I wrote few days ago about Tomi Ahonen reading Nokia’s market share decline from Kantar statistics and pointed out that even though he claims the opposite, the data just proves Nokia was losing market share prior to February 11th announcement.
But that is not only time Tomi uses Kantar statistics for something that does not make sense. Let’s see another blog post from him. This is more recent and also about Nokia-Microsoft alliance. This time he tries to track Nokia switching one platform (Symbian) to other (Windows Phone):
“My original ‘gut feeling’ for this alliance was that they would achieve something 10% by end of 2012. This was said 4 days after the alliance was announced (and I projected on that day that Nokia’s end-of-2011 market share would fall to 12% – which it did, and no other analyst dared project such a total collapse of Nokia’s share at that time). I said this was my gut feeling and not yet my formal forecast for this partnership after the new Nokia Windows based smartphones would be released.” 
Really, Tomi? Your gut feeling for this alliance? Your gut feeling 4 days after the alliance was announced?
Sorry Tomi, but you lie straight against our faces.
You had that prediction (or at least part of it) done already at end of January, about 2 weeks before Nokia announced their strategy with Windows Phone:
“So the facts. Nokia’s market share is in death-spiral, crashed from 39% to 28% in just six months and warnings from management suggest Q1 will continue the bad news, so it may end somewhere near 24% by end of March and who knows where the bottom is. Motorola was in a similar position in 2006. They had 21% market share. The crash-dive started, Motorola went from very profitable to very unprofitable, and lost customers everywhere, and the blood-letting ended in 2010 when they managed to stop the decline – and found themselves with 2% of the market.” 
And what do we know? Nokia’s Market share indeed came to be 24% at end of March!
So your “gut feeling” that “no other analyst dared to project” was indeed correct – two weeks before the announcement. And now you blame Elop and Microsoft alliance for that? (Not to mention the blatant lies involved.) That’s… I’m out of words.
But let’s get onwards:
“My critical assumption for my forecast for Nokia and Microsoft for this year 2012, when I made it before anyone had seen Lumia smartphones last summer, was that of course Nokia will be able to convert, 1-to-1 the existing Symbian users to the newer ‘highly desirable’ (haha) Windows Phone OS.” 
Ummm… Yes, Tomi. Your assumption kind of was that. This is what you said (and trust me I’ll get back to this):
“And to keep the math ‘easy’ lets say the Nokia Symbian ramp-down will exactly mirror that increase, so for every million lost in Symbian sales, Microsoft Windows Phone exactly matches that decline, picking up every lost Symbian sales.” 
“We can also give a rough model of a linear transition (only as a guide) so if Symbian starts with 10 million to 1 million at Q4 of 2012, and reverses that by Q3 of 2012, we’d get this pattern:
Q4 2011 – 11 million – 1 million WP7 + 10 million Symbian
Q1 2012 – 11 million – 4 million WP7 + 7 million Symbian
Q2 2012 – 11 million – 7 million WP7 + 4 million Symbian
Q3 2012 – 11 million – 10 million WP7 + 1 million Symbian” 
So let’s see you comparing that to what happened. Tomi, stage is yours.
“Now lets take Kantar Worldpanel data. This happens to be quite opportune timing. The June 2011 data is after the Elop Effect, so much of the shock crash of Nokia Symbian has already been factored in, and it is before any Lumia launched (which would be November) but just about the time when the first Lumia leaks started – by Elop himself. So the buzz about Lumia is just about starting in June. Lets see what story Kantar tells us of this attempt.” 
Wait – what?!? June 2011 is 5 months before Lumia sales start and at that point Nokia had not announced a single WP device. That’s “opportune timing“? For what opportunity, pray tell?
Nokia was selling only “osborned” Symbian devices at the time, how do you expect them to stop decreasing sales without the Windows Phone? Shouldn’t we pick results that are ending as close to Lumia sales start as we can?
I mean – come on, Tomi! Your original text clearly talked about Q4 2011 as a starting point! How on earth do you plan to continue after start like this?
“MARKET SHARES FOR NOKIA-MICROSOFT PARTNERSHIP (in ‘West’ Advanced market countries)
OS . . . . . . . . . . . . . . . June 2011 . . . . June 2012
Symbian . . . . . . . . . . . 13.5% . . . . . . . 3.2%
Windows Mobile . . . . . . 1.0% . . . . . . . 0.6%
Windows Phone . . . . . . 1.6% . . . . . . . 2.4%
Total partnerhship . . . . 16.1% . . . . . . . 6.3%
Total Windows . . . . . . . 2.6% . . . . . . . 3.0%
Market Share calculation by TomiAhonen Consulting July 12, 2012, from Kantar data
This data may be freely shared” 
Market shares? You’re using Market shares? Tomi, you said yourself that you expect unit sales to remain same. Smartphone market grew 72% during 2010 and you know it. It grew 80% during 2011 and you know that too! Now that you choose to compare market share from 2011 to 2012, it is not even near the same thing!
If market grows 80% during that year and Nokia sells same amount of units, Nokia’s market share would fall by 44%!!! You can’t compare market shares as if it was unit sales!(Unless you happen to be analyst that does not know even that much about statistics.)
And seriously, Tomi – in addition to the foul choice of time and weird switch to use Market shares, you choose most challenging markets of Nokia (namingly U.S., UK and Germany)?
Could we just… Take Nokia shipped amounts and compare those?
You know, if they were keeping same unit count, it means Nokia managed to keep existing users?
Should we start from Q4 (partial Lumia sales) and see if Nokia managed to keep that level during next quarters?
It is what you originally proposed, you know?
“So yes, Nokia lost 4 customers for every 1 it managed to force to take the Lumia.” 
By what math, Tomi? Let’s have look at the Nokia shipped units and try to stick to your original idea. Nokia shipped smartphone units were:
Q4 2011: 19.6M (<1M Lumia)
Q1 2012: 11.9M (>2M Lumia)
Q2 2012: 10.2M (4M Lumia)
So Nokia actually lost 1 customer for every one that chose Lumia. But mostly between Q4 and Q1. Less than 10% between Q1 and Q2.
Now Q4 is christmas sales quarter, which is always high sales. I’d say we cannot do QoQ compare with any meaningful data from that. And a single quarterly change (Q1 to Q2) is as good as nothing.
Lumia phones were sold in very limited quantities during Q4 2011. I’d say the first time where Symbian users could convert to Windows Phone users is Q1 2012. So if we want to stick to Tomi’s original analogy, we would compare Q1 2012 unit sales of Nokia to Q1 2013 unit sales of Nokia.
So Tomi, what comes to this post of yours: You made prediction about global unit sales for a year-long period that starts from Q4 2011. Now you try to prove things went wrong by using data that is not global, not unit sales and not starting from Q4 2011.
Sorry, Tomi, but this post had too high bulls#it-to-truth ratio.
Let’s see yearly comparison after we have Q1 2013 results, shall we?