100 Mbps, charge them as though they'll use it all the time. 400 Mbps? 1000 Mbps? Charge them like they'll use it. Same thing. No complicated billing changes or long implementation times. Just change the price.
3138 Though the peak-capacity billing method is preferable, this alternative model would provide similar benefits and would remove these implementation concerns voiced by the incumbents about billing systems and so forth.
3139 Finally, volume pricing is prone to error and needlessly complex to administer. Peak-capacity billing is not. This became very clear during the implementation period that followed the original decision approving end user usage-based billing.
3140 During that time, Bell provided documentation regarding how it measured volume. That documentation included numerous technical reasons as to why Bell's per-gigabyte measurements would differ from competitors' per-gigabyte measurements.
3141 At that time, it was generally agreed that it was technically impossible for the measurements to be always the same because they measured traffic in different ways in different places in the network. It quickly became clear that complex audit processes would be required and that a high number of disputes would be unavoidable.
3142 Though volume pricing is slightly different than the original end user usage-based billing model proposed by Bell, it will nonetheless be characterized by the same need for overly complex processes and a high number of disputes.
3143 To put a finer point on this, Bell provided preview usage measurements of all of Primus' GAS end users, as they did for all ISPs -- two of those ISPs, I believe -- during the implementation period that followed the decision approving Bell's original end user usage-based billing proposal.
3144 In this preview measurement provided by Bell in April 2011 -- we took a look at it and evaluated it against our own -- the usage measurements taken by Bell's calculation systems matched the usage measurements taken by Primus' for only 3.5 percent of our total user base.
3145 To take this one step further, by billing a competitive ISP based on the traffic consumption of its customers, three separate parties must also agree on that number. The wholesaler has their recorded number, the ISP has their recorded number, but so does the end user. Again, it is unlikely for these numbers to line up because the measurement is being taken at different points in the network rather than one shared interface connecting to both parties that they can both mutually monitor.
3146 In the peak-capacity billing model the capacity would be measured at the existing interface between the wholesale access provider and the competitor. For Bell, this would occur at the AHSSPI. A competitor may have any number of AHSSPIs, each of which supports many thousand customers and each with capacity usage that would have to be measured.
3147 Any dispute would require only an investigation into the capacity usage of the AHSSPIs they have and not the thousands of customers that the AHSSPI supports. Further, both the wholesale-access provider and the competitor are able to monitor their capacity usage in real time, allowing each party to not only forecast but trend and have visibility into the capacity levels.
3148 This is an efficient process in and of itself, let alone when compared to the need to audit thousands of records to determine validity of per-gigabyte volume measurements.
3149 It is clear that peak-capacity billing is superior to volume billing. It allows competitors to fully manage their traffic and has many benefits related to ongoing administration, all of which are beneficial to competition.
3150 In addition, it holds competitors accountable to the impact they have on the network, requires competitors to manage that impact, addresses the issues that usage-based billing is supposed to address and ensures the incumbent is fairly compensated for the impact competitors have on their network, all of which are beneficial to the incumbents.
3151 Finally, it allows competitors to offer unique and differentiated retail Internet services which ultimately benefits Canadians.
3152 Thank you very much.
3153 THE SECRETARY: Thank you very much for your presentation.
3154 We will now proceed with the presentation of Distributel Communications Ltd.
3155 Mr. Mel Cohen is appearing for Distributel. Please introduce your colleagues for the record, after which you have 20 minutes.
3156 MR. COHEN: Good morning, Mr. Chairman, Vice-Chairmen, Commissioners and staff. My name is Mel Cohen. I am President of Distributel Communications.
3157 With me today are Don Cavanagh, Vice-President of Telecommunications, and Anne McMillan, our Regulatory Analyst.
3158 Thank you for giving us this chance to speak to you today.
3159 Mr. Chairman, inadvertently -- this is not in my notes -- we prepared some exhibits for today's presentation and didn't realize that that would constitute evidence and I would ask your permission to enter them -- allow them to be distributed.
3160 THE CHAIRPERSON: Can I see what you want to introduce, Madam Secretary, please?
3161 MR. COHEN: The presentation in large part talks to these. I can get them.
3162 THE CHAIRPERSON: It's these two pages? This is a Backhaul Management Status Report for all the various data?
3163 MR. COHEN: I won't be talking about the content so much as the form of it.
3164 THE CHAIRPERSON: Okay. Well then introduce them as an exhibit. We will put them in the back and people can comment on it as they want.
3165 Thank you. Go ahead.
3166 MR. COHEN: Okay, thank you, Mr. Chair.
3167 I would like to clarify our appearance here with respect to CNOC. Distributel is proud to be a member of CNOC. We are very happy with the work that it has done on behalf of all of its members and we fully support its written submissions in this proceeding as well as its oral presentation at this hearing.
3168 Our main motivation for appearing separately from CNOC was to ensure that cable Internet issues get sufficient attention in this proceeding. As the Commission is well aware, TPIA uptake has lagged way behind wholesale DSL uptake. Distributel is one of just a handful of CNOC members that actually has a fair bit of TPIA experience and we wanted to make sure that we shared some of that experience with you decision-makers.
3169 Since our decision to appear was made well in advance of the release of the guidelines for structuring oral presentations released last week, I would ask the Chairman's indulgence again if our presentation does not follow that structure in a point-by-point manner. However, I assure you that what we will present is most relevant to the decisions you will be taking.
3170 The other reason that I wanted to appear separately from CNOC is that in speaking for ourselves we can be freer with the information that we disseminate. In preparing for today, we've thrown caution to the wind with respect to competitive disclosure.
3171 I hope this doesn't come back to bite us you-know-where, but I believe that we have far more to gain by giving you clear insight into our operations and decision-making than we stand to lose through disclosure of competitively sensitive information.
3172 Now, just a bit of background on Distributel as it relates to the matters before the Commission today.
3173 Distributel began as an entrepreneurial long-distance company in 1988. For the first 15 years, that remained our core business.
3174 From a regulatory and societal perspective, the introduction and facilitation of competition in long distance has been a huge success. The proliferation of competitors and competitive alternatives has been such that most people now take inexpensive or even free long distance as a matter of course.
3175 The flip side for us has been that our long-distance revenues and profits have been in steady decline for the past several years. To secure a future for the company, we realized that we had better move into areas with more growth potential and we set out to do that quickly through acquisition.
3176 In mid-2009 we purchased the assets of Cybersurf Corp. The Cybersurf asset purchase did three important things for us.
3177 First, it more than doubled the size of our Internet business, from about 12,000 high-speed Internet clients to approximately 31,000.
3178 Second, it took Distributel into Western provinces, Alberta and B.C.
3179 Third, it put us into the cable Internet business, with connections to Videotron, Rogers, Cogeco and Shaw.
3180 Cybersurf was a pioneer customer of Third Party Internet Access and, to our knowledge, was the largest TPIA client at the time of the acquisition, despite having just 7,500 end users.
3181 In 2010, we went on to acquire Acanac Inc., with roughly 50,000 DSL customers in Bell Canada territory. The Acanac acquisition plus organic growth has brought our total today to approximately 91,000 high-speed Internet clients, 75,000 on DSL and 16,000 using TPIA. This makes the Distributel group one of the larger members of CNOC and one of the largest independent ISPs in Canada.
3182 But it is important to keep this "large" status in perspective. The incumbents are still hugely dominant. For example, in response to interrogatory Shaw indicated that it had over 1.8 million high-speed Internet clients. In Shaw's territory, where we currently have fewer than 1,500 clients, they are 1,200 times our size.
3183 That concludes my background information. The next information I would like to present is relevant to the conduct letter's question numbers 2 and 3, about how best to measure usage.
3184 When we took control of the Cybersurf's operations one of the problems we encountered was congestion on the facilities used to backhaul from the cable company points of interconnection to our own Points of Presence.
3185 Prior to Decision 2010-632, cable companies were not obliged to provide interconnections in sizes greater than 100 megabits per second. We had 19 different POIs, many with multiple 100 megabit per second connections from a variety of suppliers. We also discovered that the prices we were paying for those backhaul facilities were driving our per-client costs so high as to render the services we were selling unprofitable.
3186 In addition to the cost, keeping up with growth by provisioning backhaul in increments of 100 megabits per second proved impractical. The time lapse from date of ordering to date of installation was unpredictable and we found ourselves continually embargoing new sales to different areas in order to avoid degrading service to existing clients.
3187 Please turn to Exhibit 1, which is a slightly modified version of a recent POI Backhaul Management Status Report.
3188 On that report each POI is listed together with its relevant backhaul provider, the speeds of the backhaul facility, some provisioning notes and status, peak usage to that POI, and the number of subscribers and the average peak usage per subscriber.
3189 There is also a "Flatline" column that indicates if there are extended periods of service degradation where the installed capacity is inadequate for the demand.
3190 There are no longer any Flatlines. This was due to a decision on our part to invest an additional $25,000 per month in operating expense to build enough spare capacity into our backhaul facilities so that we could accommodate more growth between augments. For the most part, this meant replacing 100 megabit per second circuits with gigabyte Ethernet facilities. For us, the Commission's requirement that the cable companies accommodate gigE interconnection was very fortunate and timely.
3191 The report shows the diversity of arrangements we have in place with the various cable companies. There are four sections, one for each company.
3192 In the case of Rogers, there are 14 active POIs, each with a 1 gigabyte backhaul facility, and 8 additional planned POIs.
3193 The gigE backhaul facilities are mapped to two 10 gig aggregation facilities that connect Rogers' Toronto POP to Distributel's Toronto POP.
3194 In the case of Cogeco, on page 2 of the exhibit, we have a single POI in one city, Windsor.
3195 In the case of Shaw, we have interconnections at three POIs, Calgary, Edmonton and Vancouver. They are still served by 100 megabit facilities, as inherited from Cybersurf.
3196 Finally, in the case of Vidéotron, which has been providing an aggregated TPIA service all along, we have a single interconnection that serves all five POIs.
3197 An interesting development is that in all cases now, the supplier of our backhaul facility is a division of the cable company to which we are interconnecting. In at least one case, we had a better quote from another supplier, but the inability of the cable company to reach agreement with that competitive provider with regard to placement of the latter's terminating equipment led us to give up for the sake of timeliness.
3198 We pressed the cable company's data division to match the price and terms of the competitor and when they finally acquiesced we abandoned the competitor. However, this kind of impasse conflicts with reliance on a competitive market for high-speed data facilities and may need to be addressed through subsequent regulation.
3199 Please note that in our POI Backhaul Management Status Report the only usage measures are those related to peak demand. Outside of peak period, the usage our customers make of our network has no impact on our facilities planning nor on our costs. Total data transfer is irrelevant to our planning, so we don't even track it.
3200 One of the things I would like the Commission to glean from the material I have presented here relates to a topic not listed in the conduct letter, but which is a live issue in this proceeding. It relates to aggregated versus non-aggregated POIs.
3201 Independent ISPs that are already TPIA customers have not been sitting on our hands waiting for regulatory issues to be resolved. Those of us already in the business are compelled to keep going. In many cases this has led us to make multi-year commitments to backhaul facilities from non-aggregated POIs.
3202 Distributel supports maintenance of both aggregated and non-aggregated POIs on a going forward basis, but should the Commission ultimately decide that future TPIA interconnections must be made only at the aggregated POIs, then some reasonable accommodation must be made for those who have already deployed on a disaggregated level.
3203 In cases where the backhaul facilities have been sold by an affiliate of the TPIA provider, this accommodation could come in the form of a Commission directive to the cable companies to have their affiliates release the ISPs from their contractual obligations and allow us to establish new interconnections without service charges. Another option might be some type of flexible grandfathering arrangement.
3204 No matter what, the introduction of a new regime should not result in independent ISPs being penalized by having to pay for backhaul facilities that they can no longer utilize.
3205 The next segment of my presentation relates to conduct letter items numbers 4 and 5, mark-ups.
3206 The rates that have been proposed for TPIA at aggregated POIs are calamitously higher than the existing non-aggregated rates. Our initial reaction was that this was due to the aggregation, but this cannot possibly be the case since, as we will reveal shortly, our own costs for backhaul aggregation are a fraction of those increases. This is true despite the fact that we are buying our backhaul from those same companies at profitable market rates. We have nowhere near their volumes, so rates based on their costs to transport traffic between their own locations should be yet another order of magnitude lower.
3207 This means that the huge rate increases proposed by the cable companies cannot possibly be attributed to POI aggregation, unless their cost studies are just plain wrong or the mark-up is too high.
3208 So what is that appropriate mark-up?
3209 I believe it should be one that makes the competitive market work properly. We will show that the rates that have been proposed for TPIA services will not permit that. Whether that is due to flaws in the cable companies' costing studies or a mark-up that is just too high under today's circumstances is, frankly, beyond Distributel's sophistication to determine, but one way or another, the Commission will have to dramatically reduce those proposed prices if it wants to satisfy the requirement for viable competition.
3210 To demonstrate this, we are going to share our business case for Western Canada, where we are just preparing to launch services under the Distributel brand.
3211 As I mentioned earlier, we currently have fewer than 1,500 cable Internet customers in Shaw's territory. Those are clients that we obtained from the Cybersurf assets purchase.
3212 Some months ago I was looking at our costs for cable Internet in Shaw's territory and I realized they were too high for us to compete profitably. Our backhaul costs based on 100 megabit circuits were too high, as were our Internet transit costs.
3213 I asked Don here -- who wears two hats at Distributel, Regulatory and Telecom Networks -- what we had to do to get our costs down. His answer was simple: volume. Don told me that if we had about 20,000 end-users we could get both costs down to a level where we could compete on price and still make a profit. So that is what we set as our minimum target and subsequently decided to try and reach within one year.
3214 Note that we start with the premise that our retail prices should be lower than the incumbents' retail prices.
3215 The ability to serve specialized markets, to offer new features or a diversity of packages, those are all important benefits of competition, but one job that competition must do to fulfil its proper role in the marketplace is to drive prices toward costs.
3216 In Shaw's territory, we expect that most of our clients would take either the 7.5 megabit per second service or the 15 megabit per second service that Shaw recently upgraded to 25 megabits per second with no change in price.
3217 Here is how we plan to position our offerings.
3218 Shaw's 7.5 megabit per second service, which includes 125 gigabytes of data transfer is priced at $49.00 per month on a standalone basis and $39.00 in a bundle. Against this we intend to position an unlimited service at $43.95.
3219 Shaw's 25 megabit per second service, which includes 250 gigabytes of data transfer is priced at $59.00 per month on standalone basis and $49.00 per month in a bundle.
3220 Currently Shaw is offering a promotional price for this service of $29.95 per month for the first six months, with no minimum contract commitment.
3221 Against this we intend to position an unlimited service at $49.95, which will be a tough sell against $29.95.
3222 To make our Western expansion happen we will have to upgrade our network equipment in Vancouver and Edmonton. Our plan is to spend about $300,000 on capital equipment and one-time service installation charges. To serve 20,000 clients, a capital expenditure of $300,000 is not a huge investment. It works out to about $15 per client.
3223 But this pales in comparison to the negative cash flow that we will invest in this business plan while we build the customer base. That investment will be in the millions.
3224 Now I will ask you to look at Exhibit 2, where we list our costs on a per end-user basis.
3225 Table 1 of Exhibit 2 shows a list of costs that we incur as we add each new cable Internet subscriber. The largest of these is the marketing acquisition cost at $145, followed by the incumbent's service charge. $63.33 represents a blend of the different rates for new installs versus transfers. Total upfront cost, $270.
3226 Table 2 of Exhibit 2 shows the monthly recurring costs. The largest of these is clearly the incumbent's tariff rate, but the other items are hardly inconsequential, amounting to roughly an additional 50 percent.
3227 Note that the backhaul and aggregation component is in the $3.00 range which, as I mentioned earlier, is just a fraction of the proposed increases for the aggregated POIs.
3228 Table 3 shows the margin calculation. You can see that we are left with roughly $12 per client at the end of each month.
3229 So you can see that we go almost two years before we recover our customer upfront costs and actually start making any money. Since many customers do not stay that long, we assume the associated risk, and hope that we make it back from those that stay longer.
3230 Now I will ask you to look at Exhibit 3, where we have reproduced the same costs, but with the interim tariffs based on retail minus 35 percent. These rates are not significantly different from those first proposed by Shaw in its December 10th filing.
3231 Note that the increases are $10.17, 48 percent, for the 7.5 megabit service, and $12.90, 52 percent, for the 25 megabit per second service. The 25 megabit per second service rate is almost $8.00 above the six-month retail promotional price of $29.95.
3232 Also note that the impact on the business plan is devastating. The time to recover on a 7.5 megabit service client jumps to almost 10 years, an impossible proposition.
3233 And the 25 megabit per second service at $49.95 has a negative margin. It simply cannot be offered.
3234 There is no room to increase our prices.
3235 So where are we in this plan? Well, we placed the orders for the capital equipment two weeks ago and we are placing the orders for the backhaul and TPIA interconnections this week. There can be no waiting to see the regulatory outcome or our entry into the market will have to be delayed indeterminably. We will simply assume the risk that the Commission will not allow the incumbents to impose unworkable rates.
3236 I believe the mandate of the Commission is clear, the government and the people of Canada want to see real competition in wireline Internet. Real competition means viable companies offering lower prices. With a duopoly supply of last mile access, built upon a monopoly legacy foundation, mandated wholesale high-speed access is the means to that end. It is up to the Commission to ensure that the rates charged to competitors compensate the incumbents for their real costs, but are no higher than necessary to ensure that competition has a reasonable chance to thrive.
3237 Thank you for your attention.
3238 THE CHAIRPERSON: Thank you for your submissions.
3239 Let me deal with Primus first.
3240 You are clearly, like all of the CNOC members, in favour of the peak period measurement. You have heard all the implementation objections that the other companies have made. I asked CNOC yesterday and they basically suggested they were overblown, that this measurement is already taking place, it can be done, it can be implemented in a limited time.
3241 What is your position on that?
3242 MR. STEIN: I would agree that it can be done in a very reasonable time.
3243 I have a couple of comments there.
3244 If you are looking at the billing system side of things, that's one thing; if you are looking at the capacity and the recording of the information, that's a little bit different.
3245 From a capacity recording, and so forth, I can't see how there are implementation challenges for doing this kind of thing. The way that you capacity manage a network is the same and therefore you could record that information.
3246 However, from a billing standpoint we did propose an alternative proposal. If it is true --
3247 THE CHAIRPERSON: I will come to that in a second, yes.
3248 MR. STEIN: Right. So if in fact it's true, but I can't comment on how complicated or not it would be to update somebody's billing system, you will have to ask them.
3249 THE CHAIRPERSON: Now, the other thing is that Bell gave you a clear picture --
3250 MR. STEIN: Yes.
3251 THE CHAIRPERSON: -- and if you just look at it one of their main points was that the contention actually occurs at points 1, 2 and 3 and yet you talk about peak capacity or the measuring will only take place at what is called A-HSSPI.
3252 MR. STEIN: Yes.
3253 THE CHAIRPERSON: How do you deal with this argument?
3254 MR. STEIN: Sure.
3255 Well, the first thing, I think we have to separate the incumbent capacity managing their own network, finding peaks, dealing with those peaks, building appropriately, and so forth.
3256 Their point that capacity problems, and so forth, are happening at 1, 2 and 3, that's okay, I accept that and if that's the case then I have to take their word for it.
3257 THE CHAIRPERSON: Yes.
3258 MR. STEIN: However, what we are talking about here is the appropriate method to charge.
3259 Because the A-HSSPI does have all the traffic go across it, we can be charged based on that amount. So, therefore, the fact that individual peaks are happening elsewhere in the network at individual times doesn't take away from the fact that by charging for peak billing we would be paying for all of it when it occurs by recording it at the A-HSSPI,
3260 The other piece to keep in mind --
3261 THE CHAIRPERSON: Stop right there.
3262 MR. STEIN: Sure.
3263 THE CHAIRPERSON: I just want to make sure I understood what you are saying.
3264 You are saying even if they didn't have any wholesale they would have contention problems as part of serving their own network and selling to you wholesale may aggravate it, but that aggravation gets caught by being measured at the A-HSSPI, you don't have to measure it at the other as well.
3265 Is that what you are --
3266 MR. STEIN: Well, I'm saying that we would not be billed at those points.
3267 THE CHAIRPERSON: Yes, okay.
3268 MR. STEIN: That's right. You measure it at this one point, you get a recording of the aggregate amount of traffic --
3269 THE CHAIRPERSON: Yes.
3270 MR. STEIN: -- going from Network "A" to Network "B" and by doing it at that one point it is consistent. Both parties see sort of either side through that little tunnel --
3271 THE CHAIRPERSON: Yes.
3272 MR. STEIN: -- and we can all record and see the same exact number.
3273 In their network, they will still have to capacity manage those individual micro peaks that occur all through the network, but that's what they already are doing.
3274 THE CHAIRPERSON: You are saying, if I understand it -- you are about 30 percent of the network, I think they said.
3275 MR. STEIN: I believe they said 17 percent of customers, 29 percent of the traffic.
3276 THE CHAIRPERSON: You're right, something like that.
3277 MR. STEIN: Yes. So certainly nowhere near half.
3278 THE CHAIRPERSON: Basically they don't have problems with those and if traffic increases by 30 percent at the A-HSSPI because of you, then they can figure out what the impact of that means for points 1, 2 and 3 because they have to do it for their own traffic as well; right?
3279 MR. STEIN: That's right.
3280 THE CHAIRPERSON: That's essentially it.
3281 So then said, "Look, I'm reasonable, I don't understand why it needs a new billing system and I'm proposing this alternate."
3282 Explain to me how this alternate works. It looks to me very much like a flat billing, but maybe I'm missing something.
3283 MR. STEIN: Well, the alternative works sort of like this, in those A-HSSPI circuits, each individual one, each one would have a certain size. If you can assume -- again let's use that 1 gigabyte per second connection.
3284 THE CHAIRPERSON: Right.
3285 MR. STEIN: In that one gigabyte per second connection, it in itself already has the maximum capacity of 1 gigabyte. If I purchase the 1 gigabyte version, I'm okay with the idea of charge me as if my peak was 1 gigabyte, because if I knew that peak would be nowhere near 1 gigabyte I could have bought the 400 megabit version, so 40 percent the size. If the 400 megabit version was too big, I could have bought the 100 megabit version. Really, the 100 megabit version is certainly small enough for the smallest of ISPs. I believe 100 megabits is the smallest that they offer today anyway.
3286 So I'm suggesting that they already have these tiers, simply change the price. Change the monthly recurring charge that we are charged for these tiers to incorporate the assumption that all the usage of that thing would be used.
3287 Remember, when you pay for the peak, if you are paying for 700 megabit, you are paying as though you have used it all month. That's why I pointed out that the most efficient, the perfect traffic distribution, is flat -- it's impossible, but it's, technically speaking, perfect.
3288 THE CHAIRPERSON: But this alternative actually may mean that you actually pay for capacity that you are not using; right?
3289 MR. STEIN: Well, remember, we are managing our network as well, we are managing our own points of contention, and so forth.
3290 THE CHAIRPERSON: Right.
3291 MR. STEIN: So we have a very clear indication and of course, like everybody, we use 95th percentile to calculate where those peaks are.
3292 THE CHAIRPERSON: Yes.
3293 MR. STEIN: So we have good indication in advance of the peak that we will reach. We don't have necessarily a volume calculation because those do not correlate, but we do know the peak that we will reach quite reasonably.
3294 So it's true that I may be needing 800 megabit and have to buy 1,000 --
3295 THE CHAIRPERSON: Yes.
3296 MR. STEIN: I may buy 2,000 megabit and have needed 1,300, but that's okay, I'm willing to accept that. That's not a difficult --
3297 THE CHAIRPERSON: And you would be doing this on a monthly basis?
3298 MR. STEIN: Well, we already do it on a monthly basis today, yes.
3299 THE CHAIRPERSON: Okay.
3300 MR. STEIN: But we calculate the capacity in peak day to day of course, but as far as purchasing new A-HSSPI's and adding to that network, that is happening every single month anyways.
3301 THE CHAIRPERSON: Is this alternative billing solely your product or does CNOC support this?
3302 Have you discussed it with CNOC?
3303 MR. STEIN: You would have to ask CNOC.
3304 THE CHAIRPERSON: Okay. I will do that. Thank you.
3305 Then, Distributel, I'm glad that you are here and explaining to us the problems of TPIA, et cetera, but if I understand it essentially there are two problems, (a) the rates are wrong, the interim rates. You can't live with them and you want us, whatever we come with as final rates, come with something that is more realistic of the market. That's really your biggest point, right?
3306 Your secondary point was the aggregated POIs?
3307 MR. COHEN: In some sense they are tied together. In Shaw's territory they have declared that the three POIs to which we are interconnected, which were the only three POIs listed in their tariff, have now become aggregated POIs.
3308 In consultation with them this week, they confirmed that there was not likely anything added to Edmonton or Vancouver and perhaps 15 percent houses past added to Calgary, yet because these are now aggregated POIs they want us to start paying the interim rates, which are 50 percent higher than the standard rates.
3309 So of course there has been -- and the interim rates are at about the same level as the permanent rates that they proposed back in December, more or less the same magnitude of increase. So essentially by requiring us to pay the interim rates starting next week, they would be they would be giving themselves the outcome of this decision, which is to approve the rates that were filed back in December.
3310 That is one issue, but certainly to the level --
3311 THE CHAIRPERSON: You are a member of CNOC and you endorse the CNOC proposal --
3312 MR. COHEN: Yes, we do.
3313 THE CHAIRPERSON: -- and Distributel, as a large TPIA user, has these two extra concerns that didn't come out in the CNOC presentation, as I understand it.
3314 MR. COHEN: Yes, that's right, just because we are more closely tied to TPIA.
3315 THE CHAIRPERSON: Okay. Thank you.
3316 Michel, I believe you have some questions.
3317 COMMISSIONER MORIN: Yes, thanks, Mr. Chair.
3318 My questions are mainly for Primus, but feel at ease to jump in if you want.
3319 You are open to an alternative model, and since the beginning I have raised the MTS Allstream model. What do you think of this model?
3320 Do you think, for instance, that they are offering enough tiers for the small ISPs?
3321 MR. STEIN: I would have to verify exactly which tiers they are offering.
3322 COMMISSIONER MORIN: They have three.
3323 MR. STEIN: As I understand it, looking over my colleague's shoulder here, it seems to say "per 32 megabytes of peak". That is certainly -- in fact, that is even a smaller, more granular tier than I was suggesting. I think that would be enough.
3324 MR. STEIN: Is your question regarding the alternative proposal for peak billing that I was talking about, where the AHSSPI size would incorporate the usage component?
3325 COMMISSIONER MORIN: No, because the MTS Allstream model is about capacity only, not usage. There is no usage component in the model proposed by MTS Allstream.
3326 MR. STEIN: Yes, we are firmly in favour of a capacity-focused approach, one where we would pay for the peak capacity that we draw from the network.
3327 COMMISSIONER MORIN: But if the CNOC model is not adopted by the Commission, do you think that the MTS Allstream model would make sense for you, that it would be an available alternative?
3328 Because it is different, you just buy from the first day of the month which capacity you want, and the responsibility is yours.
3329 MR. STEIN: Yes. I would like to come back during rebuttal, when I can address that more thoroughly. I would like to evaluate it a little bit more closely, in light of your question.
3330 COMMISSIONER MORIN: Okay.
3331 MR. STEIN: Thank you.
3332 COMMISSIONER MORIN: But, as a matter of principle, do you think the model that the Commission is used to could evolve toward another model, that we have to step forward to bring more responsibility to the small ISPs --
3333 MR. STEIN: Exactly.
3334 COMMISSIONER MORIN: -- because, up to now, it has always been the fault of the big ISP providers? Now, you will share some control over the networks, you and the big ones.
3335 MR. STEIN: Yes, absolutely. In fact, a major driver for coming here and talking to you is exactly that point. We would like the opportunity to control this. We would like the opportunity to innovate and to deliver different services.
3336 We are very limited by the incumbents' proposals to date, which are based on total volume passed as opposed to capacity.
3337 We very much are interested in being held more accountable for that amount of traffic and paying for that amount of traffic. Absolutely that is the case, yes.
3338 COMMISSIONER MORIN: So you are not afraid --
3339 MR. STEIN: No.
3340 COMMISSIONER MORIN: -- to make the right choice --
3341 MR. STEIN: No, I expect --
3342 COMMISSIONER MORIN: -- and if your forecast is wrong about the Internet and the capacity that you need, you don't care, it's your responsibility.
3343 MR. STEIN: Commissioner, I certainly care if I am wrong.
3344 COMMISSIONER MORIN: Yes.
3345 MR. STEIN: However, we make those decisions every day.
3346 As I said, at Primus we have a very large network of our own that we make these decisions on, and we have a large wholesale infrastructure.
3347 Even at previous times, before having our own network, we still were making those decisions about capacity to purchase, when, and so forth. Sometimes that capacity needs to be ordered three months in advance even.
3348 We all need to make these decisions, whether it was at the smallest ISPs that I worked for or the biggest ISPs that I worked for. The size is different, but relative to the problem and relative to the size of the company that you are in, it's the same.
3349 So you still bear risk, and these decisions are already being made.
3350 To directly answer your question, I am not worried that that will be an unfair burden to place on an ISP. An ISP should, and does today, make decisions about the capacity that they will need coming, whether it be through equipment purchases or, frankly, even with staffing, with the transport arrangements that they build --
3351 These decisions are being made today, and it should be done here as well. It should be the responsibility of the ISP.
3352 COMMISSIONER MORIN: Would you say that the system right now, with all of the small ISPs, is more mature in a way?
3353 MR. STEIN: More mature relative to --
3354 COMMISSIONER MORIN: Now you are able to make all of the decisions that, perhaps a few years ago, you were not able to do.
3355 MR. STEIN: I think that the industry is maturing. I would certainly agree with you there, but I would say that small ISPs, frankly, run some of the most sophisticated networks that I have seen.
3356 It has been brought up that Primus is also a member of CNOC. Some of the smallest ISPs in CNOC do some of the most innovative things that I have seen in the industry. I would say that they are very equipped.
3357 I really can't judge or gauge where they were years ago, but I would say that they are absolutely equipped today, as are we, as is Distributel. What we are seeking is --
3358 Frankly, up until traffic management a couple of years ago, we were managing capacity, as far as we were concerned, and then that was applied on us.
3359 The charging model was not charging adequately before. I didn't hear a lot of people --
3360 Or, I should say that it was perhaps not charging appropriately before. We are saying that we should be charged for the capacity that we draw from the network, absolutely.
3361 COMMISSIONER MORIN: When we look at the market share of the small ISPs, the result is not so evident that the competition has increased over the last few years. This is, I would say, perhaps artificial competition, because this competition has been created by the CRTC, at the end of the day.
3362 It is the CRTC that forces the big providers to offer a wholesale price to the small guys, but the market share overall is less than 10 percent, after many years of this new competition that we want to instil in the marketplace.
3363 MR. STEIN: Competition has been limited by lack of access to high speeds for a long time. Now, that will ultimately come. We are in this interim period right now, which has a lot of challenges to it, but competition had been limited and we are now, finally, coming into a period where we will be able to compete appropriately, at the full speeds that the early adopters are seeking today.
3364 COMMISSIONER MORIN: With respect to Bell's AVP proposal, what is your ability to predict monthly volume in advance?
3365 MR. STEIN: Very little. In fact, I would even tell you that users themselves, and consumers, have very little ability to predict in advance.
3366 They may be able to look at recent trends and so forth, but there isn't a correlation to the amount of volume and the peak itself, and because you pay for every gigabyte, every change has that direct impact.
3367 I don't see either correlation or a way to predict adequately long in advance.
3368 That said, the number isn't changing wildly on a per monthly basis, but I wouldn't say that that is indicative of established industry models that we can use to predict that volume. Those models are all based around the principles of capacity management, which are not based on volume at all. It has nothing to do with volume, it's all about peak.
3369 So the focus, globally, for the industry is around predicting that peak.
3370 COMMISSIONER MORIN: Those are my questions. I would appreciate it if you would give -- you know that I have made available many numbers since Monday, and today as well, about the different models and the different prices for the end user of the system, and I would appreciate your comments about what increments of capacity should be made available for the big providers.
3371 In other words, are those proposed by MTS Allstream sufficient for you to make a real business model in the marketplace.
3372 MR. STEIN: Okay. Thank you, I certainly will do that.
3373 COMMISSIONER MORIN: Thank you very much.
3374 Those are my questions, Mr. Chairman.
3375 THE CHAIRPERSON: Len...
3376 COMMISSIONER KATZ: Thank you, Mr. Chairman.
3377 Good morning. I want to take you to paragraph 29 of Primus' opening remarks. It actually begins to formulate in my mind a view that I want you to comment on.
3378 You talk about Primus currently having arrangements on a carrier-to-carrier billing arrangement, and you say:
"At Primus, we buy Internet transit from many of the companies in this room, and do so with them exclusively through a peak capacity arrangement."
3379 Is there a difference between how the facilities-based carriers would be treating you if they treated you as, in Primus' case I guess, a co-carrier?
3380 In some of the smaller CNOC players, it may be a co-provider.
3381 Is there a difference in treatment? I get the perception here that they are treating you all like customers, which is different from a carrier-to-carrier relationship, which you touch upon over here.
3382 And, if there was such a difference, what would it be?
3383 If you were a big player out there, how would they treat you differently, if you had your own infrastructure, your own network, your own business, and you were simply passing traffic back and forth, and one of you was in western Canada and one was in eastern Canada, as co-carriers?
3384 Is there a difference, in your mind, as to how you are being treated today, versus how you would be treated if you were treated as a carrier?
3385 MR. STEIN: To be honest, I don't think I have ever thought about it in that particular way. I will think about it and try to address it in reply, but my off-the-cuff answer is that, no, I don't see how they would treat us any differently than they do today.
3386 MR. HICKEY: If I may add, I think that if we were Bell and we came to the table to negotiate, or if we were in that realm, I think we would be treated differently than we are today, just based on the size and what you bring to the table, compared to bargaining power and such.
3387 COMMISSIONER KATZ: Presumably, Bell and TELUS have all sorts of arrangements across the country. Do they treat themselves the same way as they treat you when they come to the table?
3388 MR. STEIN: I don't know, honestly. You will have to ask them.
3389 COMMISSIONER KATZ: Okay. At paragraph 15 of Primus' opening remarks, you talk about the pipe, and you talk about the fact that your leasing capacity -- in one case you are talking about capacity of 1 gig, and you are saying that you can't have more than 1 gig because you are only buying 1 gig pipe.
3390 If you needed more, what would you be doing? Would you be throttling customers, imposing ITMP?
3391 How would customers who needed more capacity, en banc -- all of your customers -- get service?
3392 MR. STEIN: In fact, we have many of those, the 1 gig pipes. You can buy more than one 1 gig pipe. You could buy 5, 10, 50, whatever.
3393 There is a set price, and we are billed a set fee for each one of those today.
3394 So you could have that 1 gig pipe -- typically, as an ISP, you would have two connections, I would think, at minimum. Maybe the smallest ISPs would have only one, but you are looking for redundancy, so you would have two.
3395 So you might start with two 100s and go up to a 400 and a 100, and ultimately two 1 gigs, three 1 gigs, four 1 gigs, or, in our case, many 1 gigs.
3396 A number of ISPs have many, many of these 1 gig circuits.
3397 The customers, when they connect, are basically placed, if you will, into one of those pipes.
3398 COMMISSIONER KATZ: Okay, but if, collectively, you have X number of 1 gig pipes and you run out of capacity, do you pick up the phone --
3399 MR. STEIN: Yes.
3400 COMMISSIONER KATZ: -- and call somebody and say, "I need more right now," and they --
3401 MR. STEIN: Bring me another.
3402 COMMISSIONER KATZ: Right there, delivery on the spot?
3403 MR. STEIN: Well, no, it's hardly on the spot. It takes quite a long time, so we have to do that in advance. We order them, I think, 60 days in advance. We order them and, hopefully, they get there on time and so forth.
3404 But where your question is going is, what happens if it doesn't get there on time? Then you are right, we have to manage the network accordingly.
3405 We recognize that, "Oh, no, we might be out of capacity here," and the network, then, will naturally be restricting the amount of capacity that can be used by our customers at that point.
3406 We could employ economic ITMPs, technical ITMPs and so forth. We have many options available to us at that point.
3407 We don't run the risk, per se, of it ever going past that 1 gig. It will only use that 1 gigabyte per second --
3408 COMMISSIONER KATZ: But, as I understand it, to employ an economic or technical ITMP, you have to give your customers advance notice.
3409 MR. STEIN: Absolutely, but if things don't go from --
3410 COMMISSIONER KATZ: So what happens --
3411 MR. STEIN: First, the ISP probably should give their customers notice and say, in advance, "We, in cases of peak, will do the following," which I believe is the case -- it is already the regulation. We have to give them advance notice. And the ISP should be prepared.
3412 You see the traffic growing over time, and you don't go from 100 megabyte to 1,000 megabyte per second overnight, and you don't go from 1,000 to 2,000 overnight.
3413 So there certainly is ample time to inform and to prepare.
3414 COMMISSIONER KATZ: With regard to margins, which you talk about in paragraph 8 of your remarks, you say that there should be symmetry, but you also say toward the end of paragraph 8:
"It must be recognized though that these consistent, fair, and reasonable mark-ups should be set at a level that allows a vibrantly competitive market to form on both service types."
3415 MR. STEIN: That's right.
3416 COMMISSIONER KATZ: The fact that you say "on both service types", are you suggesting that if they are not competitive at the wholesale level, we should be considering different margins in order to balance them and make them equitable, if I could call it that?
3417 Is that what that means, or does it mean something else?
3418 MR. STEIN: I think I would have phrased it differently.
3419 COMMISSIONER KATZ: Go ahead. I am not giving evidence here, you are.
3420 MR. STEIN: Really, what I am saying is that, in principle, it should be the same. It makes sense.
3421 You have a view to all the costs, we do not. You look at it, you establish that these are the costs that are different. And if you were to apply a consistent markup across both, and one was found to be so expensive that a company as large as Primus has sat before you several times and said, "No, we are not interested. We are not going to go down that road because we can't make money with that approach," then I am encouraging you to take a look and say, "What is happening here is, for whatever reason, the costs that we are viewing must be different somehow."
3422 At the end of the day, delivering broadband by the incumbents for these two service types is competitive. Why is it not competitive in the wholesale arena? Why is the pricing for the, vaguely, around 10 meg service so wildly different? Why is the capacity so wildly different? Why are the cost structures so different?
3423 I would then encourage you to say: If we can't get access to that vibrantly competitive market with symmetry in costs, then we may need to effect those costs, or we may need to, rather, effect those markups to ensure that that vibrant competitive market forms, because that's what Canadians need.
3424 COMMISSIONER KATZ: Are you actually saying that? Do you offer TPIA today?
3425 MR. STEIN: We do not offer TPIA today. We offer DSL today, and despite having looked at TPIA many times, we have no current plans to offer TPIA.
3426 COMMISSIONER KATZ: Distributel, you do offer TPIA, because you bought Cybersurf. In retrospect, was it a good deal?
3427 COMMISSIONER KATZ: You have put your business case in front of us here, and you are basically saying that you are gambling that this Commission will change the dynamics of it. You wouldn't be putting it in front of us if you weren't basically saying: If I don't get a change, I am getting out of this business.
3428 That's the way I read it.
3429 MR. COHEN: I wasn't sure if your question was whether the whole Cybersurf deal was a good deal.
3430 The TPIA problems were just one of many.
3431 What we took on was not profitable. It was being sold at less than cost. We did make moves to improve on that, the things that were within our control; for example, doing a better job of aggregating traffic with larger pipes, using the 1 gigs, and growing the business makes it more efficient, and getting better pricing for Internet transit.
3432 But you can see that the price that the TPIA rate is set at is very substantial, two-thirds of the monthly operating cost. That is out of our control, and I guess we are just hoping that it is set fairly.
3433 We don't really understand why TPIA rates should have to jump by 50 percent from where they were traditionally to what the cable companies are proposing.
3434 COMMISSIONER KATZ: If you put Exhibit 2, which is existing TPIA rates and proposed or interim TPIA rates, side-by-side with DSL rates, is DSL a more viable business for you?
3435 MR. COHEN: To be honest, Commissioner Katz, I don't know all of the rates that have been proposed, but I suspect that the FTTN rates are significantly higher than the traditional GAS rates.
3436 The legacy GAS is viable, yes. I don't think that the FTTN rates are viable. I think they are of the same order of magnitude of the cable company rates.
3437 COMMISSIONER KATZ: Do you sell DSL in western Canada?
3438 MR. COHEN: I think we have about 30 or 40 clients in TELUS territory.
3439 COMMISSIONER KATZ: So when you go to a customer, one of those customers or a potential customer, do you pitch one over the other, or do you let the customer decide which one he wants?
3440 MR. COHEN: Generally, our business is a consumer business that is driven by advertising, and we offer both.
3441 In territories where we can't provide both, generally speaking, the DSL coverage -- the footprint is much larger, because it has been available on an aggregated basis.
3442 So if a customer calls in, in response to our advertising, and we can offer both, we will.
3443 Currently, though, there are always differences. If a cable customer calls in, they will typically say what they are interested in: They want cable. They want DSL.
3444 The speeds are also -- you know, currently DSL only goes to, in Bell territory, 5 megabytes per second, whereas cable is offered at 10 and 15.
3445 COMMISSIONER KATZ: Apparently it changed yesterday. Someone announced yesterday that Bell is now offering faster speeds.
3446 MR. COHEN: Yes, the FTTN service became available, and we have ordered one or two on an experimental basis. There are lots of ordering issues and things that have to be worked out, so we are going to learn about it.
3447 COMMISSIONER KATZ: Those are my questions, Mr. Chairman. Thank you.
3448 THE CHAIRPERSON: Thank you.
3450 COMMISSIONER PENTEFOUNTAS: Thank you.
3451 Good morning. You made a compelling case for peak traffic billing, and if I were Bell, I would go with peak traffic billing, if we were to assume that what you say is correct.
3452 Why wouldn't they? What is the motivation for them asking for volume-based and not peak traffic?
3453 MR. STEIN: I would be guessing at motivation.
3454 What they have told us is that there is just so much -- there is only so much capacity and this is a problem and they need to charge for it.
3455 But, really, what they have said is that this will create a situation where end-users will end up having this economic disincentive for use. I think, within that is the answer to your question.
3456 Bell's expectation obviously from that statement is that this will -- this system will force competitive ISPs to pass per gigabyte pricing and therefore all the pluses and minuses that it creates along to those end-users. Otherwise, how could it have actually been an economic disincentive for the end-user to use it, in Bell's words?
3457 So if that truly is their intent, then their intent is to put a structure in place where the competitive ISPs will obviously have to also charge per gigabyte and, in doing so, that will really complete it. Per gigabyte pricing will really be the only game in town. It's currently in Ontario, Quebec and pretty roundly accepted practice for the incumbents.
3458 So I think from their own words that sounds like their true intention, yes.
3459 COMMISSIONER PENTEFOUNTAS: Mr. Cohen, you want to take a round at that?
3460 MR. COHEN: It's just guessing. I have always been afraid that the total usage model is designed to discourage usage which would make applications like video, watching movies and things like that discourage usage costing.
3461 COMMISSIONER PENTEFOUNTAS: They put you guys in the same mould whereby you are all offering the same service. It doesn't allow you the flexibility to be innovative and offer a different approach than the telcos and the cablecos?
3462 MR. COHEN: That's right.
3463 COMMISSIONER PENTEFOUNTAS: Is that your answer?
3464 MR. COHEN: I would think so.
3465 COMMISSIONER PENTEFOUNTAS: Okay.
3466 MR. COHEN: I would think so.
3467 Also, I would suggest that what we really want to achieve here is that those who cost the most pay the most, not those who use the most because if your usage is outside of peak then you are not driving anybody's costs.
3468 COMMISSIONER PENTEFOUNTAS: Yes, interesting.
3469 And to get to the cablecos more specifically, and Mr. Cohen touched upon that and his relationship with Shaw, we haven't heard from Shaw yet but we did hear from the other cablecos. In speaking about peak billing their contention was that given the fact that cable network is shared in the neighbourhood -- now, correct me if I'm wrong -- that with a shared network traffic impacts all of the traffic, be it at peak or otherwise.
3470 Does that make sense to you and would that lend credence to the volume based pricing as opposed to the peak based pricing that you propose?
3471 MR. COHEN: What makes sense to me is that there are many different peaks throughout their network, many quite small serving areas and they will have differing peaks.
3472 But it doesn't make sense to me that round-the-clock traffic is a problem. The parked car analogy that was given yesterday did not make sense to me: A car parked in the right-hand lane in the busy hour is causing a big problem and in the slow time at night everybody can just go around it. I don't understand it.
3473 COMMISSIONER PENTEFOUNTAS: Well, if they are sharing a network be it at peak or could be otherwise, they are claiming you are putting pressure on their system and they have got to split the nodes and that incurs increased costs --
3474 MR. COHEN: It's true.
3475 COMMISSIONER PENTEFOUNTAS: -- and those costs aren't passed on to you.
3476 I mean you know the argument.
3477 MR. COHEN: My peak is happening at 10 o'clock at night, let's say, okay? That's when all my aggregated traffic has its peak and they have, you know, seven of their nodes that have problems at six p.m. and four a.m. or something like that, okay?
3478 Yes, it's true that --
3479 COMMISSIONER PENTEFOUNTAS: You are not contributing to that, basically?
3480 MR. COHEN: What is that?
3481 COMMISSIONER PENTEFOUNTAS: You are not contributing to that, is that what you are telling us?
3482 MR. COHEN: Well, I might be. I mean I might have a guy that's using a lot in that neighbourhood at that time and, no, that won't be reflected in my rate but a rate-setting exercise or, I should say, a rate structure selection is a choice between imperfect models.
3483 Ultimately, when you set a rate you take all the costs that are relevant to the category that you are looking at, in this case usage-generated costs, and you divide them by some denominator. That denominator might be total gigabytes passed or it might be the ISP's peak period demand.
3484 The question is which of those measures is going to be a better tool for capturing variation in the underlying costs? Which better approximates it? They are definitely all approximations.
3485 But the point is that the peak period measure will probably -- and I can't prove this scientifically -- but will probably do the better job of emulating the cost contributing factors because it's responsive. Most of the peaks are within a range of times. So if my peak is at 10 o'clock, you know, and the average of all the peaks is at 9:30, that's not going to be far off.
3486 So that's the issue, it's which of these imperfect measures better approximates the drivers of the underlying costs. I'm thoroughly convinced it's the peak usage without evidence, just intuitively.
3487 COMMISSIONER PENTEFOUNTAS: To continue on that point, the whole idea of gaming the system whereby you can sell to your clients -- and there Mr. Stein is jumping at the bit. Go ahead.
3488 MR. STEIN: Sorry, I have been expecting that.
3489 COMMISSIONER PENTEFOUNTAS: No, no, go ahead. You know where I'm going.
3490 MR. STEIN: Now, I sort of feel like I could answer any question in the world.
3491 COMMISSIONER PENTEFOUNTAS: You are helping me out there. You are saving us time.
3492 Go ahead.
3493 MR. STEIN: The idea of gaming the system --
3494 COMMISSIONER PENTEFOUNTAS: Yeah.
3495 MR. STEIN: -- I think it's a little bit silly.
3496 You know wouldn't it be terrible if we asked these ISPs to pay for the peak and when they pay for the peak we are actually going to charge them like they used that for 100 percent of the month? And then they are going to game the system -- this idea, they would game the system by using all of what we are going to charge them for. They are going to move traffic from the peak that we charged them for to another area that we are still charging them for.
3497 Because remember, like I said, the perfect traffic distribution is flat. The idea that 3 megs is more difficult to manage than a spike to 3 megs. We are saying charge us like it is flat. Charge us like the peak point we reach --
3498 COMMISSIONER PENTEFOUNTAS: Is flat.
3499 MR. STEIN: -- we use 100 percent of the time. So wherever I move it around, what I would really be doing is making more efficient use of what I purchase.
3500 In doing so, if I did move my peak down, I have made more efficient use of the ILEC network. I should pay less.
3501 Somebody who doesn't make as efficient use as me should pay more.
3502 I'm not sure if that was the question you were --
3503 COMMISSIONER PENTEFOUNTAS: That was the question.
3504 Compelling case. Good answer. Thank you.
3505 THE CHAIRPERSON: Tim?
3506 COMMISSIONER DENTON: Good morning.
3507 Mr. Cohen, in your last paragraph said that -- I shall edit -- mandated wholesale high speed access is the means to real competition. This brings to mind the debate that continuously rumbles in a subterranean way through our heads as to the legitimacy of the regulatory approach to generating competition from wholesale access.
3508 My learned colleague, Mr. Morin, has -- I think I quote him accurately saying the CRTC has created artificial competition and they are market shares low after years.
3509 One of the things that has just never settled and probably never will be settled is this question as to whether third party access competition is, in some sense, fundamentally legitimate because it's never going to get the support of Corcoran of the Financial Post or people of that ilk.
3510 So I think I would like to ask you to defend as best you can just why this regulatory approach is appropriate in the introduction of greater competition than market discipline in this country -- Cohen first, Stein next.
3511 MR. COHEN: We actually had to write up a piece to the Financial Post after they published one called "Fake Competition" after Decision 2010-632.
3512 The issue has already been decided, I think, by this Commission -- I think also by the government of the day -- that fake competition if you want to call it that is better than no competition. That's really you know, what it comes down to.
3513 Yes, I would rather that we had the wireless panacea that we are looking forward to when there are 10 facilities-based providers all providing, you know, 250 megabits per second to the home through the airwaves. It will be wonderful. We won't have to -- we will be able to negotiate with suppliers on commercial terms.
3514 But that's not the reality of the wireline competition market. So the choice is do we proceed with a duopoly or do we provide a proper framework for non facilities-based players to play? And that's what we are, you know.
3515 And if we want to revisit the decision, tell me now before I spend those millions of dollars on that business case.
3516 THE CHAIRPERSON: Let me assure you right now before we go totally off base here that is not up.
3517 The scope of this hearing is retail wholesale Internet service access and that the model for it. We are not here re-evaluating decisions that we have taken and the government has taken, as you quite rightly put that. Yes, we need the wholesale model of competition.
3518 MR. COHEN: I am sorry, Mr. Chair. I was just answering the question.
3519 THE CHAIRPERSON: No, I know, and I use the occasion to bring us all back to whereabouts what we should really concentrate on.
3520 Thank you.
3521 COMMISSIONER DENTON: I am still going to ask the question, Mr. Stein, just so we can have it on the record because -- go ahead.
3522 First of all, this is predicated on wireless competition coming in, in such a fashion, as there will be many facilities-based providers. Are you waiting for that day or do you expect it soon?
3523 When are we going to enter this period in which there is going to be many facilities-based wireless providers?
3524 MR. COHEN: We do have more than we had a couple of years ago and they do offer high speed Internet over the wires but it's certainly still not comparable.
3525 I can't predict when it will be.
3526 COMMISSIONER DENTON: Thank you.
3527 MR. STEIN: I would add on that point, it is certainly far from imminent when a wireless network will truly be able to replace both from a technical and an economic standpoint what is delivered by wireline.
3528 We are not one to two to three years away. We are several years away at best case, even with recent launches locally here.
3529 It's not just about faster speeds. It needs to be about those speeds at a price point with the usage and so forth with the same latency characteristics and so on, with the freedom to use it as people do.
3530 We are many years away from that point. We need more competition in the residential Internet market today.
3531 MR. HICKEY: Okay. And if I may, with regards just to the wireless arguments, when we sat here last summer and went through the speed matching decision, the President and CEO of Bell admitted at the time that LTE would always be behind wireline because of various spectrum constraints and technology developments.
3532 So there are various other reasons as well regarding wireless.
3533 COMMISSIONER DENTON: Thank you. Those are my questions.
3534 THE CHAIRPERSON: Candice?
3535 COMMISSIONER MOLNAR: Thank you.
3536 I am going to go back to the more mundane issues of how we structure these arrangements. And I want to -- I'm looking at graph 2 of the PRIMUS application. I just want to make sure I am understanding what you have proposed here.
3537 First of all, if you look at the three ISPs that you have put forward, can you just tell me what it is under your proposal they would be paying for?
3538 MR. STEIN: Sure. Let's use ISP A first.
3539 COMMISSIONER MOLNAR: Right, and just to be clear, I'm just speaking about the link charge that you are proposing, not the access and everything.
3540 MR. STEIN: Right, right.
3541 You will notice the overall graph has a maximum of 1,000 megabits per second if you look at the left side. In that case, ISP A would have some choices to make. They could simply purchase one 1,000 gigabit per second link. They could use that one. They will be charged as if they used it the entire time.
3542 They may also choose to purchase two 400s. If they purchase two 400s that wouldn't quite be enough so they also might purchase a 100.
3543 They have many options like this. Like I said, these exist today; simply change the fee.
3544 In the case of ISP B, they could purchase two 400s knowing that they are going to grow, and so forth.
3545 The point that I was really trying to make was that the peak is the high point for any of them. In the case of ISP C they are the only one with a perfectly efficient use. They would purchase the 400 megabit per second service and, in doing so they, by nearly using all of it, would be very efficiently using the service that they bought and it would be very highly incented to keep it as efficient as possible through whatever traffic management practices they could bring to bear.
3546 COMMISSIONER MOLNAR: Right. Just to follow up on something you said earlier, because I understand what these were intended to do in showing the different efficiency and capacity usage, but I'm trying to link it to exactly what, you know, the difference between what CNOC has proposed in being charged for that peak period and you being charged by links, how that would be a difference.
3547 Now, one I think I heard you say is potentially under your model you are paying for links that are for redundancy as well as for usage?
3548 MR. STEIN: No.
3549 COMMISSIONER MOLNAR: No.
3550 MR. STEIN: I guess I would say the price would probably be almost identical in the two models.
3551 The reason for this is the only real -- the fundamental difference in the model that I am proposing is not about whether you buy it a month in advance or you are billed a month later, which I am really neither here nor there on. It's not about whether you have one link or two and one further down. Again, that's not really it.
3552 It's about -- all I am suggesting is if it's truly that difficult to do what so many of us do and charge at 95th percentile or its other peak billing methods, if it is truly that difficult to simply add that to the billing system, well, okay, then; understood.
3553 Take something that already limits you to these capacity increments and just change the price to incorporate the capacity that you could ultimately draw. So I don't mean to make this more complicated but, really, the price would be almost the same.
3554 My proposal, or the proposal that we are tabling here, similar to the Allstream, extremely similar to the Allstream approach, has a situation where we are simply taking the cost of that capacity at 95th percentile or some other peak use calculation and are wrapping it into something we are already buying to make it easier to charge for.
3555 But we would still be paying what we effectively -- my suggestion -- we pay what we pay now for those links plus the usage that those links could drive in place of --
3556 COMMISSIONER MOLNAR: Okay, fair enough.
3557 Again, I'm just trying to make sure I fully understand the difference.
3558 MR. STEIN: Sure.
3559 COMMISSIONER MOLNAR: So effectively, the difference is that you are in charge of provisioning your capacity and a commitment to future capacity because you have ordered those links in your --
3560 MR. STEIN: Yeah, I'm in charge of --
3561 COMMISSIONER MOLNAR: -- whereas under the CNOC model that doesn't really exist.
3562 MR. STEIN: Under the CNOC model they are suggesting, as I have read it, they are suggesting that they also are provisioning their capacity. In fact, I would tell you every --
3563 COMMISSIONER MOLNAR: They are provisioning it but it fluctuates.
3564 MR. STEIN: They are provisioning it today. They may purchase way more than they actually will pay for capacity but they are provisioning these same circuits.
3565 The difference is that by adding in these costs to those circuits it will make them cost more. I think there will a need to be a little bit sharper on exactly what size of these is required for each network provider.
3566 COMMISSIONER MOLNAR: Okay, thank you.
3567 MR. STEIN: And certainly the pre-purchase that you mentioned as well, yes.
3568 COMMISSIONER MOLNAR: Right. So just a couple of more questions to ensure I am fully understanding this, the links today if I understand are not dedicated simply to residential high speed?
3569 MR. STEIN: They are not.
3570 COMMISSIONER MOLNAR: But they could be.
3571 MR. STEIN: I would tell you I don't think they should be but --
3572 COMMISSIONER MOLNAR: I understand you don't think they should be but, as you know, there is two parallel proceedings here, residence and bus.
3573 MR. STEIN: I know.
3574 COMMISSIONER MOLNAR: They could be. These could be, if necessary. These links could be --
3575 MR. STEIN: I am sorry, could I -- just so that I understand, there is a parallel proceeding you said on bus?
3576 COMMISSIONER MOLNAR: On business.
3577 MR. STEIN: Okay. Well, we can cover the bus side on the bus side but --
3578 COMMISSIONER MOLNAR: Yeah, I know.
3579 MR. STEIN: You could simply -- in fact, I will use an example in our own network.
3580 We do have some of these ISPs that we only put bus customers on today.
3581 COMMISSIONER MOLNAR: So you would be prepared to, if you needed dedicated circuits --
3582 MR. STEIN: Yes.
3583 COMMISSIONER MOLNAR: -- dedicated links to residence, then that would be reasonable?
3584 MR. STEIN: If I had to choose an ISP and put only bus on one set and res on another, I would be -- in doing that I would be denied a very effective, very commonplace in telecom approach, of taking the capacity that I have purchased and using it very effectively. You know, you can point to many examples over the years.
3585 COMMISSIONER MOLNAR: I do understand it would not be your optimal choice.
3586 MR. STEIN: That's right.
3587 COMMISSIONER MOLNAR: But you would be prepared for that?
3588 MR. STEIN: I would have to look at it closely but if that's --
3589 COMMISSIONER MOLNAR: Yeah, you need to think about it.
3590 MR. STEIN: -- I would be prepared.
3591 COMMISSIONER MOLNAR: Okay. I do have one more question, and the question is for Mr. Cohen.
3592 This model that has been proposed by PRIMUS as it regards the payment of usage it's, you know, billed on what you have said you support on the CNOC. It would work in your view within the cableco infrastructure as well?
3593 MR. COHEN: Yes, I think so. My only concern with the model is its granularity. When you actually measure peak usage you are more closely capturing the capacity that you are taking advantage of.
3594 With the pre-purchase of fixed increments of capacity, you get into a bit of a conflict because on the one hand you want to make sure that the facility you have in place has the capability to grow and, yet, at the same time, you know, when you are getting to 700 megabits and you have to buy a gig facility, you are going to be paying for a lot of extra capacity that you are not using until you need it.
3595 I would prefer to see that model separate the physical facility from a capacity commitment. It's possible to constrain a 1 gigabit facility to not pass more than 700 megabits per second traffic.
3596 So I would rather that I could put in a physical facility of a gigabit, pay for the physical facility and then independently buy capacity on that in, say, steps of 100 megabits per second.
3597 COMMISSIONER MOLNAR: Fair enough. I believe I understand why PRIMUS might have a position that would differ from some of the other CNOC models based on the size of the corporation and so on.
3598 But technically this can work. This can work as effectively within the cable company architecture as it does in the telco --
3599 MR. COHEN: Yes, absolutely. If you look at our Exhibit 1 you see all the different POIs and the connections to them are at a certain level and traffic can be measured on them.
3600 COMMISSIONER MOLNAR: Okay, thank you. Those are my questions.
3601 THE CHAIRPERSON: Okay. Before we let you go, Mr. Cohen, we have -- yes.
3602 Oh, do you? Go ahead.
3603 COMMISSIONER PATRONE: Thank you. Good morning. I will keep it brief.
3604 Mr. Cohen, if I heard Mr. Stein correctly, costs associated with the PRIMUS model and the CNOC model would be comparable or the same. Did you say that?
3605 MR. STEIN: That is the intention.
3606 COMMISSIONER PATRONE: Do you agree with that assessment, Mr. Cohen?
3607 MR. COHEN: As long as rates are calculated right. I mean if, as I said before, you will take the total -- totality of usage-sensitive costs and divide them by something.
3608 In Mr. Stein's case you would be dividing by a number of gigabit Internet capacity -- sorry, pipe sizes -- but the usage would be then spread over a slightly larger number because there would be many that would be only running at six or eight.
3609 COMMISSIONER PATRONE: But your costs as an operator --
3610 MR. COHEN: It would probably stay the same.
3611 COMMISSIONER PATRONE: -- similar or comparable under both --
3612 MR. STEIN: Again, if the rates were calculated probably based on the total -- the proper denominator.
3613 COMMISSIONER PATRONE: The idea of having the network operators charge wholesalers at peak level, which is your proposal, is that right, Mr. Stein?
3614 MR. STEIN: Right.
3615 COMMISSIONER PATRONE: If there is a network upgrade at the level of the operators, would that entail an automatic increase in the rates to the competitors if you are charging at that peak and if they have a network upgrade that allows them to increase their offerings?
3616 MR. STEIN: Not the rate, but the ultimate, sort of the tab issue, sure.
3617 COMMISSIONER PATRONE: Yeah.
3618 MR. STEIN: If, for example, the network was highly constrained and there was congestion everywhere and things are going very slowly and our customers were being impacted that way and then an upgrade were to occur in the ILEC network and all of a sudden far more traffic had passed at peak, it would and our bill would go up as a result.
3619 COMMISSIONER PATRONE: Which means that the costs to the independent ISPs would automatically go up, which would then be passed on to --
3620 MR. STEIN: Yes, exactly.
3621 COMMISSIONER PATRONE: Last question --
3622 MR. STEIN: If we use it, we should pay for it.
3623 COMMISSIONER PATRONE: Right, I understand.
3624 Do you have --
3625 MR. COHEN: May I add to that, please?
3626 COMMISSIONER PATRONE: Yes.
3627 MR. COHEN: In the case of cable, a speed upgrade to the end user will result in an increase in traffic necessarily to us because in the cable situation we do not have the ability to control traffic by end user. In the DSL environment, we do. The difference is whether or not we have the IP address of our customer.
3628 So in a DSL environment the telephone company could raise speeds, but we could rate-limit the customer and keep their speed where it was if we didn't want to incur that additional traffic. The cable situation, we would have to have more traffic from those end users because we can't limit them.
3629 MR. STEIN: Excuse me, sorry, I had interpreted the question perhaps incorrectly to be an upgrade in the middle of the network, again, sort of these major congestion points --
3630 COMMISSIONER PATRONE: I am not talking about congestion points, I am talking about --
3631 MR. STEIN: Oh, okay.
3632 On a per-user basis, I certainly agree with Mr. Cohen.
3633 COMMISSIONER PATRONE: And, Mr. Stein, you are also a network operator as well as a competitor, right?
3634 MR. STEIN: That is right.
3635 COMMISSIONER PATRONE: And do you have competitors riding on your network?
3636 MR. STEIN: Excuse me, to clarify, when you say "network operator," meaning incumbent?
3637 COMMISSIONER PATRONE: Your facilities.
3638 MR. STEIN: No, we are not facilities-based, we lease.
3639 COMMISSIONER PATRONE: Entirely?
3640 MR. STEIN: Yes.
3641 COMMISSIONER PATRONE: Thank you, those are my questions.
3642 THE CHAIRPERSON: Okay.
3643 Mr. Cohen, before I let you go, we have been trying very much to make sure that TPIA gets sort of equal use to ADSL, which of course it isn't right now. In our last decision we removed some of the incumbents.
3644 Do you see that, assuming the rates are properly set, there will be a shift from ADSL to TPIA?
3645 MR. COHEN: We have certainly seen it in our business, a lot of it due to the higher speeds that are available from cable. Perhaps with FTTN coming to bear, that may change. It is hard to say. A lot of it will depend on the practical impediments to implementation.
3646 Sometimes it is the little things that get you, you know, whether you have to send somebody out afterwards to finish the install, those kinds of things.
3647 THE CHAIRPERSON: If I understood some of the other interveners correctly, actually the change from one to the other is relatively major.
3648 And so if the availability and the prices are roughly the same, that is not going to produce a shift by itself? We will still have an overwhelming proportion using ADSL rather than TPIA?
3649 MR. COHEN: Each Star Choice seems to be different. I don't know why. For some reason we think that DSL is really difficult in TELUS territory. That has been our experience. You know, whereas in Bell's territory it hasn't been until now, but --
3650 THE CHAIRPERSON: Okay. Let me just put it bluntly. What would we have to do in order to even the load, make sure that most of the wholesale does not ride on Bell but it rides just as much on the cable companies?
3651 MR. COHEN: Can I get back to you on that?
3652 THE CHAIRPERSON: Sure.
3653 Mr. Stein?
3654 MR. STEIN: While reserving the right to add further during rebuttal I have two comments on that now.
3655 First, the rates as they currently exist and as they are proposed on cable are very high. And traditionally, one of the fundamental differences between the two approaches was that cable by its original proposal was somewhat disaggregated in that you had to go to all these individual places, whereas with DSL you can get it all at one.
3656 The get-it-all-at-one approach is great when you are just starting out because you can reach an extremely large footprint very quickly and do what you will with it. On the other hand, it starts to be constraining as you grow, which really is the essence of why we are here.
3657 In the case of DSL, through the GAS tariff and its ways, so much is being provided in that one consistent link. Disaggregating it allows it to be far more flexible. We talked about that last year.
3658 On the other hand, forcing the aggregation, which you have already done in cable, is going to go a major way. But without a major review of those rates and ensuring that those rates will create that competitive market, that is -- it won't go very far.
3659 THE CHAIRPERSON: Okay. Thank you very much, Mr. Cohen, for sharing your business case.
3660 But you don't mention at all -- and I see that you are positioning yourself against Shaw. But Shaw sells bundles and it is obviously an integrated company.
3661 Is there really still a separate market for Internet access or are customers more and more looking for bundles? And so therefore, that you may be more competitive on the Internet access is really not really relevant because they want one bill, want the whole thing bundled together.
3662 MR. COHEN: Bundles do have a very big role in the marketplace. In our experience in Eastern Canada we sell an Internet bundle with the VoIP telephone package and about 40 percent of the Internet subscribers take it, but the other 60 percent are taking standalone Internet.
3663 THE CHAIRPERSON: Okay.
3664 And lastly, Mr. Stein, you mentioned that you wanted the same approach for legacy as well as FTTN. Why?
3665 I mean, the legacy, I have heard from Bell and others, has been there, therefore a lot of people know it, they are comfortable with it, they have worked on it.
3666 Why couldn't we just adopt a new approach, be it your pipe approach or be it the peak-hour approach for FTTN and leave legacy as it is? It seems to be working. Why fix it when it is not broken?
3667 MR. STEIN: In a certain respect, I do agree. If it is not broken, why change anything?
3668 The approach that we favour bills us, as I said, for legacy and FTTN. I also said business and res, although I recognize that is not part of this. That is because we want to be held accountable for the total impact and we feel that by being held accountable we will be most efficient and we will make it more efficient and so forth.
3669 I would anticipate that once you strip out the very heavy usage component that is inherent in the GAS tariff today, I would expect that the price would go down.
3670 THE CHAIRPERSON: Okay. Thank you very much.
3671 We will take a 10-minute break before going to the next presentation.
--- Recessed at 1043
--- Resumed at 1058
3672 THE SECRETARY: Order, please. A l'ordre, s'il vous plaît.
3673 Mr. Chairman, we will now proceed with Shaw Cable Systems G.P.
3674 Please introduce yourselves for the record, after which you will have 20 minutes for your presentation.
3675 MR. BISSONNETTE: Thank you, Mr. Chairman and Members of the Panel. My name is Peter Bissonnette and I am the President of Shaw Communications.
3676 We are delighted to be here today with:
3677 - Jay Mehr, who is our Senior Vice-President of Operations;
3678 - Jean Brazeau, who is our Senior Vice-President of Regulatory Affairs;
3679 - Doug McEwen, on my right, who is the Vice-President of Engineering; and
3680 - Paul Cowling, on the far left, who is the Senior Director of Regulatory.
3681 After some introductory remarks, we will take you through our specific responses to each of the questions raised in the CRTC's letter of July the 4th.
3682 Shaw is committed to delivering an unparalleled Internet experience to Canadians. We want our customers to use the Internet more and make the Internet a bigger part of their lives. The effect of any alternative wholesale-billing model that may come out of this hearing must be to maximize the Internet experience of all customers, including retail and wholesale.
3683 The Canadian broadband market can and should lead the world in terms of quality and value, and in order to make that happen we need to encourage more usage, more investment and more innovation.
3684 It is clear to us from the last two days that the two primary models have emerged; there is the aggregated-volume-usage model and the peak-period-usage model.
3685 Shaw has proposed and prefers an aggregated-volume-usage model, but as Jay will discuss in a few moments, we are prepared to consider how to make the alternative model work.
3686 At Shaw we believe that Internet usage and the importance of the Internet to the Canadian consumer will continue to grow dramatically.
3687 Through our recent customer consultations, we had the opportunity to create a new Internet experience in collaboration with our customers across the country. Together, we have significantly enhanced our offerings with higher speeds, more comprehensive plans and more choice and flexibility.
3688 In May we completed a speed upgrade to our anchor Internet product from 15 to 25 megabits per second download and from 1 to 2.5 megabits per second upload.
3689 We also recently announced our broadband leadership strategy. Under this strategy our customers will have the option of staying with their existing plans and prices, but their current data levels will be more than doubled.
3690 Alternatively, customers may sign up for the new packages that will be rolling out over the coming months. Those new packages provide industry-leading value, data levels and speeds of up to 250 megabits per second.
3691 Our initiatives were certainly in response to customer feedback, but of course they were also driven by the intense competitive pressure we face daily in Western Canada.
3692 Canada's broadband market is uniquely characterized by dynamic competition among cable, telecommunications, wireless and alternative providers. As Minister of Industry Paradis recently acknowledged, it is largely thanks to the telecom industry and the investments we have made in keeping our infrastructure abreast of the incredible pace of technological change.
3693 Our goal is to find a reasonable path that will make the model work. There are three critical components.
3694 First, we need to ensure that we can realize a reasonable return on our investment.
3695 Second, the access rate for the wholesale service should vary by speed in accordance with our costs.
3696 And third, the access rate should reflect the shared nature of our network.
3697 This means that the dividing line or demarcation point between the access and usage portions of the network would be drawn so that all components of the CMTS are included in the access portion. We believe that this demarcation between access and usage best promotes usage.
3698 MR. MEHR: Turning to the Commission's first question, it is clear that there is no perfect billing model.
3699 In our written comments Shaw proposed an aggregated volume-based model where independent ISPs would subscribe to wholesale services at a cost-based rate.
3700 Under this approach the ISPs will have a number of benefits compared to the existing wholesale billing regime. ISPs can use the aggregated allowance to design whatever services works best for them. As a result, ISPs will have the flexibility to innovate on service design and price and can attract a diverse customer base.
3701 It is clear from the discussion that there is energy around a peak-based model, and we have worked hard to find a way to address the challenges of that model for us. The key challenge with the peak-period approach is that ISP customers may be causing peaks at a series of points in our network that will not be reflected at the aggregated point of interconnection.
3702 In addition, all usage on network has costs. In our experience, there is a strong correlation between speed packaging and customer usage. If a peak-period model included the critical criteria that we have identified, namely recovery of our investments, variable pricing by speed, an appropriate division between access and usage, that proposal could work, although it is clearly not an exact science.
3703 In going down this path, we would be assuming some risk that the usage pattern may vary from the speed.
3704 Moving to question 2, and as we indicated in our response to question 1, none of the proposed models are perfect. If the three criteria that we have identified are met -- recovery of investment, prices that vary by speed and appropriate delineation between usage and access -- we could make the 95th percentile approach work.
3705 There would be additional costs associated with implementing a peak-period model in comparison to a volume-based model. This is simply because we currently have the tools available to us to measure and bill for volume-based usage.
3706 MR. BRAZEAU: We will now turn to the other specific questions raised by the Commission and provide our responses to each of those questions.
3707 Question 3. We would support a billing model that encourages everyone to use the Internet as much as possible. Obviously, that will require significant investment. Given the shared nature of the cable network, a volume-based approach encourages responsible packaging. All usage on our network impacts other users and can cause congestion and drive up our investment costs. The independent ISPs' peak traffic may not be a complete story about where we must upgrade our network.
3708 In designing our Internet services, we are conscious of network utilization and capacity limits. Difficult compromises are always required when we make marketing and network investment decisions, which are complicated and risky. If the independent ISPs have no similar risk, Shaw will be forced to meet capacity needs driven by marketing decisions we have no control over. Now, let us be clear. Shaw is not asking for control over those decisions. We are simply observing that, given the shared nature of the cable network, the ISPs must face packaging risks, just like Shaw.
3709 Now, considering question 4, the Commission cannot lose sight of the following key facts:
3710 - The retail Internet service market is highly competitive.
3711 - Wholesale access is a conditionally mandated non-essential service.
3712 - Any mandated wholesale billing regime must uphold regulatory symmetry and technological neutrality to comply with the Policy Direction.
3713 For these reasons, there must be symmetry with respect to the mark-up on costs for all network providers offering the same services.
3714 Question 5. For reasons provided in response to question 4, the same mark-up should apply to monthly access and usage rates.
3715 It is important to emphasize that all aspects of the TPIA service are, and always have been, categorized as non-essential. Given the shared nature of the network, there is no basis for applying a different mark-up to access and usage costs.
3716 Finally, question 6. The principles determined by the Commission in this proceeding must be symmetrical and technologically neutral in order to comply with the Policy Direction's requirements. At Shaw, we are constantly investing in next-generation networks that lead the industry, including fibre to the home. Therefore, there is no basis for concluding that any of Shaw's services should be treated as "legacy services." In fact, our current Internet services are far more advanced than any next-generation services offered by the phone companies.
3717 MR. BISSONNETTE: In conclusion, Mr. Chairman and Commissioners, the key goals of this proceeding are:
3718 - to maximize the use of the Internet whenever and however it suits the customer;
3719 - to enhance consumer benefit and choice through innovation and facilities-based competition; and
3720 - to encourage the network investment that fuels more capacity, faster speeds and greater value.
3721 At Shaw, those goals guide our strategic decisions, operational execution and network investments. That is precisely why we took time to understand our customers' needs and designed innovative and consumer-friendly packages that lead the market.
3722 The wholesale-billing model that emerges from this proceeding will have to maximize consumer use of the Internet and independent ISP flexibility without discouraging or distorting facilities-based competition or undermining incentives to invest. In other words, the CRTC, we think, must focus on what matters most to consumers.
3723 We have identified three key criteria for any model to work:
3724 - we must be able to recover our investments;
3725 - we must be able to charge more for higher speeds; and
3726 - there must be an appropriate demarcation between access and usage.
3727 If these criteria are met, we believe it is possible that either the aggregated-usage or peak-period model satisfy the goals of this proceeding.
3728 We look forward to your comments and questions. Thank you.
3729 THE CHAIRPERSON: Now, let me start off by asking you what I asked Bell and the cable companies. If I adopt your proposal what does it do for consumers?
3730 MR. BISSONNETTE: Thank you, Mr. Chairman.
3731 You know, our focus in life is how we continue to offer a competitive product to our consumers and I think one of the things that has been clear to the Chairman is that we have a very, very competitive business in Western Canada.
3732 I mean TELUS is a very formidable competitor and to the extent that they now have a full bag of arrows, we need to make sure that our consumers and our relationships with our consumers is what's driving our behaviour.
3733 And to the extent that you can maintain some balance, if you will, in your decision, that doesn't bring an imbalance to our ability to compete, there's no doubt what we're trying to do is to satisfy what our customers have told us is important to them.
3734 They want to know what they're paying every month. They want to know that they have the best service that's suitable to their needs and that they're going to be treated in a respectful way.
3735 So what you're doing right now plays into, if you will, our overall suite of services and conduct with respect to how we satisfy our customers' needs. And they tell us what they want.
3736 THE CHAIRPERSON: Yes, but you tell me that you prefer a volume-based model but you could live with a peak-based model.
3737 In terms of benefit to consumers, which one brings greater benefit or are they neutral?
3738 MR. BRAZEAU: Well, I think both will certainly have benefits and I think our volume-based would probably -- gives the independent ISPs enough flexibility to package their own services to more effectively compete against us in the marketplace.
3739 And by doing so, by giving this flexibility, they will be able to provide a different suite of services, different prices, different packaging, different bandwidth to consumers.
3740 So I think at the end of the day all consumers will benefit from that additional competition from the independent ISPs, given our model.
3741 THE CHAIRPERSON: Yes. And your model will be encouraging more diversity, more options to consumers than the peak model?
3742 MR. BRAZEAU: I think both models provide the --
3743 THE CHAIRPERSON: That's what I asked.
3744 MR. BRAZEAU: Yes.
3745 THE CHAIRPERSON: I understand your -- you have always been clear that you are customer-driven, you want to give the consumer the best. I'm just wondering whether there's a difference which model we adopt in terms of customer benefit.
3746 MR. BRAZEAU: On the customer side, I don't think so. I think on the implementation side and on the cost side the peak model will probably -- we would incur higher costs in implementing that model given the billing challenges we would have for that.
3747 THE CHAIRPERSON: Okay.
3748 MR. BRAZEAU: But for the customer, the end customer, the advantage would be very similar.
3749 THE CHAIRPERSON: Now, I am still not quite clear on this one point which you make again on page 4 of your submission here where you say:
"The key challenge with the peak period approach is that ISP customers may be causing peaks at a series of points in our network that will not be reflected at the aggregated point of interconnection."
3750 We had a very helpful chart from Bell on day one, which suggested that the peak is measured here at the beginning, at the AHSSPI point, and yet congestion could be --
3751 You just heard me asking Primus and Distributel about this and they said, well no, the traffic comes here and in effect if there are congestion points at points 1, 2 and 3 on the chart, yes, it will be caused but it will be caused by Bell's traffic as well as ours.
3752 The key is what comes in up to a point. The rest of it is really part of sort of normal management of a network, and, you know, so much will be your traffic, so much will be others, but those are sort of predictable consequential downstream effects.
3753 It sounds logical to me but I don't operate networks and you do. So therefore, I want to ask you what is your position on this?
3754 MR. MEHR: I think the discussion has evolved in a way that people are getting a great understanding of network. I think what folks might be missing is the magnitude of the two pieces.
3755 Perhaps to use a practical example, the Lower Mainland in Vancouver, including everything from Whistler to Abbotsford, would have a single point of interconnection in our network. Of course, the vast majority of our network is downstream of that point.
3756 We serve well over one million households out of that point of interconnection and so the opportunity for congestion around the node that serves the area around the University of British Columbia is very different than the opportunity for congestion in White Rock and it's going to happen at different times in the day and cause different things to occur.
3757 So I think there is recognition that peaks are still the issue. They just exist at various parts of very diverse networks.
3758 I think what we were saying is if we needed to go there we're prepared to assume the risk that that usage is still going to correlate mostly with speed, and therefore higher-speed customers will use more and that could be built into the access piece of the model as opposed to the usage piece.
3759 THE CHAIRPERSON: But taking your example of Lower Mainland and UBC, would you be measuring peaks at all those three points too? No, you wouldn't. You would just make a sort of prediction on what's happening at those points given what's coming in by way of speed; would you not?
3760 MR. MEHR: Yes, and what is happening will be quite diverse depending on how customers use their service, right?
3761 THE CHAIRPERSON: Right.
3762 MR. MEHR: So what the peak measure gives you is it gives you the overall net capacity on the transport system coming at that part of the network, and there are certainly some commonality in terms of peaks.
3763 So, for example, most network peaks will be in the evening between 9:00 and 11:00 and will sort of coincide with video streaming as a primary application.
3764 There are a number of other applications that don't follow those same patterns and depending on the blend of users in your network and what they're using it for, you could end up with very different peaks.
3765 That's the difference in the model.
3766 That having been said, as we've said throughout, there is no perfect model and the problem is if you try and measure at 20,000 different points, you're just going to cause complexity as well.
3767 THE CHAIRPERSON: And the Primus fallback model that we just heard about five minutes ago, is that sort of a halfway house?
3768 MR. BRAZEAU: Well, I am not sure if the Primus model would work for TPIA. We don't have the equivalent of links coming into our network at the POI and we don't charge like the telephone companies do.
3769 So I'm not quite sure -- I guess there is a connection there and there is a volume of traffic there and we would measure that but we would have to change the way we price our services because we don't charge for those --
3770 THE CHAIRPERSON: I thought I heard somebody say -- but I mean you just -- maybe in your written reply when you come back next week you could reflect on it, whether this is workable. I appreciate you just heard about it a few minutes ago.
3771 MR. BRAZEAU: Sure.
3772 THE CHAIRPERSON: Okay, Tom, you have some questions?
3773 COMMISSIONER PENTEFOUNTAS: Yes, thank you. I will be brief.
3774 Good morning. I noticed a substantial degree of flexibility with respect to the billing models. You know, much of the debate over the week has been volume-based as opposed to peak-based and I noticed a flexibility on your behalf and that runs contrary to what we heard from your cable colleagues earlier in the week.
3775 One issue that was raised was the fact that, you know, given the fact that cable networks share a neighbourhood and that with shared networks all traffic impacts all other traffic, be it peak or otherwise. Do you want to sort of comment on that issue?
3776 MR. BISSONNETTE: Well, we appreciate you seeing the flexibility. This is --
3777 COMMISSIONER PENTEFOUNTAS: Well, it may be just me, but I see it.
3778 MR. BISSONNETTE: We have been here in attendance for three days and we have heard the various arguments, both pro and con, for peak. And I guess if you really took -- you know, if you wanted not to do peak and you said that I would have to measure at 25,000 points to make peak very accurate, so it's how accurate is peak going to be.
3779 To the extent that we do do peak measurements right now on our network, we have some infrastructure that we think by augmenting that infrastructure we would be able to provide a reasonably close proximity to what actually is happening on the network, that, you know, with some consideration and compromise we could make that work.
3780 And we think that the difference between the aggregated volume basis and comparing that to peak may not be all that different anyway at the end of the day if in fact we have those facilities in place to do the measurements in a meaningful way, with of course the principles we described in here, which is, you know, there's a return on our investment, that speed is involved and that the demarcation is appropriate.
3781 COMMISSIONER PENTEFOUNTAS: You want to be compensated for the costs of implementation obviously?
3782 MR. BISSONNETTE: Yes.
3783 COMMISSIONER PENTEFOUNTAS: And they are there and they're significant from what I understand?
3784 MR. BISSONNETTE: Yes. And there's an IT side too, as you know. So all of the costs associated with the augmenting would be -- you know, would be fair game, if you will, in terms of cumulating the costs to do that.
3785 COMMISSIONER PENTEFOUNTAS: Including node segmentation --
3786 MR. BISSONNETTE: Yes.
3787 COMMISSIONER PENTEFOUNTAS: -- that will be caused by different peaks?
3788 Do you see any value -- you also mentioned at some point in your presentation that you would like to see independent ISPs share some of the risk.
3789 Have you sort of worked on a model of how that could come to fruition, how that can become a reality?
3790 MR. BISSONNETTE: No, but we can certainly provide something.
3791 But, you know, the risks are associated with the way you package your services. I mean we went through this, frankly, when we were -- well, after we met with our customers.
3792 And we said, so how do we provide them with what they really want, which is, we'll call it, a non-UBB regime where you price things at an appropriate spot for customers and give them sufficient choice that they can choose how their behaviour fits into what the packages are that we offer.
3793 And there's a certain degree when you offer unlimited services, maybe not so much in the light speed because that's almost self-limiting, but on a very fast 250-megabit service with unlimited capacity, in order for us to offer that service and to provide the services on a reasonable basis, we have to invest, first of all, about $100 million over the next 16 months to actually accommodate the bandwidth and the capacity to provide that service.
3794 There's a risk to that. So if we didn't offer unlimited services, then we would mitigate -- the risk wouldn't be as great as what we're doing now.
3795 The ISPs have a responsibility as well. I mean if an ISP wants to offer an unlimited service, you know, there are costs associated with that and there's a responsibility if they offer that package to manage their side of the network in a really -- in a very, very defined way so that they understand what impact their customers are having on the network and the costs associated with that.
3796 COMMISSIONER PENTEFOUNTAS: So that kind of shared risk would be more in the sense of managing their network and not necessarily becoming, let's say, investment partners with you on your network? That issue was also raised.
3797 MR. BISSONNETTE: That's ours. I mean we recognize --
3798 COMMISSIONER PENTEFOUNTAS: When you talk about sharing risk, that's the kind of risk you're talking about?
3799 MR. BISSONNETTE: Yes.
3800 COMMISSIONER PENTEFOUNTAS: Okay.
3801 Is there any -- do you see any value in -- you know, we talked about the different models and we also talked about one aspect of the Bell model to have prepaid and having the independent ISPs predetermine or guesstimate what their usage would be.
3802 How would that approach suit your concerns?
3803 MR. BRAZEAU: Well, with the model we are suggesting there's no -- you don't have to buy any packages of utilization or bandwidth or anything of that sort. So it would be on a per-gig basis and there would be some allocation at the beginning, but then any additional charges would be based on a per-gigabit basis. So that's our model. I think that's the model that works the best.
3804 COMMISSIONER PENTEFOUNTAS: I appreciate that.
3805 Have you actively pursued the independent ISPs? Are you trying to pursue that clientele? Do you see it as a clientele that you would like to incorporate in your business model?
3806 MR. BISSONNETTE: Frankly?
3807 COMMISSIONER PENTEFOUNTAS: Please.
3808 MR. BISSONNETTE: You know, we have 1.8 million customers, as Mr. Cohen mentioned, and it takes all we can to continue to provide those customers with the services that they are requesting. And our door is always open to TPIA -- you know, people who want a TPIA relationship.
3809 We don't provide a lot of energy in seeking them but we have a sales team now on the SME side of the business who are actually knocking on doors and happy knocking on doors looking for opportunities to go, both from a TPIA point of view as well as from other business products.
3810 COMMISSIONER PENTEFOUNTAS: My learned Chairman said it's mandated. I understand it and it could be mandated with the welcome mat or it could be mandated with the full bag of arrows, as you mentioned earlier.
3811 MR. BISSONNETTE: No. I think people find us a reasonably good company to deal with, and when they come to us and we know it's mandated, then we do what we can to facilitate it.
3812 COMMISSIONER PENTEFOUNTAS: Well, thank you, Mr. Chairman.
3813 THE CHAIRPERSON: Just one clarification.
3814 Under your model, they would purchase a certain block and if they need more, they have to buy additional, but there would be a different rate for the additional than the original purchase, right?
3815 MR. BRAZEAU: No. It would be --
3816 THE CHAIRPERSON: Well, that's what it says on page 4 of your original reply. So you changed it? I just looked at it. It says here:
"If the independent ISP customer's aggregate usage at the end of the month exceeds the aggregate base amount of bandwidth and any access usage blocks that have been prepurchased, the independent would pay a per-gigabit that would be higher than the rate charged for the prepurchased blocks."
3817 MR. BRAZEAU: We've become more accommodating. The rate would be the same.
3818 THE CHAIRPERSON: The same price?
3819 MR. BRAZEAU: The same price.
3820 THE CHAIRPERSON: Okay. That's what I wanted to know.
3822 COMMISSIONER KATZ: Thank you, Mr. Chairman.
3823 Good afternoon. I have a couple of questions regarding your presentation this morning.
3824 You talk about three principles or three critical components. One of them has to do with speed and you say:
"...the access rate for the wholesale service should vary by speed in accordance with our costs."
3825 We've heard some parties say there is no difference in cost, it's primarily usage or throughput. Can you clarify what different cost components you would be facing?
3826 MR. MEHR: Sure. There is a tremendous amount of scaling that goes to provide the high-speed services that Canadians want and demand today, and so if you're looking at the difference between a 3-meg service and a 5-meg service I'm not in a position to speak to that because we don't really offer -- we have such a small portion of our customers that would have services at that speed or below.
3827 As you get to moving a 7.5-meg service to a 25- or 50-meg level, what that requires is a complete rebuilt of your network. It requires additional channels on the cable network to be redeployed, including moving those channels that currently exist today. It requires a different modem and router in the home.
3828 So when we talk about the sort of hundreds of millions of dollars of investment in the next period of time that we're undertaking, we're really undertaking it in service of delivering extraordinary speed to our customers.
3829 So if you base the cost on the point of the network beyond the interconnection point on speed, they will roughly vary with the same things that you have to do in order to add capacity.
3830 COMMISSIONER KATZ: But when you say in your remarks here that in May you completed an upgrade to your Internet product and you upgraded customers from 15-25 megs and didn't charge them any more money, presumably, although you had a capital cost, your actual operating costs were the same?
3831 MR. BISSONNETTE: The competitive environment, I think, we're trying to describe is -- and I don't know how closely you follow Shaw in terms of the market, but there was a period where there was a -- you know, and I've spent time with our analysts talking about this.
3832 There was a misperception that somehow our primary competitor had kind of leap-frogged us in terms of technologies, you know, they had an IP offering. And we've tried to explain to them that IP doesn't mean better pictures, better quality, et cetera, et cetera.
3833 So in order to differentiate ourselves, you know, we differentiate ourselves on customer service. We also have to differentiate ourselves on our cornerstone product, which is the Internet now. It's not video, which it was five years ago.
3834 The Internet is what drives our customers and provides us with the value perception of our service. And so to the extent that we did what we just did, we came out of a consultation, we said customers are looking for greater speed and we have been working -- like this is like the rosary, you say it every day.
3835 Fibre node segmentation is now part of our life. Every year we spend at least $30 million segmenting nodes because the capacity of the network is being challenged by the changing patterns of customers using now streaming video, going to Google TV, YouTube, Netflix.
3836 All of those guys are putting capacity constraints on our network. If our customers want to do that, we want to accommodate it and so we have been doing that. So every year we're spending, as I said, about $30 million.
3837 In addition to that, we've said we need probably to quadruple our network capacity in the next two years in anticipation of what's happening and that's this $100-million digital network upgrade.
3838 But there is an ongoing capital requirement in our business, and fortunately we were able to do what we've done in terms of increasing the speed from the work we had just done prior to that.
3839 COMMISSIONER KATZ: Okay. But presumably, based on what your evidence is, you've given us or you will give us costing sufficient to differentiate your speeds to warrant a difference in speed at the wholesale level?
3840 MR. BISSONNETTE: Yes, we will be filing that with you.
3841 COMMISSIONER KATZ: Okay.
3842 On page 5 you talk about additional costs associated with implementing a peak-period model, and I think you've been through it a couple of times now with other folks as well.
3843 My one question is: Do you currently charge people based on peak in other parts of your business?
3844 MR. McEWEN: We currently monitor for peak on our backbone networks and therefore would have the ability to charge for peak on any high-speed aggregated link to a large commercial customer.
3845 We currently monitor the large number of nodes and customers only at an hourly basis and therefore at that point in the network we don't have the ability to do the 95th percentile calculation today.
3846 COMMISSIONER KATZ: Are you familiar with your Shaw big pipe business?
3847 MR. McEWEN: We are.
3848 COMMISSIONER KATZ: Do you bill customers based on peak usage in that business?
3849 MR. McEWEN: I'm not aware of whether we do or not. I do know we have the capability there.
3850 COMMISSIONER KATZ: Could you find out --
3851 MR. McEWEN: Yes.
3852 COMMISSIONER KATZ: -- when you're back next week, I guess. Okay, that would be useful.
3853 And the last question is on page 6, just before your response to question number 4 you say:
"We are simply observing that, given the shared nature of the cable network, the ISPs must face packaging risks, just like Shaw."
3854 So you're, I think, saying here that your network is an integrated network and whether it's traffic coming from an independent ISP or from your own customers it does down one pipe?
3855 MR. BISSONNETTE: That's correct.
3856 COMMISSIONER KATZ: And therefore, if you have to use whatever measures you use, they would face the same reality?
3857 MR. BISSONNETTE: That's correct. We're basically one symmetrical body.
3858 COMMISSIONER KATZ: So this is very different than what I guess Primus advocated about an hour and a half ago, where they said: Charge us for the pipe. We're using the entire pipe and therefore there shouldn't be a problem because if we need more pipe, we'll order more pipe. But the pipe is the pipe.
3859 You're saying: There is no such thing as their pipe and your pipe, it's one pipe.
3860 MR. BISSONNETTE: That's right. It's one network shared by all customers. So we have, as Jay mentioned, 1.8 million Internet customers using various different patterns and passages and events but they all impact the network.
3861 And, as you know, at the lowest point of the network is our node and typically now we have between 600 and 700 customers who share that segmented node, and so within that 600-customer mix you're going to find every kind of apple and banana in terms of usage.
3862 COMMISSIONER KATZ: Okay. Those are my questions. Thank you.
3863 THE CHAIRPERSON: Candice?
3864 COMMISSIONER MOLNAR: Good morning.
3865 I appreciate that you've come here and are very accommodating as to let's find a solution, and you have 1.8 million customers and you're focused on meeting their needs and wholesale is not a large customer base for you. I mean that's what I have heard you say here today.
3866 But for these little wholesalers you're a very big deal to them. We had Distributel up before you and I'm certain that you heard their comments. One of their issues was the existing POIs versus moving them to new POIs and the impact that will have on their business.
3867 Do you have a comment on that? Is that an issue for you? Can you just leave them on their existing POIs today?
3868 MR. BRAZEAU: We were ordered by the Commission to aggregate the POIs. We did that and that's -- I mean that's the reality and we've done that.
3869 Also, we were required to provide GigE services and we did that.
3870 So I think that was all that we heard from Distributel.
3871 COMMISSIONER MOLNAR: So it's all in our hands if we order you to maintain existing POIs in some manner for some transition or to deal with the transition? It's in your hands?
3872 MR. BRAZEAU: We argued for that at the last hearing and we were -- the Commission took a decision that we should aggregate the POIs and so we did.
3873 COMMISSIONER MOLNAR: Are you particularly concerned as to how we might address this, given the large -- given the size of this issue on you?
3874 MR. MEHR: No.
3875 MR. BRAZEAU: No.
3876 COMMISSIONER MOLNAR: No. We can just address it?
3877 MR. BRAZEAU: Yes.
3878 COMMISSIONER MOLNAR: Thank you very much.
3879 LE PRÉSIDENT : Michel?
3880 COMMISSIONER MORIN: Thanks, Mr. Chair.
3881 Madame la Secrétaire, est-ce que vous avez... As I've done over the last few days, I have the numbers of your service in comparison with MTS Allstream, and now we have all the comparisons with Cogeco, Rogers, Videotron, Shaw, and all the hypotheses are there.
3882 I would appreciate your comment next week about those tables because I want to make sure that they are the real costs and I think it's important.
3883 As far as you're concerned, with the lower 15 megabytes, you're the first to be lower than MTS Allstream. As far as the 50 megabytes is concerned, you're -- as the other ones.
3884 The question I will ask you is if the alternative model is adopted and specifically the MTS Allstream, do you think that it will stimulate for you and the small ISPs more investment in the system than if we use the AVP model from Bell or aggregated volume?
3885 MR. BISSONNETTE: Well, we will file our comments with you, but I don't know that we can do much more investing than we're doing right now.
3886 MR. BISSONNETTE: I think, you know, the facts of life are that we're in an extremely competitive business and in that business, in order to continue to attract customers, we have to invest and we do that recognizing that there's a benefit to doing that, that ultimately there's going to be a return.
3887 But, you know, I don't know anywhere in North America actually who's investing in their Internet customers as much as Shaw is. And so to the extent that we have, frankly, a small portion of our customers who are wholesale, our behaviour isn't driven by the wholesale relationship. It's driven by our customers irrespective of whether wholesale or retail.
3888 COMMISSIONER MORIN: But if the small ISPs have to make their own decision about the capacity they want, it's an hypothesis but I think that perhaps they will invest more in their own networks because they will have to make their own decisions, they will have to take more risk. So if they invest more perhaps you will have to invest more in your own network, too.
3889 MR. BISSONNETTE: We don't see the TPIA as investing a lot in their networks. I mean, frankly, they go and somebody hooks up something that's already in existence, which is our network. So there might be one-time installation charges to hook up a customer or whatever, but there is no infrastructure that they have, other than the interconnect at the POI.
3890 COMMISSIONER MORIN: Okay.
3891 Those are my questions.
3892 MR. BRAZEAU: Just one point on the specific MTS, we will certainly respond in writing to it, but I think one of the elements of the MTS model is that it's speed agnostic and so that would violate our basic principle -- or one of our basic principles, which speed does matter and speed does have a cost. So we will give you further comments on that, but that is certainly one observation we have for right now.
3893 COMMISSIONER MORIN: Thank you.
3894 THE CHAIRPERSON: Marc...?
3895 COMMISSIONER PATRONE: Thank you. Thank you, Mr. Chairman and good morning.
3896 Your capacity to invest in your network -- because you spoke at length about this, Mr. Bissonnette -- would that be affected adversely in any way in the event that the competitors were to find your service, your TPIA offering, a lot more attractive than they currently do, in other words to the point where they took -- well, Rogers talked about 2 percent for them. I don't know what it is for you.
3897 MR. BISSONNETTE: We are just slightly under. I think we are about 1 percent actually.
3898 COMMISSIONER PATRONE: One percent. What if it theoretically went up to 5 or 10?
3899 MR. BISSONNETTE: You know, if the pricing, both the access and usage, were fair pricing and there was a fair mark-up on those, that would be a business for us. It's not something we shy away from. We are not disenabling the desire of TPIA people to come to us, they just haven't come to us.
3900 We don't have the footprint, sorry, on -- as you heard, of course the other factor is even TELUS in western Canada has a very disproportionate size to the wholesale as, say, what Bell would have in the east. There aren't a lot of wholesale residential kinds of customers that come to either of us.
3901 COMMISSIONER PATRONE: But they have told us -- we have heard consistently that they would happily any day of the week trade for a retail customer in exchange for a wholesale customer because there is simply more money in it on the retail side than there would ever be on the wholesale side, for obvious reasons.
3902 That's why I ask whether or not Shaw's capacity to continue its build out, at what you say is a rate that exceeds anybody else in North America, would be impeded in any way as a result of having a wholesale component increase exponentially over time?
3903 MR. BRAZEAU: Yes and no. If the principles that we mentioned about us recovering our costs then, as Peter pointed out, maybe we might encourage. That would be another customer base that we could certainly go after.
3904 However, if it's as what I think I have heard Distributel suggest this morning, that somehow you have to manage their margins and that somehow impedes us from adjusting our prices and not responding to the real competition we are facing in the marketplace, which is TELUS -- and certainly Jay can speak to that issue -- then yes, that would certainly seriously impede the amount of investment that we could make in our network.
3905 Again, it depends on the details of -- you know, the devil is always in the details, but how do we get there? If it is managing margins -- and resellers have been asking for that from the Commission from day one, to somehow keep the prices and their costs in proportion -- then yes, this would have a significant negative impact on investment.
3906 COMMISSIONER PATRONE: Okay.
3907 MR. BRAZEAU: Okay?
3908 COMMISSIONER PATRONE: Thank you very much.
3909 THE CHAIRPERSON: Just clarify one thing for me.
3910 You said making the model work you have three principles. First, reasonable return on investment; second, an access rate for the whole service should vary by speed in accordance with our costs; and third, the access rate should reflect the shared nature of our network.
3911 If I accept those, when then at question 5 we ask you: What should be the mark-up for monthly access and usage rate, it should be the same.
3912 How do you apply these principles? I mean monthly access is a one-time cost connecting some of the usages, obviously something that fluctuates depends on how much somebody is using.
3913 So explain to me how applying those principles you come to the conclusion it should be the same for both.
3914 MR. BRAZEAU: Well, I think it's a little bit the answer that Mr. Engelhart gave yesterday. We don't build an access network separate from a usage network, so it's the same network. Because of that, we think that the mark-up should apply equally -- it should be the same mark-up, I should say, for access or for usage because it is, again, the same network.
3915 THE CHAIRPERSON: That's because of the different network. Bell gave exactly the opposite answer, they said that it should vary -- the usage one should be variable, the other one not.
3916 So you have no problem with us adopting sort of one pricing model for ADSL and a different one for you?
3917 MR. BRAZEAU: They are very different networks.
3918 THE CHAIRPERSON: Okay, thank you very much. I think those were our questions.
3919 I guess it's 20 to 12:00, we will have an early lunch and we will start at 1 o'clock.
3920 Thank you.
3921 THE SECRETARY: Did you say 1 o'clock, Mr. Chairman?
3922 THE CHAIRPERSON: Yes.
--- Recessed at 1145
--- Resumed at 1305
3923 THE SECRETARY: Order, please. À l'ordre, s'il vous plaît.
3924 Ready to start, Mr. Chairman?
3925 THE CHAIRPERSON: Yes.
3926 THE SECRETARY: We will now proceed with the presentation of PIAC, Pubic Interest Advocacy Centre. Mr. John Lawford is appearing for PIAC.
3927 Please introduce your colleague for the record and you have 20 minutes.
3928 MR. LAWFORD: Thank you, Madam Secretary.
3929 My name is John Lawford and I am legal counsel of the Public Interest Advocacy Centre and with me today is Jean-François Léger, outside regulatory counsel, and our articling student, Roxanne Gunning.
3930 We are here today on behalf of the Public Interest Advocacy Centre and the Consumers' Association of Canada.
3931 We are here to speak to you as advocates for retail, principally residential, consumers. We are not telecommunications network operators or Internet service providers.
3932 In the organization and conduct letter of July 4th the Commission requested that parties structure their presentations in accordance with the approach provided. We have done so, but we note that the focus of the Commission's questions appear to be upon service providers, as suppliers and as users of wholesale broadband Internet services. This comes as no surprise since this proceeding focuses upon wholesale service arrangements. Service providers, however, are not consumers of retail services.
3933 We therefore welcome this opportunity to provide to the Commission a retail consumers' perspective. The response of thousands, indeed hundreds of thousands of Canadian consumers to wholesale service offerings deployed by incumbents triggered the launch of this proceeding and we thank consumers for expressing their views to you directly in such large numbers.
3934 The objectives of Canada's telecommunications policy as set out in the Telecommunications Act include several which focus upon the needs of retail users of telecom services, notably those set out in paragraphs (a), (b), (c), (f), (g) and (h) of section 7.
3935 We also note and will discuss the effect of the Policy Direction's requirement to rely upon market forces for the attainment of these objectives.
3936 In our view, however, the underlying purpose of policies, those that promote competition, is clearly the delivery of benefits of competition to retail users, which in turn supports the realization of the policy goals in section 7 of the Telecommunications Act.
3937 We have, as representatives of consumer groups, expressed, in this and other proceedings, our opposition to policies that penalize users based on their usage of the Internet. Canadian consumers and businesses should be encouraged, not discouraged, from making use of the growing range of applications, sources of information and entertainment made available by the Internet. Canadians should also be encouraged to develop innovative services delivered over the Internet.
3939 MR. LÉGER: Thank you.
3940 We have advocated in numerous proceedings in the last few years that the current market structure in the retail Internet services marketplace is dominated by incumbent telephone and cable companies and that this effective duopoly does not adequately protect the interests of Canadian users of the Internet. The Commission reached the same conclusion, we are pleased to note, in TRP 2010-632.
3941 In this proceeding, we are here to assist the Commission, mindful of the Commission's determination in its recent speed matching telecom policy that in the absence of competition from ISPs other than the incumbents and their affiliates:
"...retail Internet service competition would not continue to be sufficient to protect consumers' interests..."
3942 We are also mindful of the principles set out in TNC 2011-77 that:
3943 As a general rule ordinary consumers served by small ISPs should not have to fund the bandwidth used by the heaviest retail Internet service consumers; and
3944 That it's in the best interest of consumers that small ISPs, which offer competitive alternatives to the incumbent carriers, should continue to do so.
3945 We note as well the Commission's direction to parties that when they propose changes to the regulatory regime they should be able to demonstrate that such changes will:
"Ensure that Small ISPs retain flexibility and continue to be a source of innovation in the industry."
3946 Now, in response to the Commission's first question regarding the billing model to adopt, as consumer representatives we encourage the Commission to focus upon specific objectives. These are:
3947 1) The billing model selected by the Commission should maximize ISPs' ability to distinguish their own retail services from those of the incumbents, as well as from those of other ISPs to the greatest extent feasible.
3948 We submit that a billing model that builds into a basic charge a given amount of usage should be avoided. Charging wholesale customers for a given level of usage, whether the ISP's retail customer achieves that level of usage or not, is inconsistent with the promotion of customer choice and, more generally, with competition.
3949 The Bell companies have admitted in their appearance that they are seeking to use economic measures to influence retail customers' usage.
3950 While the Bell companies can choose to do this with their retail customers, the wholesale regime should not enable them to force this outcome upon independent ISPs' retail customers. The notion that a supplier should purport to be able to increase unit rates as demand goes up is clearly, in our view, a sign that market forces are not at work.
3951 We also believe that rates for wholesale Internet access services such as TPIA and the ILECs' GAS service or its equivalents should be cost-based.
3952 The notion put forward by some incumbents that their rates should be "market-based" baffles us, frankly. While we acknowledge that it's not the Commission's intention in this proceeding to re-examine its wholesale service regime, we also note that there appear to be no market forces to discipline the incumbents when they set prices for their wholesale Internet access services. Indeed, the customers for these services are the incumbents' competitors whom the incumbents would likely rather presumably see disappear.
3953 Cable companies have been quite forthright in this hearing to point out that they still see no need for independent ISPs.
3954 As we view the retail Internet services marketplace, the incumbent telcos and cablecos are equally incented to drive independent ISPs out of that marketplace and they have, we believe, been quite effective at achieving this objective. By the Commission's own account, independents now -- that is between all of them -- hold approximately 6 percent of the residential high-speed Internet services market.
3955 This is not a share which we believe is tenable and, as the Commission stated in TNC 2011-77 and in TRP 2010-632, independent ISPs are needed as a source of innovation and to protect the interests of consumers. As a result, the Commission cannot rely solely on "market forces" to determine rates for wholesale services such as TPIA and GAS.
3956 Third, wholesale tariffs need to promote the development of the independent ISP sector. We encourage the Commission to consider features such as volume and term discounts. We believe that such discounts could potentially provide an additional vehicle for ISPs to distinguish their service offerings from those of other ISPs.
3957 Incidently, in our written comments we have also suggested that the Commission consider productivity adjustments as time goes on. We have noticed that in their comments in this hearing a number of other parties have suggested likewise.
3958 We also believe that such discounts could promote consolidation. While we are strong believers in a highly competitive marketplace with several competitors vying for consumers' business, we also question whether hundreds of small ISPs can effectively compete in the long term with the entrenched incumbents. We do not expect that effective competition to the incumbents is likely to develop if the independent ISP sector remains fractionalized among hundreds of very small firms. Measures that allow market forces to promote some consolidation may be desirable.
3959 After a decade of diminishing independent ISP presence more wholesale customer-friendly services are needed from the incumbents. TRP 2010-632 was a step in the right direction, although it may still be too early to tell, particularly given the slow pace at which some of the Commission's determinations are being implemented.
3960 As we look back at the lengthy and contentious process that led to TRP 2010-632 and related disputes, it seems to us that a key element appears to be missing, that is the supply side of the debate.
3961 In retail markets, suppliers devote considerable attention to identifying their customers' needs and to developing services that meet those needs. In retail business markets, service providers seek to develop cost-effective solutions for their corporate customers. In their televised and print advertising campaigns, for example, the Bell companies regularly tout how they work closely with their corporate customers in developing services that contribute to these customers' business success. Other incumbents do likewise with their own corporate customers.
3962 Now, we heard the Bell companies and TELUS tell this panel how difficult it has been to establish constructive relations with independent ISPs. However, these ILECs' determined campaigns before the Commission and then before Cabinet to deny independent ISPs matching speeds cannot have helped, in our view, create a positive business climate.
3963 When we read the incumbents' annual and quarterly reports to their shareholders and investors, the focus appears to be entirely upon achievement of success in the retail marketplaces. Very little, if any, of the incumbents' considerable engineering, service development and marketing resources appear to be devoted to deployment of wholesale Internet access services. While it's no surprise that this is so, we invite the Commission to consider the establishment of incentives for incumbents to devote more effort and resources towards the deployment of successful wholesale services.
3964 In our written comments earlier this year we invited the Commission to consider developments such as the implementation of functional separation. While we are not advocating the implementation of functional separation at this time, there are potential benefits, we believe, inherent in separating, to some extent at least, an incumbents' retail and wholesale objectives and reducing the inherent tensions within an incumbent's organization between the pursuit of its retail and wholesale objectives.
3965 A decade plus of experience observing and often participating in the regulatory history of wholesale services such as GAS and TPIA suggests to us that the incumbents' retail objectives are considerably more important than the development of their wholesale businesses.
3966 The Commission should therefore keep close oversight of the wholesale Internet services marketplace. The Commission has taken what we consider to be positive steps with its matching speeds determinations and its directions regarding TPIA, but we see independent ISPs as vulnerable customers in a wholesale marketplace in which market forces are, in our view, inadequate to discipline the incumbents.
3967 Regarding the Commission's second question concerning usage measurements for the purpose of setting wholesale rates, we are unfortunately unable to provide much input.
3968 We note in this respect that very little cost information has been made available on the public record. Moreover, we appear before you as consumer advocates and not as ISPs or network operators.
3969 Regarding the Commission's question as to whether there are significant differences between the billing models in terms of impacts on incentives to build out networks and manage traffic, we submit that, consistent with the policy objectives in the Act and the Policy Direction, the Commission should, to the greatest extent possible, let market forces determine service providers' network build-out plans and traffic management strategies.
3970 Carriers such as the Bell companies have consistently been explicit with their investors and shareholders in identifying the incentives they pursue in relation to network construction.
3971 We submit that the increasing lead of the incumbent cable companies in the Internet services marketplace and the ILECs' search for competitive alternatives to the cable companies' TV offerings have provided the ILECs all of the incentives they need to engage in significant network improvements. For several quarters now BCE Inc., for example, has identified as part of its five "Strategic Imperatives", the improvement of customer service, what they refer as:
"Leverag[ing their] Wireline Momentum, and Invest[ing] in Broadband Networks and Services."
3972 TELUS and the other incumbents appear to be motivated by similar considerations.
3973 The Commission, in our view, should be sceptical of suggestions by incumbents that the deployment of wholesale services that provide wholesale customers a realistic opportunity to compete will dis-incent them from building-out networks.
3974 We believe that the strength of incentives for service providers to build out networks will likely be determined to a significant extent by their customers' ability to choose to migrate to other service providers if these customers are unhappy with their current supplier. We also believe, however, that a duopoly composed of the incumbent cable and telephone companies does not provide sufficient choice.
3975 We have been encouraged to see the Commission has also agreed that the interests of consumers would not be adequately protected by a duopoly comprised of the incumbents.
3976 Regarding the Commission's reference to "incentives" to manage traffic, we similarly believe that a vigorously competitive marketplace should provide all of the incentives ISPs need to manage their retail traffic.
3977 The Commission forbore from regulating retail rates over a decade ago and we do not think that the Commission should be adopting rules imposed by the incumbents that force ISPs to mimic one or another of the incumbents' own retail traffic management practices.
3978 Instead of encouraging ISPs to adopt what we consider to be counter-productive measures, such as the punitive UBB measures introduced by some incumbents, we believe that market forces should be incenting ISPs to offer more, not less, value to consumers for their hard-earned dollars.
3979 With reference to the Commission's fourth question regarding symmetry with respect to mark-ups for network providers, again, in the absence of any significant information regarding the network providers' respective costs, we consider it difficult to provide views regarding mark-ups.
3980 That being said, although regulatory symmetry is a guiding principle under the Policy Direction for economic regulation and in relation to network interconnection and access, some flexibility may not be inconsistent with these directions given the goal of the promotion of competition and the development of an environment in which supplier behaviour is dictated to the greatest extent possible by market forces.
3981 In short, symmetry may be important, but we also believe that the achievement of the policy objectives of the Act, to which we referred earlier, should be the Commission's ultimate objective.
3982 Regarding the Commission's fifth question, concerning mark-ups underlying access and usage rates, again without access to significant data regarding the incumbents' costs, we find it difficult to provide advice.
3983 We are, however, sceptical regarding the impact on competition and choice of relatively high access rates. One objective of this proceeding is the achievement of wholesale serving arrangements which maximize the flexibility enjoyed by independent ISPs to design and tailor their service offerings to their retail customers' needs. It's evident to us that the Bell companies, in their AVP proposal for example, are seeking much more than to simply recover their incremental costs inherent in the provision of access facilities.
3984 Regarding the Commission's sixth question, concerning the applicability of the principles set out by the Commission in this proceeding to new and to legacy services, we reject the contention put forward by some of the incumbents that fibre-based facilities were built at greater economic risk and wholesale services that use such facilities should therefore warrant a premium.
3985 As we see things, the ILECs have been building out fibre facilities largely to support IPTV and to remain competitive. The ILECs had little choice but to implement technology which would enable them to compete effectively with the cablecos.
3986 The investment risk faced by the ILECs was likely greater in doing nothing rather than implementing new technology. It's unclear to us why the incumbents should receive a premium for these facilities.
3987 In any event, and whether the Commission agrees or disagrees with our assessment of the risk faced by the incumbents, then imposing a handicap on independent ISPs if they try to compete with the incumbents in the marketplace for new and more advanced services as these become available is likely to prove counter-productive if the Commission's objective is to preserve an Internet service marketplace that is served by more than a duopoly.
3989 MR. LAWFORD: In closing, we believe that the Commission should also consider increasing its efforts in the area of research into marketplace trends, including retail prices, terms and conditions.
3990 Studies prepared by third party observers, such as those produced by the OECD, our own analysis in previous hearings and even research commissioned by the Commission, both on the public record in the TNC 2009-261 proceeding and in recent updates on the Commission website, have portrayed Canadian incumbent ISPs' performance as "middle of the pack" at best.
3991 We believe that timely, up-to-date information regarding the performance of Canada's ISPs would be useful in assessing the performance of Canada's Internet services marketplace. This would provide a valuable tool to ensure that Canada's policies are effective and efficient, consistent with the directives in the Act and the policy direction.
3992 Thank you. We would now be pleased to take your questions.
3993 THE CHAIRPERSON: Mr. Lawford, how can you repeat this middle of the pack legend? You know that there are all sorts of problems with the OECD data, and there have been several studies recently showing that it depends on what you look at and how you look at it, et cetera. We are actually doing quite well.
3994 So what is the use of repeating this?
3995 MR. LÉGER: We think -- and we have said this in our written comments -- that Canadians need better than middle of the pack performance.
3996 Now, we recognize that there have been challenges to the OECD studies. This is why we are inviting the Commission perhaps to consider devoting more resources to being able to generate the type of information that would permit closer scrutiny of what is going on in this jurisdiction and in others.
3997 We are not suggesting that the Commission isn't doing a good job at that now, but we are suggesting that there appears to be some controversy with respect to some of these -- particularly the OECD studies.
3998 Now, we also saw the Commission publish -- it was dated June 2011 -- the update to the Wall Study, and as we look at this study, we are not suggesting that Canada is at the back, trailing the pack, but, frankly, Canadians are increasingly losing the Internet for a wide variety of purposes, and we think that Canadians need better than service that is average, or prices that are toward the bottom -- from low to high. They are among the higher from a sample of advanced economies.
3999 We hope that a more competitive marketplace will achieve this, but we think that more data is perhaps helpful also.
4000 THE CHAIRPERSON: Basically, you are saying that we are doing okay, but we could do better.
4001 MR. LÉGER: That's correct.
4002 THE CHAIRPERSON: On page 13 of your presentation today you say:
"While it is no surprise that this is so, we invite the Commission to consider the establishment of incentives for incumbents to devote more effort and resources towards the deployment of successful wholesale services."
4003 What exactly do you have in mind?
4004 MR. LÉGER: We have struggled with this quite a bit. There have been a number of instances over the years where the Commission has looked at changes to stimulate the incumbent, typically the telephone companies, to develop more competition-friendly policies.
4005 We look back at a time when forbearance, for example, was held as -- well, I will use the word "quid pro quo", which is probably not a very accurate description, but the prospect of forbearance provided some incentive for carriers to develop more competition-friendly services.
4006 As we look at the current situation, we think that there may be opportunities, again, for the Commission to look at ways for the incumbents to put more effort into developing services.
4007 Now, I am not providing you any answers here, you have probably noticed. We are, frankly, struggling with what is left today for the Commission to be able to incent carriers, but things like a further degree -- we have spoken of functional separation. We suspect that the incumbents don't like that prospect too much and we are not suggesting --
4008 THE CHAIRPERSON: They certainly wouldn't regard that as an incentive.
4009 MR. LÉGER: Well, I think the Commission used to call these things incentives. I remember reading, in an earlier career, some of the Commission's decisions, and the way it characterized incentives sometimes had a negative connotation.
4010 THE CHAIRPERSON: But isn't it the reality that -- obviously, the telcos or cable companies regard the wholesalers as competitors, which they are, after all.
4011 And they will do what we ask them to do, what we oblige them to do, but I can't see that they would voluntarily go one iota beyond that, because it is to their own detriment.
4012 That's just the nature of the beast.
4013 MR. LÉGER: Yes, but I guess we need to look back at how -- I guess I will put it this way -- at how serious the Commission is about having a third option for consumers to go to.
4014 It has been, frankly, a pretty tough uphill struggle to date. We pointed out earlier -- and the Commission is well aware of this -- the 6 percent market share that these competitors hold, and we don't see a long future, with 300 service providers sharing that kind of market share.
4015 THE CHAIRPERSON: Okay. So you are not using the word "incentive", you are talking measures. You really want measures to --
4016 MR. LÉGER: The Commission itself has used the word "incentive" in the past, so we are using the Commission's language.
4017 THE CHAIRPERSON: It's a euphemistic expression. I appreciate that. I thought you had meant a positive incentive, and I couldn't figure out how this would work.
4018 Len, I believe you have some questions.
4019 MR. LÉGER: I'm sorry, could I add something?
4020 COMMISSIONER KATZ: Yes, go ahead.
4021 MR. LÉGER: If we look back at history, there have been times when the Commission has, in effect, directed the industry to develop capabilities -- number portability is one of them; arrangements for CO access -- which, in effect, gave the industry the responsibility to develop these arrangements.
4022 Now, the Commission needed to exercise a fairly close and strong control over what happened, so these things wouldn't drag on and fall off the rails, but there are precedents in which the Commission has turned to the industry, the incumbents and the competitors, and has set a task for them, kept close track, and in a number of instances the results were quite good.
4023 THE CHAIRPERSON: Yes. You are thinking of 911 and things like that.
4024 MR. LÉGER: There were a number of things that were done within CISC over the years successfully.
4025 THE CHAIRPERSON: Len...
4026 COMMISSIONER KATZ: Thank you very much, Mr. Chairman.
4027 I have to tell you that I was taken aback by your suggestion that we create a regulatory framework that would promote consolidation in the industry at a time when we are trying to create more competition, and you indicate that these discounts would incent people to come together, and you want us to do that?
4028 MR. LÉGER: No, we are not -- we need to be careful here. What we are concerned about is the existence of very large numbers of very small ISPs. We are proponents of greater competition, but we wonder how, in the long term, all of these -- I will call them mom and pop-type operations -- will be able to compete with the very, very powerful incumbents.
4029 What we are saying to you is that we have an expectation that, in time, there may be some benefit to having larger entities compete with the incumbents, and that there are perhaps some opportunities to promote that.
4030 But we are not suggesting that the Commission put a framework in place to compel consolidation, certainly not amongst the larger players, obviously.
4031 COMMISSIONER KATZ: But let the marketplace take care of itself. Those who want to get into the market are free to get in, knowing full well what the rules of the road are.
4032 But for us to create an environment that says, "We will give you volume discounts the bigger you are," basically goes counter to that, because now we are creating a regulatory fiat, a regulatory framework that would, in fact, penalize certain players in the industry through regulatory oversight.
4033 I don't understand it.
4034 MR. LÉGER: Let me speak to that a little bit. Volume and term discounts are not things that we have invented. This is common practice in the marketplace, in virtually every conceivable marketplace out there. So we don't see what is necessarily revolutionary about that.
4035 There are many, many precedents in telecommunications.
4036 Vendors that are motivated by market forces will, typically, entice customers to purchase larger volumes or to make longer term commitments by offering greater prices.
4037 We are saying that a side benefit of that may be that it would promote -- it may help promote some consolidation, and help create larger players.
4038 We are not asking the Commission to create a framework for that purpose, but we are saying that, as volume commitments grow or --
4039 We have heard a number of parties express concern about the concept of commitments to the business or commitments to technology, and here is an obvious example where having a greater commitment may produce benefits for the customer.
4040 And, in our view, again, only a side benefit may be to help encourage the development of stronger players, who can compete on their own and have the wherewithal to invest more in their facilities.
4041 COMMISSIONER KATZ: But you have to think about where all of that leads at the end of the day.
4042 MR. LAWFORD: Yes, we take your point, and I understand that.
4043 It also is largely in this submission because it is, if you will, a contrast to what I would think of almost as a reverse volume and term discount that the Bell AVP proposal is: You buy this much, and if you don't make it, you lose. If you don't buy enough, you lose.
4044 So to the extent that we are promoting this view, you could also look at Bell's as being the mirror image of this and trying to win on their end.
4045 Our note here is also just to have people think that in a regular market you would be having these sorts of term and volume discounts.
4046 Asking you to mandate it, I understand where you are coming from, but it is also in stark contrast to Bell. But I think they are really trying to do the same thing in their own way.
4047 COMMISSIONER KATZ: Do you actually believe that having term and volume discounts is pro-consumer?
4048 Would that be in the consumer's best interest at the end of the day?
4049 MR. LAWFORD: Our thinking was that, as Jean-François noted, there have been, for ten years or more, a number of small players, some of whom have gotten larger, but not terribly larger, not providing a credible threat to the incumbents, and they are here saying, "If you can fix the pricing, we will do better." We are saying that they will do better if you can fix the pricing, but perhaps you could encourage them to do even better and provide a quicker credible threat, if you have a few of them grow larger.
4050 As I said, if that is going a step too far, we are not wedded to the idea.
4051 COMMISSIONER KATZ: You stayed away from answering many of the questions that we posed, saying that you are representing consumers and you are not here to deal with the substantive issues of network topology and engineering, but I want to take you to page 7 of your remarks, where you say that the billing model selected by the Commission should maximize ISPs' ability to distinguish their own retail services from those of the incumbents, as well as from those of other ISPs, to the greatest extent possible.
4052 I think you have been following this hearing for the last couple of days. We had the cable companies in here yesterday, we had Shaw in here this morning, and they categorically said that it's one big, fat pipe.
4053 Unlike the DSL technology, where you can buy smaller pipes and different size pipes, this is one big pipe.
4054 So it's nice to say this, but we are faced with the reality, as well.
4055 Is there anything that you can offer us in terms of your views as to which way to go between these scenarios, because what you have proposed here would be great, if it could be done, but what we heard from the cable companies, subject to someone refuting it, is that it can't be done, they have one big pipe.
4056 MR. LÉGER: I am at pains to reiterate that we do not come here as network engineers. We have heard a lot of assertions this week. We, frankly, can't tell you which ones are correct and which ones are incorrect, as to what carriers can do in their networks.
4057 We, again, come at it from a consumer's perspective and, for us, the objective should be to provide wholesale services that preserve or maximize the flexibility that ISPs have to tailor their service offerings to their customers' needs. That is how they will compete successfully. They need to be able to distinguish themselves from the incumbents and from one another, and service arrangements that do that, in our view, help consumers.
4058 But is it the Bell solution, is it the CNOC solution, is it another specific solution? Frankly, we don't have the skills to be able to tell you, from a network standpoint, what can be implemented and what can't.
4059 COMMISSIONER KATZ: In your remarks -- and, again, I will direct you to page 13, paragraph 17 -- you touch upon functional separation, and you had a discussion with the Chairman about incentives, or negative incentives.
4060 The sentence in the middle says: "While we are not advocating the implementation of functional separation at this time, there are potential benefits inherent in separating...retail and wholesale..."
4061 And people have used the Ofcom model many times.
4062 Do you know of any country in the world that has the degree of facilities-based competition in markets, overlapping markets -- in the case of Quebec, you have Bell and Videotron, you have Cogeco overlapping, and you have other independents overlapping, as well -- where they have taken the step of finding the benefits of structural separation outweighing the cost, even if it could be done?
4063 MR. LÉGER: This will, maybe, sound a little cute, but, arguably, this is what the Americans did back in the early eighties, isn't it?
4064 Now, before you jump all over that, we recognize that the circumstances there were vastly different than what we are talking about here, but do we know another jurisdiction at present? Well, no, we don't.
4065 But, again, we think that we have some ways to go in Canada to have a high-performance retail Internet services marketplace that delivers prices and speeds that we think Canadians should be entitled to.
4066 So we are looking for outcomes, and we think that providing these incentives -- they are, admittedly, negative incentives, but providing these incentives may help ameliorate the situation.
4067 COMMISSIONER KATZ: But I would put to you that if there is a need for incentives, negative or positive, they should be reasonably achievable, as well.
4068 And I would put to you that functional separation, the way this country is structured, with the sophisticated, facilities-based competition that exists, is probably not attainable.
4069 MR. LÉGER: Again, we are not advocating it. We have suggested to the Commission that there may be forms of separation or some --
4070 For example, the carriers, or at least the incumbents, were required to establish carrier services groups at the onset of interexchange competition. There is evidence that creating these groups, although they were initially designed to protect competitor information, in fact, may have had more positive effects, because they also became a means of marketing wholesale services that might not have been as effective if these services were marketed as part of the carriers' retail organizations.
4071 Now, we don't know enough to be able to provide a precise assessment, but there are some precedents for types of efforts or initiatives that reduce the impact of the retail side of the company on its wholesale side, and I guess we are inviting the Commission to think about those.
4072 COMMISSIONER KATZ: Those are my questions. Thank you very much.
4073 THE CHAIRPERSON: I will ask you the question that I have asked everybody else. For consumers, do you see any difference between the two principal models that are before us, the AVP model and the other one?
4074 MR. LAWFORD: Yes, we do, Mr. Chair. We view the Bell AVP model as inferior, because it is more likely to dampen usage, and the CNOC model as probably more effective for consumers, although neither is ideal.
4075 I think that Jean-François has something to add.
4076 MR. LÉGER: We also recognize, however, that there may be some work to be done, in terms of assessing the feasibility of other models.
4077 However, at the end of the day, we urge the Commission to be driven by the flexibility that ISPs need to have to be able to design their service offerings in response to market forces.
4078 At this point we have a lot of concerns about the AVP proposal because of a number of elements, including what we perceive to be penalties for running more traffic, and the built-in usage for at least a component of the proposal.
4079 So, from that standpoint, we felt that, again, the AVP proposal seemed to be at one end of the continuum, and a number of others seemed to offer more flexibility -- potential flexibility for ISPs.
4080 THE CHAIRPERSON: Thank you very much. Those are our questions.
4081 We will take a five-minute break before we hear the next one.
--- Recessed at 1345
--- Resumed at 1354
4082 THE SECRETARY: We are ready to start, Mr. Chairman.
4083 So we will now proceed with the presentation of Mr. Kimberly Smith to present his intervention by videoconference from our regional office in Dartmouth, Nova Scotia.
4084 Good afternoon, Mr. Smith. Can you hear me well?
4085 MR. SMITH: I can. Can you hear me all right?
4086 THE SECRETARY: We can hear you very well.
4087 The panel is now ready to hear your presentation. You may now proceed. You have five minutes.
4088 MR. SMITH: Thank you very much.
4089 MR. SMITH: Thank you.
4090 I am not a technologist. I am a cultural worker so I have prepared a short, more philosophical statement to help with this thinking. I will just read what I have and then you can ask me questions about it after.
4091 I advise banning user-based billing because it thwarts Canadian cultural and economic activity at our grassroots which is definitely not good for any of us.
4092 CRTC regulations, you probably already know, must be designed to support the democratic and economic interests of all Canadian citizens.
4093 Our telecommunication system must empower civil society at least as well as it consolidates financial and political powers that be.
4094 The term "consume" must be stricken from all CRTC documents. Canadian citizens are people. It dehumanizes us to call us consumers, but more importantly it is fundamentally inaccurate. Canadian people don't just consume. Each one of us can also create, innovate, love, nurture. Need I say more about that?
4095 Asking industry players how to regulate is fair, but their advice is guaranteed to be skewed. Their motives are well known. They have a vested interest in maintaining and strengthening their power to make money by exploiting the attention of "consumers".
4096 CRTC must improve how it communicates to all Canadian citizens. According to Scott Murray, President of Data Angel Policy Research, Adult literacy is one of the primary determinants of economic health and social engagement. Our literacy rate hovers around 40 percent. This means CRTC has to work harder with video and audio messaging to engage more Canadian citizens in the important issues of regulation. This will facilitate inviting a more balanced and accurate perspective across Canada.
4097 At this point in the hearings, we already know that user-based billing for high speed Internet service has very little to do with congestion and perhaps more to do with competition.
4098 If it does -- if it does have something to do with congestion, I strongly advise the proof to be given to us. I support Michael Geist's little suggestion on user-based billing, his modest proposal. You have probably seen it online. I put a link on the page.
4099 Finally, I urge everyone to recognize the fact that net neutrality is good business for every one, not just the big businesses but little businesses like mine that operate in rural Nova Scotia helping to build community and engage people in our local economy like the local farmers market and other small little things that are happening in our community that used to be facilitated by public access television. Public access television, yes, through the cable networks that have been shutdown by the cable networks as they consolidate their holdings. And so, ironically, now YouTube is where we have public access television happening, at least in my community.
4100 So I think it would be very, very important to avoid user-based billing because it will essentially thwart this kind of grassroots cultural and economic activity.
4101 That's basically my statement. Thanks for your time.
4102 THE CHAIRPERSON: What do you understand by user-based billing? How do you use that term? Many people have used it before us meaning something different, so I want to make sure we have the same terminology.
4103 MR. SMITH: I am basing my understanding of user-based billing on the way Michael Geist suggested.
4104 I also think of it as the idea that beyond the charge that I pay for having access to the Internet, it seems to be another charge over and above that bill, kind of like my long distance charge on my telephone bill. It seems that if it gets to the point where when I want to put original video content out from my community to bill, every time I upload a video to YouTube or do any of those kinds of routine things that I need to do to engage my local audience, I'm going to be charged every time I do that and that may make it more difficult to do what I have to do.
4105 THE CHAIRPERSON: The caps that you are really -- you know but the principle of user-based billing is also used as a term that the more you use the more you pay, but the way you are being billed is dependent on the amount of content or use you make of the Internet.
4106 Do you have any problem with that?
4107 MR. SMITH: I do. I have a problem with that because what it does is it sets up an artificial kind of relationship between the network builders and owners and those of us who are involved in the business of content creation of being engaged in cultural activity like creating travel programs, music; Canadian content. All of that could be chilled if this kind of user-based billing system comes in.
4108 I suggest that Bell and the other large players in this business focus their attention on providing the best, most transparent network they possibly can. It's fair for them to charge us, you know, a base rate to use that system because they have to pay for building it.
4109 I also suggest that if they want to be involved in the business of attracting the attention of Canadian citizens and thereby earn advertising dollars, that instead of them being involved in creating television programs and music and all of that kind of thing, that they should instead develop excellent web portals like YouTube or Vimeo or any of these other ones.
4110 If they have done their job well and they have attracted Canada's content creators, artists, musicians and they start to post their material on their network on their web portal and then the audiences start to watch that, then they will earn more money and therefore be able to pay for their infrastructure. YouTube knows this very well.
4111 THE CHAIRPERSON: But I presume we have no problem with being charged different rates for getting access at different speeds. If your speed on yours is 10 megabytes download rather than 20; that you pay a lesser amount, is that concept that you are billed according to the speed that the facility provides, does that offend you too?
4112 MR. SMITH: In principle I understand the nature of that. I, idealistically, would like to see equal access and pay for every citizen regardless of where they fit in the business hierarchy.
4113 Although it may be as we go forward with this and as new technology is developed, it may actually become a moot point. As we already know with the fibre-optic pipe it seems to make it very inexpensive to have it available.
4114 So I would say it would be more an equal sort of sharing of every citizen in terms of paying for the building of this infrastructure.
4115 THE CHAIRPERSON: Marc, you have some questions?
4116 COMMISSIONER PATRONE: Thank you, Mr. Chairman.
4117 Hello, Mr. Smith.
4118 MR. SMITH: Hi.
4119 COMMISSIONER PATRONE: I would like you to expand a little bit more. You clearly favour a flat rate. Is that right, sir?
4120 MR. SMITH: Yes, I do.
4121 COMMISSIONER PATRONE: Could you elaborate a little bit more?
4122 You have addressed it to a point about the whole premise that heavy users -- or that, rather, that light users should not be subsidizing heavy ones. We have heard some express that sentiment but you feel that somehow that shouldn't be a driver as far as policy is concerned, going forward?
4123 MR. SMITH: Well, from the way I look at it, I think what we are talking about is an infrastructure that facilitates communication between all people in Canada. And the problem that we are faced with is a lot of the business thinking that we have now, the dehumanizing business thinking that turns people into consumers and just thinks in terms of us that way, is actually a financial mistake.
4124 I think that if the businesses that are involved in this work now go back and regroup and take a look at what happened when we had free-to-air television, the way those broadcasters were able to pay for their infrastructure was by making their signal equally available to all citizens in their area and then to be able to, through ratings, justify the advertising dollars they had on that free system.
4125 Now, we are in a world where we are all hardwired together. There are very serious civil implications in this because if, for instance, we get into a situation where, heaven forbid, our government becomes corrupt or something terribly wrong goes that way and the citizenry decides they want to stop that and they start talking about that and then the government just turns it off, that's not safe for a democratic due process.
4126 So it's not just simply business we are talking about. We are talking about civil society and we are talking about the health of Canada's culture and its ultimate ability to collectively create prosperity.
4127 COMMISSIONER PATRONE: Sorry, sir, we do have several -- yeah, I thank you. We do have several models that have been presented before us.
4128 Have you had a chance to monitor this proceeding up to this point and, further to that if you have, do you have any thoughts about the models that have been presented before us that are being considered by the Commission?
4129 MR. SMITH: I am really delighted to see the variety. I think that there isn't one yet that stands out as being the perfect solution.
4130 Part of me thinks that the idea of regulating the Internet infrastructure, the idea of these various networks may be that one of the areas that we need to look at is how those different networks interact with each other. So how does the Videotron network interact with Shaw's or with Bell's or all of that?
4131 COMMISSIONER PATRONE: Well, they compete against one another, sir, to a degree.
4132 MR. SMITH: I think it is in the public -- pardon me?
4133 COMMISSIONER PATRONE: Yeah, there is competitive pressures between them, but I take it you think that they should be cooperating or is that --
4134 MR. SMITH: What I'm getting at is I think there should be a regulation about how those networks interact with each other that is guided by the principle that the public should be able to communicate in an unimpeded way with each other. So if I happen to be in one network provider's territory and I'm communicating with a citizen in another network provider's territory, the regulations should make sure that that communication is unimpeded, that there isn't any change in the speed and direction.
4135 And so where the real competition comes up is in how intelligently each network provider builds their system, how transparent they make that system, how secure they make that system, how fast they make that system.
4136 But, you know, at a certain point all of this stuff has to connect together for the civil good.
4137 That's, I think, where the CRTC needs to think about it. It basically takes on the role of making sure everybody is able to communicate clearly and easily without any kind of tiered system.
4138 COMMISSIONER PATRONE: Okay, sir. I'm just going to throw one last question at you, and I do appreciate your comments.
4139 You said that you believe congestion is a red herring in this proceeding. Did I hear you correctly?
4140 MR. SMITH: I think it is.
4141 COMMISSIONER PATRONE: Are you on Eastlink?
4142 MR. SMITH: Another thing we have -- pardon me?
4143 COMMISSIONER PATRONE: Are you on Eastlink now, sir?
4144 MR. SMITH: No, I use Bell Aliant.
4145 COMMISSIONER PATRONE: Okay. Do you have any congestion there at all? Have you had any congestion?
4146 MR. SMITH: There have been times when the service was spotty but it's not really an issue to me.
4147 COMMISSIONER PATRONE: Okay.
4148 MR. SMITH: It's generally been stable and easily accessible and it has facilitated me in my video business.
4149 COMMISSIONER PATRONE: Okay. Well, thank you and good luck with your business, sir.
4150 MR. SMITH: Thank you very much for hearing me.
4151 THE CHAIRPERSON: Thank you.
4152 By the way, I think your quote on literacy rates is reversed. 40 percent of Canadians have literacy problems, not that only 40 percent are literate. It's the other way around, 40 percent have literacy problems in various degrees. I just checked it on the --
4153 MR. SMITH: That's correct.
4154 THE CHAIRPERSON: Okay. Thank you.
4155 MR. SMITH: In fact the number is a bit higher.
4156 THE CHAIRPERSON: Thank you for your submission.
4157 That's all, Madam Secretary?
4158 THE SECRETARY: No, we need to take a short break again, Mr. Chairman, please.
4159 MR. SMITH: Thank you.
4160 THE SECRETARY: Before the next videoconference.
4161 THE CHAIRPERSON: Okay.
--- Recessed at 1412
--- Resumed at 1421
4162 LE PRÉSIDENT : Madame la Secrétaire, commençons.
4163 THE SECRETARY: All right, Mr. Chairman.
4164 We now have Mr. Andrew Moore from the Regional Office in Montreal.
4165 Good afternoon. How are you? Can you hear us well?
4166 MR. MOORE: Yes. Can you hear me?
4167 THE SECRETARY: Yes, very well.
4168 MR. MOORE: Perfect.
4169 THE SECRETARY: You have five minutes, Mr. Moore. You may now proceed with your presentation.
4170 MR. MOORE: Thank you for having me.
4171 My name is Andrew Moore, I am an independent IT consultant and a so-called heavy user. The availability of affordable unlimited packages is of great personal concern and the outcome of this audience will ultimately affect the pricing and availability of those packages.
4172 Chers commissaires, veuillez noter que je suis aussi disponible pour répondre à vos questions en français.
4173 First, I believe that the reason put forth by Bell Canada and other incumbents for adopting usage-based pricing on the wholesale level are deceitful.
4174 Over-the-top services have been threatening Bell Canada's market share in sectors other than Internet services and usage-based billing was, from my point of view, introduced to their customers as a way to curtail OTT's growth. Some customers responded to that change by transferring their business to independents that were willing to offer them unlimited packages.
4175 This is, in my opinion, and the opinion of many other Canadians, the reason why we are sitting here today.
4176 The fact is, there is little to no correlation between what is being called usage and congestion. Congestion is a product of a surplus of data being transferred on a node relative to its capacity at a given time.
4177 Overall monthly usage has little to do with it. A user transferring 120 gigabytes in a month at a throughput of 0.36 megabytes per second has little to no impact compared to the user who, every week night during peak times, consumes 1 gigabyte of data at a throughput of 5 megabytes per second, for a total 20 gigabytes a month.
4178 It is, therefore, illogical to implement any type of usage-based pricing to prevent congestion when usage has nothing to do with it.
4179 To put that into perspective, let's compare data pipes to a highway system where the amount of data consumed is the distance travelled by its users and the throughput is the size of the vehicle used.
4180 A highway is congested when too many cars try to use it for its current capacity. The distance travelled in a month by the road's users has nothing to do with its current status. However, the type of traffic on that road -- small cars versus trucks -- and its users habits of using that highway at very specific times of the day, because they are going to work for example, is what contributes to congestion.
4181 Congestion is simply a product of convenience. Putting a limit on how far users can travel or charging per kilometre will not affect the convenience of using it at specific times. You will only see an increase of traffic at off peak periods while your congestion periods are still very well congested. The end result, heavy travellers will pay for casual travellers which only use the service while convenient. Heavy users will be subsidizing casual users.
4182 Also, usage-based billing and the new aggregated volume pricing both calculate usage on a per-subscriber basis when a subscriber itself cannot cause congestion. To me this is a direct attack on the consumer who chose to leave Bell and is in a decision driven by cost.
4183 Due to these problems, I believe that the industry standard 95th percentile billing method proposed by CNOC is the proper way to charge for wholesale last mile access. 95th percentile more accurately bills ISPs based on the real metric that drives congestion, throughput, and incites both the independent and the incumbent to invest in their network for two different reasons.
4184 The incumbent will be pushed towards increasing capacity in order to provide a way for the independent to consume more and therefore pay more.
4185 On its side, the independent will be pushed towards controlling their spikes more efficiently to drive their costs down. It's a win-win situation.
4186 The decision that will take place at the end of this audience will ultimately affect the competitive forces into play on the retail market. Therefore, I feel compelled to remind the Commissioners that competition is there to ensure that Canadian consumers hold the big end of the stick, not the big telcos.
4187 As demonstrated by OpenMedia Stop the Meter Petition, Canadians do not want metered Internet and a push towards a metered wholesale, regardless of the method of usage-based pricing used, will mean the end of unlimited Internet in Canada. The Canadian footprint on the web in terms of video content and other forms of content made by Canadians.
4188 Canada's place in the digital economy depends on the access of affordable Internet services. Therefore, I must remind the Commission of their mandate.
"The CRTC's mandate is to ensure that both the broadcasting and telecommunications systems serve the Canadian public."
4189 This is a direct citation from your own website.
4190 A decision for wholesale usage-based pricing would go against the interest of the Canadian public and could have grave consequences on the Canadian digital landscape.
4191 Thank you.
4192 THE CHAIRPERSON: Thank you.
4193 Just so we make sure we are talking about the same thing, usage-based billing, when you talk about it are you using the term generically or are you using it in reference to a specific billing method that Bell had put forward?
4194 MR. MOORE: In fact, in my opening statement I used the expression "usage-based pricing for wholesale services".
4195 THE CHAIRPERSON: Okay. Just so I have that right.
4196 MR. MOORE: So it's really a broad term for the wholesale services.
4197 THE CHAIRPERSON: Now, explain to me, you say:
"Usage-Based Billing unfairly penalizes so-called 'bandwidth hogs' which do not truly have an important impact on Bell's network."
4198 Why do bandwidth hogs not have an impact?
4199 I thought we have heard a lot here the whole week about installing a system measuring peak usage, because that's really what would best reflect how much something is being used, and surely peak usage would be driven by bandwidth hogs so it would actually show up that way.
4200 So I don't quite understand your statement here.
4201 MR. MOORE: Yes. The reason why actually bandwidth hogs is in quotations in my document is because I'm more referring to usage hogs, people that consume a lot of data.
4202 THE CHAIRPERSON: Yes.
4203 MR. MOORE: The thing you have to understand is that we do not consume bits, we do not consume bytes. On a network what is being consumed is network capacity --
4204 THE CHAIRPERSON: Right.
4205 MR. MOORE: -- which is measured in bits per second.
4206 The whole idea of usage-based pricing calculates on the wrong metric, which is why I support the 95th percentile. What you are using is not bits, bytes, gigabytes, what you are using is bits per second.
4207 THE CHAIRPERSON: Oh, so that's what your point is.
4208 Okay. Thank you.
4209 Tim, I believe you have some questions.
4210 COMMISSIONER DENTON: Really only one. Hi, and thank you for appearing.
4211 Your conclusion in relation to usage-based billing is that:
"It unfairly targets users who do not significantly contribute to network congestion. The minority of users gets penalized for congestion caused by the majority who use Bell's network during peak time."
4212 So really what you are saying, then, is that it's because of -- it's usage by many people at peak hours that causes the problem that needs to be solved.
4213 MR. MOORE: Yes. A user by itself -- and Bell actually said it this week, a user by itself cannot cause congestion, it is a huge number of users using the service at the same time which actually causes congestion. It's high usage by many users at the same time.
4214 COMMISSIONER DENTON: Yes. Please don't take it wrong, but this issue has been very extensively canvased this morning and in previous days and I don't have further questions of you because there is you and a lot of others who are in agreement with you so I don't think it would be useful to ask you any further questions. I'm fine with what you have said.
4215 Thank you.
4216 MR. MOORE: Thank you.
4217 I just want to add for the record, for everybody listening, that I'm not against the idea of paying for what you use.
4218 COMMISSIONER DENTON: Yes.
4219 MR. MOORE: I'm against the idea of the consumer being seen as using bits and bytes. That's not what we are using. What we are using is network capacity, that's measured in bits per second, which is why I'm against any proposal of usage-based pricing.
4220 COMMISSIONER DENTON: No, your message is entirely clear and we have received it.
4221 Thank you.
4222 MR. MOORE: Thank you.
4223 THE CHAIRPERSON: Thank you.
4224 Michel, I believe you have one question?
4225 CONSEILLER MORIN : Oui. Merci, Monsieur le Président.
4226 Vous dites que vous êtes un consultant en ce qui concerne l'Internet. Avez-vous pris connaissance de d'autres modèles de prix de gros à travers le monde, sur quelle base on facture habituellement? Est-ce que c'est les bits par seconde ou par l'usage, l'utilisation? Est-ce que vous avez certaines informations que vous pourriez nous donner là-dessus?
4227 M. MOORE : Oui. En fait, le standard de l'industrie en matière de transport, c'est vraiment les bits par seconde. C'est la méthode du 95th percentile qu'on appelle, que CNOC a proposée. C'est le standard de l'industrie.
4228 Ce qui se passe dans le cas de Bell Canada, c'est qu'il y a aussi le côté accès, vu que, oui, ils font du transfert de données, sauf qu'ils donnent aussi accès à une connexion à l'utilisateur même.
4229 Je n'ai aucun problème au fait que l'accès soit un coût fixe ou dépendant de la vitesse qui est donnée à l'utilisateur, sauf qu'en tant que telle, la méthode de transport, la période de transport doit être calculée dans une méthode qui est reconnue par l'industrie, et cette méthode-là, c'est le burstable billing au 95th percentile que CNOC a proposée.
4230 CONSEILLER MORIN : Et est-ce qu'il y a d'autres expériences semblables à travers le monde, à votre connaissance, sur cette base-là?
4231 M. MOORE : Pas à ma connaissance personnelle. Pas à ma connaissance. Non, malheureusement pas. Je n'ai pas d'information là-dessus.
4232 CONSEILLER MORIN : Alors, en tout cas, si jamais vous en aviez, je serais intéressé à en prendre connaissance.
4233 M. MOORE : Je vais faire ma recherche et je vais vous apporter ça la semaine prochaine.
4234 CONSEILLER MORIN : Merci beaucoup.
4235 C'est tout, Monsieur le Président.
4236 M. MOORE : Bienvenue.
4237 THE CHAIRPERSON: Thank you for participating. Those are all our questions.
4238 Madame la Secrétaire...?
4239 THE SECRETARY: Yes, Mr. Chairman.
4240 Just for the record, I would like to indicate M. François-Xavier Richard-Choquette nous a indiqué qu'il ne serait pas présent aujourd'hui. Il devait comparaître en personne et a annulé un peu plus tôt cet après-midi.
4241 Alors, c'est tout pour aujourd'hui.
4242 We will resume tomorrow at 9:00 a.m.
4243 Thank you.
--- Whereupon the hearing adjourned at 1432, to resume on Thursday, July 14, 2011 at 0900
- Date modified: