Wireless haters take note: our home broadband still sucks too

6755-portobello-1bLondon, Portobello market, August 2013


Fresh data show Canada is still a mediocre performer among the dozens of nations measured continuously by Ookla on 5 broadband performance variables. Will the CRTC’s 2013 Communications Monitoring Report, due out this week, keep up the old tradition of pretending Canadian broadband is just fine?


[Sept 24: added pointers to Ookla]

For weeks now, we’ve been pummeled by tales from the wireless wars. As recently as last Thursday, Michael Geist was reminding us how two-faced and hypocritical the incumbents can be, as if that was a surprise. Nothing gets the incumbents foaming at the mouth – behaving like “raving lunatics,” as Tony Lacavera put it – like the prospect of being disciplined by real competition.

I say it’s time to think again about the equally dismal and depressing state of wireline broadband in Canada. Wireline isn’t nearly as sexy as it used to be – not as fodder for controversy I mean. A couple of years ago, we started hearing forecasts from the likes of Cisco pointing to the triumphant rise of mobile everywhere. The mobile forecasts are holding (see Cisco’s mobile forecast for 2012-2017 here). But even 4G LTE isn’t going to make a lot of subs give up their residential DSL or DOCSIS any time soon. So let’s get back to making invidious international broadband comparisons, this time courtesy of Ookla and its ongoing Net Index broadband usage project.


We’re going to look at 5 different metrics in all: download speed, upload speed, quality, value for money, and promised vs actual performance. But let’s start with the usual starting point, and a skill-testing question. What do Andorra, Taiwan, Macau, Lithuania, Latvia, the Republic of Moldova, the Åland Islands, Estonia, Portugal, the United States and Malta all have in common?…

They and 24 other countries have all shown in the 30-day period ending yesterday that they have faster average download speeds than Canada. They’re not alone. Canada on this dimension stands at 36th place in the world broadband ratings, comprising 184 countries.


The ranking for Canada is based on a download mean of 18.57 Mbps. That mean is not only worse than the US, at 19.55 Mbps. It also fails to live up to 3 collective averages: the EU, the G8 and the OECD. There was a time a few years ago when our government and the incumbents loved to brag about Canada topping the G8; now we don’t even hit the average.

~ Ookla’s household download index (means in Mbps) ~

  • #1 – Hong Kong – 61.85 Mbps
  • EU – 20.66 Mbps
  • G8 – 19.14 Mbps
  • OECD – 18.99 Mbps
  • #36 – Canada – 18.57 Mbps
  • Mean – 14.85 Mbps
  • #184 – Burkina Faso – 0.20 Mbps
  • Canada in top 10 nations for speed – NO
  • Canada in top 30 int’l cities for speed – NO

Check out some of the attributes of Ookla’s research:

  • Its speed tests reflect actual not reported measures of throughput.
  • Rankings are based on rolling 30-day tests not occasional snapshots.
  • Ookla has millions of meansurements to draw from.
  • Ookla pools data from up to 184 countries.
  • Enough data are collected in some countries to measure down to the city level.

Upload speed (Mbps)

On the uplink side, Canada does much worse, ranking #59 with a score of 4.18 Mbps out of the same 184 countries. Even the US, which has major broadband problems of its own, ranks #41 on mean uploads (5.92 Mbps). And Canada doesn’t come near the collective upload averages for the G8, EU and OECD.

Broadband quality (R-factor – 0 to 100)

Broadband quality is a more complex measure that accounts for variables like packet loss and latency. Canada ranks 28th here, though that’s in a much smaller group of countries, totalling only 48. With a quality score of 83.86, we come in a little below the mean of 84.72. That rank again brings us in below the G8, EU and OECD. Then there’s a twist. For territories where Ookla has lot of data (in this case at least 10,000 tests for packet loss), the “top 10” has Canada suddenly bouncing into rank #6, behind the UK, Italy, the US, Poland and France. (Note to CRTC: get out there and start testing.)

The value index (see: money)

certOokla’s value index is designed to provide a cost-benefit ratio, i.e. how much speed you get for how much money. As any wonk will tell you, cross-national comparisons are tricky when money is involved – in this case what subs pay their ISPs for access. Ookla normalizes the many different currencies by translating them all into USD. On that basis, Canada ranks #30 out of the 60 countries in this sub-index, which is the median (not mean) cost per megabit per second – in our case US$3.85/Mbps. This is one of the 3 value indices provided here, and it’s spread across a wide range of values: from Bulgaria at the top (#1 at $0.49 USD) to Venezuela at the bottom (#60 at $34.28 USD). The main value index page is here.

At the OECD, the broadband team tries to eliminate discrepancies by applying the formula for purchasing power parity. In Ookla’s case, 2 similar smoothing formulas are applied: RELATIVE COST PER MBPS – The median cost per megabit per second (download) divided by the Gross Domestic Product per Capita. And the RELATIVE COST OF BROADBAND – The mean broadband subscription cost divided by the Gross Domestic Product per Capita.

  • On the first of these – median cost per Mbps divided by GDP – Canada stands at #20 out of 64 countries.
  • On the mean cost of a subscription divided by GDP, Canada stands at #24 out of 64.

Promise of performance

The 5th and final index, another median measure, is computed by finding the ratio of actual (measured) download speeds to promised (i.e. advertised) speeds. This issue often goes by the “up to” designation: the download speed ISPs suggest you may get, but will never really promise, given all the qualifiers in the fine print (depends on your browsing habits, congestion, server latency, bla bla).

The numbers show that here too Canada is no better than a middling candidate: we rank #25 out of 64 countries. At 93.55%, we at least have the honor of beating out the 4 collectives: APEC, EU, OECD and the G8. On the other hand, we don’t make it into the top 10, nor does any Canadian city appear on the top-30 list of cities. The main promise index is here.

The Gospel according to the CRTC

In the next couple of days, the CRTC will be releasing the 2013 edition of its Communications Monitoring Report. As I’ve mentioned frequently on this page, the CRTC has done a great disservice to the Canadian public through the picture it paints in the CMR of our national media and telecoms (see e.g. The CMR: CRTC’s annual exercise in pseudo-science, posted in August 2011). On reading the CMR, you could be forgiven for thinking that in Canada, we have no consumers, only service providers. And that we have no problems to solve either. In Gatineau, life is always a bowl of cherries.

This week, the CMR is coming to us courtesy of the CRTC’s Consumer Affairs and Strategic Planning staff. It won’t contain any magic formulas for fixing what ails our broadband industry. But it would be nice if the Commission finally has the guts to acknowledge that all is not well in broadband land. And that their wishful figures about the broadband access purportedly available to 99% of our population don’t begin to tell the real story. It’s time for the Consumer Affairs branch to start doing some hard thinking about their forgotten constituency.