In the 1999 headmap manifesto, Ben Russell coined the term “location-aware devices”, in a beautiful techno-romantic vision:
location aware, networked, mobile devices make possible invisible notes attached to spaces, places, people and things. […]
places can have histories ‘attached’ to them (i.e. the collection of notes left at a given place sorted according to when they were left).
inanimate objects can become more animate (if you know where a tree is and you know when someone is walking past it you could make it burst into song).
Today, of course, for any gadget-freak this is not a prophetic vision but an everyday reality – incarnated in an iPhone, Android, Nokia phone or pda or digital camera – or even some laptops. But how many are the freaks, and how many are the ones for whom this is still just an expensive toy that they don’t really feel they need? Put differently: What is the potential audience for a locative media application today?
I’ve been trying to do some research on this, but it turns out these figures are not too easy to find. Device makers and OS makers guard their sales numbers, and even the research department of Norwegian operator Telenor could not share their numbers with me.
(NOTE: Those of you more interested in global numbers than Norwegian-specific ones, skip ahead to the last part, updated 3 May 2010.)
According to Statistics Norway, the number of people aged 16-74 who own a 3G mobile device in Norway is 12%. If we consider that location-aware devices are still a fairly modest subset of all 3G devices (think of all the “smartphones”, “music phones” etc that were sold before GPS became common), we may guess that location-aware devices are less than 5%. But it would be nice to do better than guessing.
Looking at international statistics, separate reports from Gartner and Canalysis, indicates that worldwide, ca. 41 out of 309 million phones sold in Q3 2009 were “smartphones” – that is, 13%. Breaking down the numbers further, we see that iPhone represents 7.4m (I’m assuming this figure doesn’t include iPod touch, which uses iPhone OS but doesn’t have GPS nor ubiquitous network connection). Android devices make up ca. 1.5m. As for Blackberries and Nokias they sold 8.5m and 19.1m, respectively, but how many of those had GPS’s? My guess (again, just guessing) would be that Nokia sold at least as many GPS devices as the total of Android devices, while the number of GPS Blackberries might be a little lower. Say around 3-4m for those two together, giving us a ballpark number of 12-3m devices. That means, less than a third of all smartphones, and just 4% of all mobile phones sold in Q3 last year were location-aware devices. And those are the people getting new phones – all those people who are still using their old Nokia from 3 years ago and have no plans to change are not included in this statistic.
Add to that, of course, that if you make an iPhone app it won’t work on Android, Nokia etc and vice versa in all dimensions – you’re looking at a potential audience of something like 1% of the population, or less, for any application that requires a GPS.
(OK, so perhaps in a country like Norway the numbers are a little higher than worldwide, due to our generally high use of mobile phones and broadband internet. Still, with just 12% 3G devices in total the difference can’t be too large.)
This number sounds not just a little depressing (for us who do want to play around with location-aware applications), but also almost incredible. But I can’t really find the error in the math? Anyone who can help me? Anyone got more detailed (and reliable) figures?
Certainly, this would help explain why no-one can make money from making mobile apps in Norway… (Norwegian link)
UPDATE: I just saw the sales numbers for 3G phones last year in Norway (Norwegian link). 63% out of all the 2.4m phones sold – 1.5m devices! Even in the longer perspective, 2007-2009, the sales share is about 52% (on average). How can then the share of people owning one be only 12%? Even if everyone who bought a 3G phone bought two devices each year, and it were the same people every year, the percentage would be over 12%… Who is wrong, the industry or the statistics bureau?
UPDATE 2: These numbers from the European statistics bureau Eurostat lend support to the figures from Statistics Norway: 4-5% on EU average, Sweden topping the list with 14%. But then again, their data sources and methodology is probably identical to that of Statistics Norway. It seems likely to me that the discrepancy can be a combination of two factors: Inflated industry numbers, and probably some under-reporting in the official statistics (if they called my dad and asked him if he has a 3G phone, his answer would probably be “I have no idea”…
UPDATE 3: Just had a chat with senior advisor Kjell Lorentzen at Statistics Norway. That made me much more skeptical to the numbers from both Statistics Norway and Eurostat (same methodology). They have called around asking people: “Does the household use mobile broadband such as UMTS, turbo-3G or ICE?” (In Norwegian: “Bruker husholdningen mobilt bredbånd som UMTS, turbo-3G eller ICE?”) I think I would have answered that one correctly myself, but only after pausing to think – and I am doing a PhD on this stuff. So it seems pretty likely that the Eurostat numbers are too low. So, if the industry numbers are correct, we might assume that something like 30-50 % of Norwegian mobile users have 3G phones/smartphones, and something like a third of those have GPS’s. A rough guess then would be somewhere around 10 %. A slightly safer figure is that a third of the 3G-phones sold in Norway last year were GPS phones (judging by my estimation that roughly a third of smartphones sold worldwide were GPS phones). These are guesses – only way to verify them is probably with numbers from the operators.
Oh, but worth noting though: The Statistics Norway numbers indicate that twice as many had 3G phones in April-May 2009 as in 2008 (from 5 to 12 %). This might be another indicator of error – industry sales numbers increased much more moderately from 2008 to 2009, even from 2007.
UPDATE, 3 May 2010: I finally found some more reliable figures for at least part of this calculation. According to the Tomi Ahonen Almanac 2010, the worldwide installed base of smartphones was 13% for the year 2009 (p. 33). That is, of all the mobile phones in use in the world last year, 13% were smartphones. About 9% of the smartphones were iphones – in other words, ca. 1% of all the phones in the world are iphones (of which a small proportion are the first generation, which do not have GPS). Judging by this, it seems reasonable to guess that the total installed base for GPS phones worldwide is somewhere around 2-4%.
However, these are worldwide numbers. In the leading mobile markets (Japan, Western Europe) the percentage is likely to be higher. Ahonen does give me support in my critique of Eurostat/Statistics Norway: According to the almanac, 56% of Norwegians own a 3G phone! If that is right, even my guess of 30-50% was a little too conservative. The numbers vary greatly for the rest of Scandinavia – from 71% in Sweden and 55% in Finland to 38% in Denmark. Japan is on top with 78%, the US is way down with 31%. Worldwide average: 29%.
Also worth noting from these numbers, is that less than half of all 3G phones count as smartphones (defined as phones on which you can install your own applications). My guess, however, is that the number of 3G “dumbphones” stems primarily from certain markets like Japan, where the handsets tend to be technically advanced but locked down. According to recent news, the iphone took 72% of the smartphone market in Japan last year – corresponding to 4.9% of the total mobile market. That must mean that the total smartphone market is only about 6.8% of the total mobile market – or about half of the world average. So while 78% of all Japanes mobile users have 3G, only 6.8% of them have “smartphones”! One could certainly ask if the current definition of “smartphone” is a good one when it excludes most of the users in the most advanced mobile market of the world. However, the important point to note in this context is that it lends support to the idea that in Europe, most 3G phones are likely to also be smartphones. (Which makes sense when you think about Nokia’s dominance in Europe – most Nokia 3G phones use symbian, and are therefore smartphones.)
If so, for a country like Norway we can once again guess that the installed base of smartphones (per capita) is ca. 50-55%. And if we assume that the proportion of GPS phones to all smartphones is roughly the same in Norway as in the rest of the world, we may deduce that the installed base of GPS phones in Norway is somewhere in the range of 5-20%. If I were to wager a more precise prognosis, I would guess that around 1/10 of the smartphones are iphones (as they are worldwide), in other words 5% of all mobile phones. (And pretty much all of these will have GPS, since the first generation came so late to Norway.) Furthermore I’d guess that the total number of gps phones is somewhere around 2-3 times the number of iphones – in other words, around 10-15% of the population in Norway have GPS cell phones. Judging by the 3G penetration rates, that number is likely to be roughly the same for Finland and Austria, a little higher in Sweden, and a little lower in other European countries (and much lower in the US).