Why the public consultation on broadband (and telephony) was a bad idea.

August 20 marked the last official day of the CRTC’s 4-week online public consultation on the “obligation to serve.” It was initiated “to hear Canadians’ thoughts about their access to basic telephone and Internet services.”

That’s the stated goal. The truth is the consultation couldn’t make up its mind what it’s trying to be. It’s a shaky combination of crowdsourcing (to unearth policy ideas); surveying (to see what Canadians think); and marketing (to inform visitors what the CRTC thinks). Let’s look at the list of things that went wrong:

Methodology.An online, self-selected consultation has all the disadvantages of a townhall meeting or focus group and none of the advantages, particularly the ability of organizers and participants to engage in a real-time conversation. The 5 main questions asked of visitors are either far too technical or far too obvious; there is nothing representative about the responses; and the consultation was fielded during the dog days of summer, making participation even less likely.

Content.The material on the site reads like an administrative law textbook, starting with the site title itself: “Consultation on the Obligation to Serve.” Non-expert visitors can’t be expected to make the connection between better broadband (or basic telephony) and the regulatory principle of universal service obligations. And this is the Web. Visitors are in a hurry. You don’t want to give them extra reasons for not sticking around.

Policy assumptions.The CRTC is sticking by the current government’s time-honored message: the only real issue in broadband is Canada has too much geography. While we worry about connecting rural Canadians, broadband is available "virtually everywhere in urban centres.” Speed and pricing are conspicuous by their absence, as is any recognition that “availability” includes awareness and affordability. I have an available Cartier boutique right in my building; that doesn’t mean I buy my watches there. 

Social networking sites (SNS). In a gesture to Web 2.0, the site has a YouTube video imbed and links to several sites like Twitter. As I’ll show from the Web server analytics, this aspect of the site is a poorly planned shot in the dark that yields nothing of substance - little traffic, no SNS dialog, no viral activity.

Questionable questions

Let’s begin with the “ask.” For an agency that thrives on posing sharp questions in PNs and at hearings, the list of five questions leaves much to be desired:

1. What services should be included as part of your basic telephone services today?

2. In the context of this objective, what role, if any, should the CRTC play in ensuring that all Canadians have access to broadband Internet service at comparable rates?

3. Do you think that cellphone service can be a substitute for traditional home phone landline service? Explain why or why not.

4. Do you think that wireless services (e.g. Wi-Fi, 3G networks or satellite) can be substitutes for landline services to connect to the Internet? Explain why or why not.

5. For what activities do you use or expect to use your Internet service?

These questions are a strange mix of either pointless or way too difficult. I put 1, 3 and 5 in the former category, and 2 and 4 in the latter. Whereas #3 could be answered by any 20-something, a cogent answer to #4 presupposes a knowledge of next-gen networks, last-mile business issues, and line-of-site and other barriers to wireless deployment. Let’s say 500 respondents were to argue in their comments why they think cellphones cannot be a substitute for traditional telephony. Would that change anything?

I have more serious issues with the two broadband questions (2 and 5), starting with the sleight-of-hand in #2 about broadband rates:

2. In the context of this objective, what role, if any, should the CRTC play in ensuring all Canadians have access to broadband Internet service at comparable rates? (emphasis mine)

This question assumes rates should not necessarily be affordable but only "comparable." This is a significant policy choice on the Commission’s part. Recent data from the OECD, Harvard’s Berkman Center, Waverman’s Connectivity Report and others all confirm that Canadian broadband prices are way too high compared to other developed countries. This is simply not a debate any more. Why does the consultation site pretend this problem doesn’t exist?

And the ultimate irony. This consultation has excluded the very people the CRTC says most need help: those in rural Canada without broadband. They belong to two sub-groups: those who are still on dialup and those who aren't online at all. Even if you are online on rural dialup, you're very unlikely to participate in this forum, because you haven't got the juice. In other words, Canadians who are affected by high prices and slow speeds in urban areas are not being asked to comment on issues that actually affect them, but on an issue that doesn't affect them. 

Broadband that’s too basic

While not addressed explicitly in the questions, there’s another questionable assumption lurking in the Glossary: the CRTC’s putative definition of “broadband”...

Low-speed Internet service includes speeds that are below 128 kilobits per second (128 Kbps). High-speed Internet service includes speeds at or above 128 kilobits per second (128 Kbps). Broadband Internet access service includes speeds that are above 1.5 megabits per second (Mbps) [see consultation Glossary, Broadband Internet].

The problem here is there is no accepted definition of broadband (see e.g. the Wikipedia entry). These happen to be the CRTC's benchmark speeds (used in some form in the Commission's Communications Monitoring Report: see e.g. the 2009 version, footnotes 229 and 230, p.213). We also know that when definitions are proposed, they reflect self-interest: public interest advocates want thresholds high, ISPs want them low.

The same week the consultation was being launched, the FCC's Democratic majority signed off on the SIXTH BROADBAND DEPLOYMENT REPORT (GN Docket No. 09-137; download file FCC-10-129A1.pdf), which made a widely-discussed change to the rules (pp.7-8): a redefinition of broadband as 4 Mbps in the downlink and 1 Mbps in the uplink. The document is worth quoting because of its consumer-oriented reasoning:

“Today, Americans increasingly are using their broadband connections to access high-quality video, and we anticipate that this demand will only continue to grow in the future. For example, many Americans now communicate with their families and friends through desktop videoconference calls. Many users also now post their own videos and view others’ on such sites as YouTube and Hulu. Instead of reading articles online, Americans often watch videos of today’s top stories [...].

“Thus, for purposes of this report, we update the Commission’s broadband speed threshold. Specifically, we benchmark broadband as a transmission service that actually enables an end user to download content from the Internet at 4 Mbps and to upload such content at 1 Mbps over the broadband provider’s network.”

The FCC has availed itself of data that indicate last-mile bandwidth needs are climbing fast. One such source is Cisco's Visual Networking Index: Forecast and Methodology, 2009–2014 (various links here). Cisco's estimates for North America indicate we're undergoing a 30% annual increase (CAGR) over this period in combined IP traffic - from about 5 exabytes in 2009 to a projected 19 exabytes in 2014 (p.7). Five exabytes is the equivalent of a transcript of all the words ever spoken by humans. And that was last year's traffic in North America alone!

How do the Commission and Industry Canada propose to manage these increases in usage levels while suggesting 80% of Canadians can manage fine with the broadband they have? 

Leaping onto the social network bandwagon

The consultation site came equipped with some basic Web 2.0 tools, i.e. a video with a YouTube imbed and links to several social networking sites (SNS). With over 500 million people now on Facebook, and the use of other social networking sites (SNS) growing, it would seem there’s a compelling case for having SNS links on any big site. There is. But SNS badges aren’t medicine for a site that’s not reaching its audience.

That’s where the CRTC has run into problems. It has done nothing in creating material for the consultation to expand its audience outside those with a professional stake in regulatory affairs.  Nor has it understood the need to overcome the inertia of demographics.

Even when the site is trying to help, as in the FAQ, the non-expert reader is unlikely to take comfort. Reading suggestions include two CRTC telecomm notices, one of which (CRTC 2010-43) concerns the larger consultation on the obligation to serve. It begins thusly:

In this notice, the Commission initiates a proceeding to review issues associated with access to basic telecommunications services, including the obligation to serve, the basic service objective, and local service subsidy. This proceeding will also re-examine the local competition and wireless number portability frameworks in the territories of the small incumbent local exchange carriers. In addition, the Commission will re-examine the appropriateness of the existing forbearance framework for mobile wireless data services.

Why would anyone put ILECs and the existing forbearance framework for mobile wireless data services anywhere near a site devoted to a vox pop consultation? If the Commission is not making these arcane but important issues accessible to the public, it’s not only doing a disservice to taxpayers; it’s also failing to look after its own relevance as a Canadian institution.

Despite the increased use of social media by older adults, the Pew Internet’s research demonstrates that one trend remains unchanged. In 2009 (see chart, above), the most over-represented age group on social nets was 18-24 (when share of SNS use is compared to share of the general Internet population); the most under-represented groups were 45-54 and 55-64. The site authors would have done well to realize that young adults will give you lots of buzz online - but not when this involves reading long documents that look suspiciously like homework. I know; my students told me so.

What the server stats say

If you go to the consultation site homepage, you discover it’s hosted by a firm called Publivate.  A look at the Alexa analytics for the consultation site confirms they’re actually for Publivate (I’m using the Alexa toolbar for Firefox, which isn’t 100% accurate, but better than nothing).

I’ve noticed an interesting change from the time I first landed on the consultation page and clicked through to the Alexa stats (week of August 9). Alexa’s world ranking for publivate.ca was 3,155,803. At the time of writing it stood at 2,338,761, an increase of about 35% in less than two weeks. That increase, which I assume came mostly from consultation visitors, has to be judged in the context of other Web stats.

SNS links need to incent visitors to do more than post a pointer on Facebook or make it tweetable. They have to get people talking, expressing likes and dislikes. The links also have to ride the network effect - what we call going viral, which denotes an exponential rather than linear rate of growth in popularity or awareness (put another way, each transmitter must “infect” more than one person at a time).

So what does the evidence tell us about the CRTC’s experiment with SNS?

I clicked through the five SNS links at several points. I didn’t go hunting for comments, but, as I’ll show in a moment, I did find some revealing details about referrals and traffic in the server stats generated by the site’s YouTube video page. 


I used my own accounts on Facebook and Twitter to get to the inside landing pages. Of course nothing is provided on these pages but a pointer and a few lines of text. The first few times I went to Facebook via the site homepage (August 12 and 13), I found an item from Publivate about the BP oil spill, dated May 31, 2010. Three weeks in, no one had made the Facebook page point to the consultation, a pretty serious oversight.

There wasn’t much else to see via the SNS, except at StumbleUpon. Despite its size (nearly 10.8 million stumblers), StumbleUpon had attracted only one view and not a single like, three weeks after going live. At this writing, it’s scored two views.

While this information doesn’t add up to much, the YouTube video analytics put a nail in the site’s coffin. This is a rather unusual page of information (found below the video player), which I’ll divide into several categories:

- Number of views: at the official end of the consultation, about 1,460.

- Buzz: Ratings - 0, Likes - 0, Dislikes - 0, Honors - 0, Favorites - 1, Comments - disabled.

- Audience: Most popular with males, 35-44, 45-54, 55-64 (no source or methodology is given).

- Links: Over three-quarters of the views came directly from the consultation site. Other referrals were small in number and scope. For example, a Twitter referral on July 23 generated a total of only 11 views, a poor result considering the size of its base (190 million monthly visitors) and the ease with which pointers can be retweeted.

- Growth of views: A time-series graph on this same page shows the views went up to about 1,000 in three or four days, then hit a plateau and struggled for more than three weeks to gain a mere 460 hits. This is the big disappointment. These results show that, far from going viral, the YouTube video experienced a continuing drop in the daily rate at which it added views for most of the consultation period.

What’s next?

I’ve tried to describe the numerous shortcomings in the consultation methodology, site materials and the use of social software. At the same time, I think there’s more to be learned here than how to make better use of Web video or Twitter.

First, the odd slant of the questions and poor execution shouldn’t be allowed to conceal the very weak policy infrastructure that underpins this dialog with the public. The framework for broadband in this country has gone from weak to unacceptable, as we continue to slip in the international rankings. The consultation makes assumptions about speeds, quality, competition and availability that simply don’t stand up to the evidence. No amount of consulting with the public will fix that.

Second, if we go back to the original consultation goal, we might find a new and better direction for policy reform. The idea that the CRTC should collect the “thoughts” of its citizens might have some rhetorical charm. But it’s no way to regulate. So let’s pass on the attitudinal stuff and get back to basics.

In a recent technical paper, the FCC has confirmed what Ofcom established in its own study last year: namely, that US broadband subscribers are getting no more than 50% of the bandwidth promised by their ISPs. The advertised “up to” speeds averaged about 6.7 Mbps in 2009, but both the median and mean actual figures were way below that level, at about 3 and 4 Mbps respectively. While this round was based on data from Akamai and comScore, the FCC is now going out to the field to install meters that will measure actual line speeds in thousands of homes. It’s expected this research will lead to a set of truth-in-labelling regulations.

Right now the CRTC’s idea of consultation has fallen between two stools: it isn’t creating any rapport or dialog with its constituents, nor is it generating information that could conceivably be used to advance any real policy agenda. The Commission needs to make a much greater effort to separate outreach and marketing communications from actual research, of the kind the FCC is committed to.

One good thing might have come from this failed exercise in consultation. Now that it’s over, the Commission and others - both friend and foe - need to ask what we got for our money. As we do, my hope is this online fad will be pushed aside in favor of smarter and more effective ways of talking to Canadians about their communication needs.