Short warning: I’ll give you a “too long, didn’t read” summary as this is a long and mostly boring blog post. Lots of quotes and then some comments from me. It may interest some readers but most likely not all.


Tomi Ahonen – an author and a consultant, according to himself (although not by others) most accurate forecaster in mobile – made his forecast on Apple smartphone unit sales and market share for calendar year 2015. [1] Those were:

Unit sales: 215 million
Market share: 14% (revised from original forecast of 12% [3])

In reality [2] the Apple smartphone shipments for year 2015 were

Unit sales: 231.5 million (Tomi is off by 7%)
Market share: 16.2% (Tomi is off by 16%)
Without the revised forecast Tomi would’ve been off by 26%.

The main point of the long text below is that Mr. Ahonen forecasted Apple to lose market share (from 14.8% to 12% 14%) but instead Apple gained market share (from 14.8% to 16.2%). This rise was visible throughout the whole year yet he was unable to accept it as it would have been against his “most accurate forecast”.

The math above (2.2 points of percentage presented as 16%) was questioned in Twitter. I do admit that if we have e.g. a forecast where market share is expected to be 30% and reality turns out to be 10%, we could say that the forecast is off by 20% (the difference between the two as in points of percentage) or by 67% (reality was 33% of the forecast).

In this post I’m merely following the method introduced by Mr. Ahonen himself. It abandons both manners above and instead that forecast is off by 200% (as the forecast was 3 times as big as the reality turned out to be). We may not like it but a man needs to be evaluated by his own metrics, eh?


We should start from Tomi’s review of iPhone 6. It’s crucial if you want to understand the rest of his actions. We have all past iPhone forecasts from Tomi and this one might be the worst:

this means three things. It means the news customer acqusition to iPhone will be weakest its ever been. The growth in absolute unit sales will be worst its been, the iPhone market share for Christmas 2014 and Spring 2015 will be worst its been in years. (when we combine that with declining iPod sales, flat iPad sales and delayed launch of the iWatch to next year, it means very modest Christmas news for Apple). These two smartphones are priced at the top of the price pyramid where competition is rough and rivals are tough. The consumers of smartphones are not now on their first smartphone, they have grown to appreciate certain things and will seek the model they want. These two iPhone 6 models will not set the world on fire in the way that previous new iPhone model upgrades have. Yes the iPhone will set a new sales record of course for Christmas but its market share will be the worst the iPhone has seen for many years and the total annual market share for 2014 will be somewhere down near 14%. As China transitions away from handset subsidies, these two iPhone 6 models will not be very compelling to the Chinese market where local rivals already have competitive offerings and will upgrade their portfolio with even better (more competitive) models for the Chinese gift-giving season (Chinese New Year in early 2015).

Yes the iPhoen6 models will sell very well to existing loyal iPhone owners. But they are spectacularly weak compared to all rivals out there now, in terms of winning new customers to Apple. Apple’s bleeding market share will continue and I think this year Apple ends with something around 14% market share (full year) and next year powered mostly by this model range, the iPhone market share continues to fall to something near 12%.

–Tomi Ahonen, September 16, 2014 [3]

I have written about this before (visit link) but the fact is that this – eh – forecast is as far off as it can be and in every imaginable way. It’s as if Tomi did not get iPhone 6 at all…
…which undoubtedly is exactly the case. Ever since the original iPhone Tomi has given his reviews and every time iPhone is lacking towards competition and will not be a hit. If you find Tomi saying that the iPhone would set the world on fire, it will be when he talks about the past models…
…but he did NOT rate those very same models as hit products when he originally reviewed them.

A month later Tomi shows us his unbeatable forecasting skills and tells us how much iPhone 6 will sell on fourth quarter of 2014. More precisely, he sticks with his view that Apple must lose market share and then just goes by that assumption:

But for Apple to achieve a stabilization of its market share erosion year-on-year, would require roughly 64 million iPhone sales for the Christmas Quarter now. Even that would only give a market share annually of about 14.5%. And it would require a herculean jump from the 51 million iPhones sold last year in the Christmas quarter. I think more reasonable is something nearer 58 to 59 million.

–Tomi Ahonen, October 29, 2014 [4]

That forecast – naturally – was also way off. If Tomi would bother to listen those who know about things…
benbajarin_twitter [5]
…he would realize he has been wrong all the time. But Tomi has never been much of a listener, really. So we get Christmas results. Apple sells amazing 74.5 million iPhones, just an inch less than Samsung. Tomi however is not going to admit that he would have been wrong. Heavens no!

The big spike sales is OF COURSE driven by the loyalists who have to attend iChurch every year. Next scheduled iChurch gathering is April for the Midnight Mass of the hailing of the Apple Watch.

–Tomi Ahonen, January 28, 2015 [6]

In the same text Tomi explains that you should not compare Apple to the performance of previous quarter, but to the same quarter last year (and he is right in that):

Because of the peculiar launch pattern by Apple that produces that huge sales spike for one quarter going from Q3 to Q4, and then essentially flat sales for the next 9 months of diminishing Quarterly market share, the relevant measure with Apple is always the annual comparison. So compared to Q4 of 2013, the unit sales of iPhones are up 46% (better than the industry overall) and the market share is also up from Q4 of 2013 which was 17.8%.

–Tomi Ahonen, January 28, 2015 [6]

Although Tomi forgot he had made failed forecast about fourth quarter, he admits iPhone gained market share on fourth quarter against his forecast. And if you look back, you’ll notice Tomi originally forecasted 12% full year market share for Apple in 2015 but by the time of his annual forecast (done after Q4 results) he had to tune it up by two points of percentage! That “fine-tune” by 17% just magically happens, no explanation given. Obviously Tomi had to fix it after the monster sales of Q4 2014 but iPhone 6 being a success would be against Tomi’s forecast so he just silently adjusts the number.
(That is okay, I won’t pick on him as he will get more stubborn from now on.)

Now how about the first quarter results? They come out after Tomi’s annual forecast and it turns out that iPhone sales were not driven by “iChurch” as Tomi stated since Apple STILL increased market share in Q1, compared to the same quarter of previous year which Tomi said was the preferred metric.

Apple results out obviously. iPhone sales down 11% from Q4 to 61.1M and prelim market share is down to 17% from 20% last quarter. The large screen bump to the iPhone is now subsiding clearly. And obviously the silly stories of iPhone being bigger than Samsung are now well past us.

–Tomi Ahonen, April 30, 2015 [7]

Now you may notice that Tomi suddenly switches to comparing Apple to previous quarter, which he has already said is the wrong metric. That was pointed to him in the comments of his blog but Tomi silenced those comments quickly:

We cannot measure Apple growth only looking at one quarter, because of the seasonal spike pattern. Apple past 12 months (April 2014 to March 2015) did NOT grow 40%. Apple grew from 159.7 million to 210.1 million iPhones. You do the math, that is not 40% that is 31.5% growth.

–Tomi Ahonen, May 05, 2015 [8]

So as Tomi’s forecast on market share decline did not hold, he changes the metric to moving 12 month average. How convenient. But you may notice the moving average already now shows unit sales of 210 million, just 5 million less than Tomi’s forecast. At least he has noticed it as he does a bold prediction:

If Apple does an iPhone 6C update, I think its still safe to assume Q4 2015 will see actual unit growth vs Q4 of 2014 but if Apple only do premium phone version updates (to iPhone 6 and 6 Plus) we could see first quarter ever that iPhone – Quarter-vs-year-ago will be down. As you know, that is not the way to measure iPhone performance (that is 12 month moving average) but a ‘this quarter one year ago’ metric would be a massive shock to Apple loyalists and might signal a minor panic around January 2016. It is too early to tell, we will know if this is even plausible only after we see the new iPhone lineup in September – BUT – the overall smartphone industry sales will stop growing one day. Before that happens, because they only serve the top price tier, iPhone absolute unit sales growth WILL END. Their market share has already long ago stopped growing so iPhone is growing at less speed than the industry. What comes next, is the ‘one quarter vs same quarter year ago’ measure – which will preceed the actual annual UNIT SALES PEAK which we will see in coming years, might come as soon as 2016 or 2017 but certainly by 2019.

–Tomi Ahonen, May 05, 2015 [8]

Let’s move to second quarter. AGAIN Apple grows market share compared to the same quarter year ago. This means that Tomi no longer cannot use 12 month moving average to show decline in iPhone market share. What does Tomi say?

So Apple came in with 47.5 million units of iPhone sales in Q2 (calendar quarter ie Apr-Jun) which is down 22% from Q1 when it was 61.1 million. This gives Apple a current market share of 13.4% down from 17.9%. As always – ALWAYS – you CANNOT do Apple analysis quarter-on-quarter, you could not do that on the great Christmas quarter and you can’t do it now in the bad summer quarters either. Apple can only be measured for its annual performance and its again – as I’ve been saying – on track to lose (modestly) market share (again) this year.

If you do want to do current analysis of Apple you have to use a statistical tool called 12 month moving average – but then you’d also need to do that for all the rivals and the industry, so yeah, I’m not gonna bother doing that every quarter. That is done automatically once per year when full annual numbers come in , and the truth is, Apple is bleeding market share annually, has been for years, and the rest of the hysteria about iPhone 6 Plus massive jump was just noise. Stay on the real numbers, you won’t be misled by the hype.

–Tomi Ahonen, July 22, 2015 [9]

It should be noted that Tomi now “is not gonna bother doing” moving average. Two of the commenters in his blog did. And it shows Apple has grown market share from 14.6% to 16.5%. Once again a place for Tomi to admit he has been wrong about iPhone 6. And once again he just can’t do that. He merely acknowledges the numbers and reads exactly the opposite conclusion from them:

thank you both for doing the math for our readers.
With that minor correction, thats what the iWorld looks like and the phablet surge is well past us

–Tomi Ahonen, July 22, 2015 [9]

When we get to third quarter results, Tomi has already banned the 12 months moving average from the allowed toolset. Instead he sets new rules for estimating Apple – he will only do it for full calendar year:

Apple launch always creates a big spike, for Christmas. Then the sales gradually decline in the next three quarters. If Apple launched in January every year, it totally wouldn’t matter. But because Apple launches, and its huge spike is for Christmas, then that big jump falls to ‘last year’ ie 2014, and then the next three quarters of declining sales arrive here in the new year ie 2015. Yes, looking back at the same quarter last time, can yield a higher market share, for one quarter or several. But to convert that into a CALENDAR year ie January to December, which is how we do it on this blog, every year. I will not bother to recalibrate and recalculate all data for whatever bizarre fiscal years some companies have, I am doing this for free. I will only do it once, that is by calendar year. So by calendar year, we have to wait until THIS quarter, Q4, the October-December calendar quarter is finished.

–Tomi Ahonen, October 29, 2015 [10]

It seems Tomi put his hopes on iPhone 6S being a dud (which would save him from the disaster-of-a-forecast he had made). Unfortunately he was not that lucky as Apple sales were not completely tied to iPhone 6S. Once again, there are people who could tell this to him:
But Tomi usually sticks to his stand no matter what and that is also the case with his annual market share decline forecast (and the forecast of decrease in unit sales he did in May):

On iPhone – as someone said in the thread, its possible for Apple unit sales in Q4 to be down vs last year (ie quarterly peak iPhone) and yet iPhone to have a slight gain in market share for the calendar year 2015 vs 2014. I still expect that market share to be slightly down and no data so far has suggested to me that it would not be so.

–Tomi Ahonen, November 23, 2015 [12]

I really like how this very same comment was summed up in Twitter: No data except Apple guidance and analyst consensus. 😉
It’s amazing how difficult it is for someone to just admit he’s been wrong – it had been over a year since Tomi’s iPhone 6 review, all the time every news have indicated that Tomi’s forecast was a failure and yet he still insists on it.
Please note this latest step further to denial took place AFTER we had Q3 results available. At that point Tomi’s own estimate was that year 2015 total smartphone sales will be 1,530 million to 1,550 million and it was already then overly optimistic compared to general consensus. E.g. IDC forecast was that market will grow 10.4% YoY which indicated about 1,450 million units – hundred million less than Tomi’s forecast. [13] (And reality turned out to be even less – 1,433 million units (growth of 10.1%). [2])
But even with Tomi’s overly optimistic 1.55B units it was a mathematical fact that Apple needed to ship less than 72.3 million iPhones on Q4 if the market share was to fall from 14.8% that Apple had in 2014. Anything above that 72.3 million marker was going to be a rise to the market share. General analyst consensus in November [14] was that Apple would ship 76 million iPhones on fourth quarter. And in addition to that Apple CEO Tim Cook said himself that Q4 unit sales will be up compared to Q4 of 2014 [15] (and Apple never misses their guidance except up). Yet Tomi says no data suggests the market share could be anything but down. That of course means unit sales in Q4 need to be down (against Apple’s own guidance) so Tomi keeps that stand too:
As usual, Tomi could of course give up on his stubborn view and listen to those who know about things:
(if you are interested on details, please check the comments, picture works as a link) [17]

But let’s stop beating the dead horse. Year 2015 was yet another disaster for Tomi’s iPhone forecasts and I can’t say I would be surprised. But don’t worry, year 2016 has lots of nice stuff coming in: I’m pretty sure that none of you knew that iPhone 7 will have 7 inch display!
…or perhaps after all this it is more certain to assume that it won’t. 😉

(EDIT: I found the actual IDC report, it’s now reference [19].)