So as I released the TomiAhonen Mobile Forecast 2014-2018 last week and celebrated that with new forecasts of upcoming milestones, now is a good time to look back at my past forecasts (and who dares do that in mobile forecasting haha?)
ITS EASY TO MAKE FORECASTS, ITS NOT EASY BEING ACCURATE
So yeah. Just for context some of the wild forecasts promised in 2012 by some of my esteemed peers in the forecasting racket. Did you see any of these return to how accurate their forecast was and why it went wrong? in 2012 iHS iSupply promised us Windows Phone would zoom up to 15% market share (reality? stayed at 3%. 5-fold error). Abi forecasted in 2012 that smartphone unit sales would pass dumbphones in four years in 2016 (reality was the very next year 2013, so error was 4x). In 2012 Ovum told us iPhone would grow global market share and reach 27% by 2017. Well it didn't grow its been falling and is now at 15%. Maybe Ovum turns out still correct but highly unlikely. As it now stands, they are off 80%. Credit Suisse told us in 2012 that global smartphone sales would hit the 1B level in 2015 (reality this year will be 1.5B, so an error of 50%). Forrester told us in 2012 that smartphone apps would be worth 56 Billion dollars this year 2015 (reality will be about half). Research & Markets forecasted in 2012 that the global smartphone market would be worth 218 Billion dollars this year 2015. Reality is nearly twice that at about $400B (so error of say 83%). Gartner in 2012 promised that Windows Phone would pass iOS to become second biggest smartphone OS in 2015. Apple has 15%, Microsoft has 3%, thats a 5-fold error. In 2012 Morgan Stanley forecasted for us that Nokia would sell 80 million smartphones in 2013. They sold 30.5 million (error 162%). IDC in 2012 promised us that year 2016 would see Android OS pass Windows on all new digital devices sold. It actually happened the next year 2013 (so it was an error by a factor of 4).
I could go on, but you get my point. Its not easy being in the forecasting business, we all get it wrong from time to time, and sometimes we make big booboos. But most of my 'peers' do not bother to come back and remind you about how their past forecasts were, nor to try to examine why that forecast went wrong. Now, I would be remiss, in discussing errors in mobile industry forecasts, if I didn't remind the readers of the worst forecast of the mobile industry in this century so far, that by the utterly incompetent Pyramid and their lunatic 'expert' Stela Bokun the biggest fraudster in the mobile stats and facts industry (whatever happened to Stela Bokun, worst creep of this difficult industry). So yes, that was a forecast made in 2011, but they promised the world that WIndows Phone would grow past Android by 2013 (which tells us a lot about who must have been paying Pyramid for this garbage) and a 'forecast' based on utterly imaginary starting points where their base year numbers were off at the extreme by 120% (yes, that would be 120% error somehow magically in favor of Microsoft. Yes its the Microsoft Bullshit Brigade again at work with their propaganda here). Pyramid was widely ridiculed for that worst forecast of the industry and for the record in 2013 Android sold 767 million units of smartphones and Windows Phone managed 33 million. That error is as far as I know the biggest error in any forecast in mobile ever made, at 2,200 percent !!! in a very short term forecast period covering only 2 years !!!!!!!!! Yeah Pyramid, you should be ashamed of yourself and nobody reading my blog should ever ever buy another Pyramid forecast or report for anything in or near mobile or tech .Shame! (But I told you so at the time that it was an epic ridiculous disaster of a forecast. As I wrote in 2011 about Pyramid's forecast: "That is patently false. It is a lie. Pyramid Research are lying to its customers. They are not just incompetent, they are liers!" If you want a good laugh, go read the original blog). But yeah, we know Microsoft try to buy good news, probably some money exchanged hands on this outrageous, ridiculous, fantastical and utterly unbelievable 'forecast' that was instantly lambasted by the industry. Two THOUSAND and 200 percent error in a two-year forecast.
SO HOW IS THE T-DAWG?
Ok, after that little bit of context to remind readers nobody is ever perfect in forecasting, lets go to my 2012 blog forecasting big milestones to come in mobile. In that blog I made 10 milestone forecasts (what moment is passed in that given year) through years 2013 to 2015. Obviously the 2015 forecasts cannot be evaluated yet but we have 8 items we can examine. Here we do arrive at a problem, however. Many of those forecasted data points are a bit difficult to verify from 'independent' sources as some of those are statistics that only TomiAhonen Consulting even reports into the public domain. So its not exactly fair for me to claim that my company has verified this number haha.. Lets see which of the 8 forecasts DO have some independent number and what can we tell from that.
First off the easy ones. I said in 2012 that in year 2013 the world will see 1 Billion smartphones sold. And we do have the count, it wasn't quite that. The average of the big analyst houses reporting that statistic gave us the annual count at 990 million. I was wrong. I was off by 1% on this, one of the most seismic statistics for the industry. I'll take that. 1% error is as good as perfect.
Then I forecasted in 2012 that year 2014 the world will see total handset sales hit 2 Billion units. Here I was more off. According to Gartner, the total was only 1.88 Billion so I was off by 6%. Thats ok but not exactly spot-on. I personally happen to think the actual number was more than 1.88 - as Gartner is very conservative in their numbers but I also do agree it wasn't 2B last year, so fair enough. I was off by 6%.
Then I forecasted in 2012 that the total number of SMS active users would exceed human population by 2014. Here I was quite a bit off. SMS active user counts are rarely reported but we do have a global count by Acision from 2013 which was 6 Billion. And now an analysis of the numbers based on Pew global survey of Emerging World mobile users, and adding rich world numbers to that, gives an SMS text messaging active user base at about 5.6 Billion in 2014. My own consultancy also agrees it was under 6 Billion, I counted it at 5.8B. So if we ignore my number and take the average of those two, ie 5.8 Billion, that is still way off from the human population at 7.2 Billion last year. That is an error of 19%. That is quite a lof of error and I do apologize. What have I learned? SMS user growth stalled just about when I posted that forecast, with the rapid adoption of Whatsapp. But because most OTT messaging users also continue to use SMS occasionally, even that doesn't explain the full matter of the stalling of SMS user growth. I think we've just hit the natural limits that come from literacy and the use of the same services across multiple accounts. Many people might have two or three accounts where one is optimized for SMS (with say an unlimited SMS bundle) and therefore that user won't bother to use the other account(s) for SMS while using them for say voice or internet browsing or whatnot. Still yes, I was off by 19% on this count. That being said, SMS is still by far the most used data application on the planet, not just on mobile. SMS is also the most used feature on mobile phones far ahead of voice, camera and clock. And SMS dwarfs all other messaging platforms like MMS, email, OTT instant messaging etc. By branded platforms SMS is four and a half times bigger than Facebook by active users and obviously 8 times bigger than Whatsapp (by active users, the traffic generated by that small passionate user base of Whatsapp is far more than SMS users naturally).
I said in 2012 that half of all new phones sold in 2014 will have touch screens. That was a rather easy one as global smartphone sales went from 990 million to 1.3 Billion into 2014 and nearly all smartphones have touch screens. So this was a 100% accurate forecast (while not necessarily very 'bold' one haha).
So that was the four easy numbers to verify. I also made four other forecasts that we can't easily prove to be absolutely true but have at least some signs of were a couple in the ballpark or not.
I forecasted in 2012 that by end of year 2013, one third of all phones in use worldwide would have WiFi ability. I have tried to find an independent count of that number to no avail. Lots of stats about WiFi radio shipments or WiFi traffic but no count of WiFi mobile phones either installed base or new sales. BUT, we do have a close proxy. Back in 2013 most recently-sold smartphones (but not quite all) had WiFi and also some high-end 'featurephones' like Nokia Asha series had WiFi. We can thus take as a close estimate the total count of WiFi as the proxy for WiFi enabled devices (today that is ever less accurately so, as ever cheaper basic smartphones are sold that don't include WiFi). So in 2013, what was the final count of smartphones in use? 1.6 Billion according to this blog, the only site that reports installed base of smartphones globally. And of all phones in use in 2013? That was according to the TomiAhonen Almanac 2014, 3.9 Billion phones. 41% is roughly the number. That was under one third in 2012 so yeah, this is pretty solidly 'likely true' number and date.
But then the internet user forecast. I said in 2012 that in year 2014 half of internet users never use a PC and only use a moblle. For this again, there are not easy numbers. The ITU which is often quoted, only gives 'broadband' mobile internet count on 3G and faster, so they ignore much of say African and Indian mobile internet use on slower networks running 2.5G like GPRS and EDGE. Another popular way to count mobile internet users is smartphone users, but that ignores dumbphones and many of the mid-price dumbphones have full HTML browsers such as the Asha line of Nokia featurephones. And beyond that, yes there is still WAP, the simple mobile-optimized subset of HTML that was the early mobile internet and is still a significant part of internet use in Africa and poor parts of the world. But what we do know that in 2014 the total internet user count passed 3 Billion. If I was to be right, 1.5 Billion should never use a PC. Meanwhile the PC population passed 1.5 Billion and most definitely many who own a PC and a smartphone use both to access the internet, so I was definitely wrong. But this is a difficult number to pin down. There is a reported number by IAB that says 1 Billion internet users will be mobile only, and that is a 2015 number for 2015. Sounds pretty solid except it is only for 'mobile broadband' ie 3G and faster use. Definitely mobile internet user base is signficantly larger when 2.5G and WAP are added but how big? I don't have a number for you out of any published source other than my own. But I will grant you, in 2014, we did not reach the point where half of internet users acccess via mobile-only. Maybe that number was about 1.2B out of 3B so I was off by about 25%. But like I said, this is a number very hard to pin down.
Then there were two numbers we really don't have a good verification for. The camera use ahead of voice calls. Its an inevitable trend. The UK has for example already passed that but is the global number there yet? We don't get enough cameraphone user global stats reported to give that count accurately.
Same goes for TV voting stats. It used to be more widely reported but recently we haven't seen those numbers anymore. Likely yes, considering how well premium SMS revenues keep growing and the popularity of various vote-to-TV show formats like American Idol continue but I don't have any recent number to report to verify. So these two last items we have to leave as unverifyable (as of now).
AND THE VERDICT IS
So the final count. Of 10 forecasted mileposts, 8 have passed. For 6 of those we can determine precisely or at least approximately if they were accurate. On three I was solidly on, excellent forecasts (smartphone sales 2013, touch screens 2014 and WiFi mobile use 2013). On the total handset sales 2014 forecast I was off a bit more with error of 6% but still good. And on two forecasts I was off far more, with SMS users 19% and with mobile-only internet users 25%. Four good or excellent calls out of six, two somewhat off calls but still in the right direction. No really booboo calls here, off by 50% or 100% or 400% or gosh, 2,200% haha... (yeah, I still got it..)
Thats what the most accurate forecaster does for you, here on this blog, all the time. Nobody is perfect in forecasting, that is literally impossble. But somebody has gotta be the most accurate in this business, and that is still me. And if you need more numbers, my brand new TomiAhonen Mobile Forecast 2014-2018 is out. Some highlights of the forecast are again free in this blog article.