“By gum Watson, high Canadian ARPU really doesn’t reflect usage!”
So many wireless studies, so many different conclusions? It’s no wonder people are confused. The cynic might think they’re all being done on purpose.
In the space of a week, there have been two separate reports suggesting that wireless prices aren’t so bad in Canada. The first came last week in the form of a study from Ottawa-based telecom consultancy Wall Communications, on behalf of Industry Canada and the Canadian Radio-television and Telecommunications Commission, whose headline finding was that cellphone prices had decreased (links to PDF) over the past year to the point where Canada finds itself “middle of the pack” among compared countries. The second was in the form of a Globe and Mail article by two University of Calgary professors, Jeffrey Church and Andrew Wilkins, arguing again that prices in Canada are low and that services are great.
Let’s start with the Wall study. The report grouped wireless subscribers into three categories based on monthly usage – low, medium and heavy users – and found that prices have declined for them by 11%, 13% and 5%, respectively, over the past year. New wireless companies, including Wind, Mobilicity and Public Mobile, offer prices that are lower by 19%, 14% and 39% across those respective baskets. In Wall’s view, medium users tend to compare well against other countries charted in the study, while low and heavy users are still coming in on the expensive side.
The study also looked at mobile data prices specifically, which found that Canadian prices fall on the high side of compared countries, and are highest of all once above five gigabytes a month.
Some of this runs counter to a report back in May from J.D. Power, which found that Canadians were paying about 13 per cent more for wireless service than they did a year previous. I spoke to Gerry Wall, principal of Wall Communications, about the discrepancy and he was at a loss to explain it, other than to say that different methodologies will turn up different results. It’s hard to say what methods J.D. Power used, but Wall’s findings do seem to closer reflect what has actually been happening, at least in big cities. While many may gripe that prices in Canada are still high, they’re at least somewhat better than they used to be.
The problem with the Wall report lies in interpreting it as comprehensive, because it isn’t. When the study says that Canada is “middle of the pack” in terms of pricing, that actually refers to only a very limited group of six compared countries: the United States, the United Kingdom, Japan, Australia and France. Three of those countries – Canada, Japan and the United States – are like the John Olerud, Paul Molitor and Roberto Alomar of wireless, in that they’re first, second and third in total monthly revenue gleaned per subscriber (points to anyone who recognizes the reference).
The U.K., France and Australia are more middle of the pack by that measure, according to the Bank of America Merrill Lynch Global Wireless Matrix, or the wireless industry’s bible. When additional countries are included in comparison – the Wireless Matrix compares 50, including 19 developed nations – Canada’s standing isn’t so middle-of-the-pack. Wall says he wishes he could include more countries in his comparison, but he simply isn’t given the budget to do so.
One of the other apparent issues with the report is which service providers are compared in those six countries. On the wireless side, the new Canadian entrants are included, yet down in the United States, only AT&T, Verizon and Sprint are counted. I asked Wall why this was – why weren’t smaller U.S. carriers such as T-Mobile and MetroPCS not compared? They generally offer better prices than their bigger rivals, so I wondered whether including them might might make the U.S. look a shade better.
He explained that the small Canadian players were included because his clients – the government and CRTC – want to see how they’re doing. But the results are weighted to market share, meaning that the likes of Wind and Mobilicity don’t actually affect the overall price results. “As a matter of aggregating them into the Canadian price index, the market share is so small that they’re really not material,” he said. That doesn’t mean they haven’t had a real effect on pricing – surely their very presence has forced the big incumbents to lower their prices somewhat.
Wall added that the same rationalization was used for U.S. carriers, as well as for comparisons on the wired broadband side. While big international internet providers such as BT, Telstra and Orange were compared, the likes of Tesco (the U.K. supermarket chain that sells dirt-cheap broadband), Free (France) and iiNet (Australia) were not. Most surprisingly, ISPs in Kansas City, Missouri were studied, but not Google Fiber, which is selling ultra-fast access relatively cheaply in that city. “When we do an index of prices, we want to do something that captures the price to the majority of people who live in that area,” Wall says.
(In case anyone was wondering, the study found that Canadian broadband prices compare favourably at lower levels, but are higher at faster speeds than all countries except the United States. Broadband prices have also come down 1%, 6%, 4% and 12% across the four service/speed tiers, although instinctively, I can’t see how that might true. While wireless prices have certainly edged downward, internet access has only gone upward.)
Getting back to the mobile comparisons, I’m not sure Wall’s explanation of the weighted averages makes sense, at least as far as the United States is concerned. While AT&T, Verizon and Sprint certainly do put significant weight on price indices with their respective 100 million customers (about 50 million in Sprint’s case), the smaller players can’t be discounted. T-Mobile has about 35 million subscribers while MetroPCS has about 9 million. Including them would likely move the U.S. needle downward somewhat.
Regardless of that, prices in Canada do indeed seem to have marginally improved over the United States. A basic plan from AT&T with 450 minutes and 300 megabytes of data costs about $60, while a comparable one in Canada from Telus (with much fewer minutes, but unlimited evenings and weekends) runs about $55. Which brings us to the Globe and Mail story.
Church and Wilkins again argue that prices are not necessarily the same thing as average revenue per user (ARPU), which is a measure that Canada leads the world in. They say that high ARPU is the result of high smartphone usage in Canada; with more people voluntarily using more expensive data plans, the ARPU that Canadian carriers are posting is naturally going to be higher.
I debunked that particular notion last month by using voice minutes from the Wireless Matrix and data usage figures from Cisco’s Virtual Networking Index. Church and Wilkins used the same numbers, but slightly differently in the case of data, to arrive at entirely different conclusions.
The Australian results are a good example. I got my numbers directly from Cisco, which stated that the typical Australian mobile data connection used 652 MB per month, which was almost exactly the same as the 656 MB used by Canadians in the same time frame. Australian voice usage was also close to Canada’s, yet ARPU was lower there by a third, indicating that prices are indeed lower in Australia.
Church and Wilkins, however, used Cisco’s online VNI, which lists very different results – sometimes dramatically so. In the case of Australia, the average mobile connection generated only 367 MB of traffic in 2012. Canada is similarly off, with each user posting only 536 MB of total traffic per month, as are many countries.
I asked Cisco for an explanation and it’s actually pretty simple. The numbers they gave me included data usage across all mobile devices, not just smartphones, which are naturally higher. “The average number of devices per user is higher in Australia, so users will spread their mobile usage among multiple devices, bringing down the per device usage average while maintaining a high per user average,” a spokesperson says. Those multiple devices can include smartphones, tablets, mobile sticks for laptops, but they all ultimately add up to the same thing: ARPU for wireless carriers.
Multiple devices complicate comparisons, but not terribly. Here’s a chart I put together back in April that factors the effect of multiple devices into ARPU. For those keeping score at home, I multiplied penetration levels – the percentage of a country’s population that has a mobile device, which in every case but Canada’s is above 100 per cent – by that country’s ARPU. Australia is actually the top dog here, but Canada still comes in at a comfortable second. Note that the United States – with its higher data usage and much higher voice usage – is still decently below Canada:
The conclusion here, as it was in my previous post on this subject, is that high usage does not explain high ARPU. As I said back in June, the Sherlock Holmes-ian answer is that when all other possibilities have been eliminated, the probable remains: Canada simply has high wireless prices.
Beyond that, I don’t actually disagree with the headline on the Globe article, which suggests that high prices are part and parcel of quality networks. The article argues that Canadian carriers have been among world leaders in spending on fourth-generation HSPA+ and LTE networks and thus have the quality to show for it. I can’t see why that’s not true, but I do have a problem with the premise as to why such spending has happened. It hasn’t been because of fierce competition, as the industry likes to profess, but rather because of technological reasons and some bad investment decisions that went with them.
It was only a few short years ago when Bell and Telus were getting pummeled by Rogers, thanks to that company’s chosen technology. Rogers, like most of the carriers in the world, went with GSM network technology while Bell and Telus opted for CDMA instead. Without getting technical, GSM won, and Apple put the exclamation point on the battle in 2007 in the form of the iPhone. Unable to offer the latest and greatest devices, including that quintessential and hotly desired device, Bell and Telus moved quickly to upgrade to the next greatest and latest 4G technology. Rogers followed suit. The same is happening in the United States, with Sprint and Verizon – both former CDMA users – both spending heavily on LTE.
Network investment in both Canada and the United States does not reflect the competitiveness of either market, but rather phone makers’ decisions on technologies. Carriers are simply being pulled along for the ride.
One thing I may indeed have been wrong about in the past is how high prices were mainly the result of the lack of foreign competition in Canada, which wasn’t legally allowed until last year. The poor technological choices made by a number of carriers can’t be discounted as a factor. The industry is now waving the billions they’re having to spend to correct those mistakes in the faces of consumers and government, with prices – be they as they are – the necessary rationalization.