When I originally promised to visit Qt/MeeGo/Meltemi/N9/N950 topic, I planned a short glance. Instead I ended up doing a full review of Nokia strategy. So this is a long post then, hopefully rewarding one.
Let’s very first point it out that I actually am very much on same track with Tomi Ahonen what comes to strategy choice of Nokia – Qt powered Symbian/MeeGo would have been the way to go. Difference between me and Mr. Ahonen is that I do not spread rumours and false claims because it did not become reality. I try to stick with facts and I will tell when I go to speculations – do not refer to those as “trusted source”.
That said, let’s get to topic.
NOKIA STRATEGY AS IT WAS
I’ll start by talking a bit about Nokia Strategy in 2010. It was in Brief about Symbian, MeeGo and Qt. First we’ll have Qt told by Tomi Ahonen (don’t worry, this is factual data for a change):
“Qt is effectively an ecosystem. Developers can make apps via Qt that work on Nokia’s Symbian, Maemo and MeeGo smartphones, today. These have an combined installed base of 300 million smartphones currently in use worldwide. That is the biggest installed base of any smartphone platform.” 
So Qt is a C++ framework that makes it possible to develop application once and deploy it to several platforms (Windows, Mac, Linux, Symbian, MeeGo, …). However, it is essential to understand that Qt does not make Symbian applications run directly in MeeGo devices – it allows same source code to be used when building application to both. Therefore having successful developer ecosystem and application offering for MeeGo needed motivated developers willing to build application on multiple platforms. It is not a big deal when application is done, but may need some additional work when you end up to see different display sizes etc. This is essential and we will return to it during this post.
So the Nokia strategy at 2010 was rather straightforward: Keep making the Symbian phones that outsell any other platform. Get developers to do Qt applications to Symbian, lure them to also build them for MeeGo, make products of both type, profit.
But that did not last. So what happened?
WHEN STRATEGY STARTS TO COLLAPSE
Now remember, what I’m addressing next is still in 2010. No Microsoft, no Windows Phone – heck, no Stephen Elop for the starters! And I’ll let Tomi Ahonen do the introduction again:
“In June of 2010, after Q2, 2010, Nokia’s smartphone market share stood at 39%. In Q3 it fell dramatically to 33%. Now in Q4 it fell further to 28%. Yes. In only six months, Nokia’s market share in smartphones fell from 39% to 28% – Nokia lost more than one quarter of its total market, in just six months!“ 
“I have never witnessed such a wholesale destruction of any company’s market share in a similar period of time. The nearest I can find, is Motorola which lost a quarter of its market in all handsets from 2006 to 2007 in a period of one year. Nokia smartphones today is twice as bad, as Motorola Razr led mobile phones was when facing the iPhone in 2007.” 
“Let me show with a familiar example. Lets go to my favorite analogy to phones: cars. The world’s biggest car-maker in 2010 is (by a slim margin still) Toyota ahead of GM. What would happen if Toyota lost one quarter of its market – I am not destroying 11 market share points because Toyota’s total market share is only 12%, so we can’t drop them to 1% haha, that is destroying nine out of ten customers. No, the same ‘ratio’ of loss. If Toyota lost one quarter of its market, and instead of selling 8.4 million cars, they would only sell 6.3 million cars. That is the same proportion of loss as Nokia going from 39 million to 28 million sales. If Toyota fell from 8.4 million to 6.3 million, Toyota would be in the news headlines all over as the ultimate catastrophic collapse of a global market juggernaut – Toyota would fall behind General Motors, AND Renault-Nissan AND even behind Volkswagen Group, landing in 4th place, just ahead of Ford Motor Company. Do you see what I mean? It is a totally exceptional situation in mobile telecoms, that Nokia has such a massive lead in its own industry, that even when it loses one quarter of its market, we ‘do not even notice’ because before this happened, Nokia was the biggest, and after it happened, Nokia is still the biggest. Nokia still today is bigger than RIM and Apple, the two nearest rivals making smartphones – combined.” 
“We saw it in the UK customer satisfaction survey last year, when even those Nokia current owners who did say, that they wanted their next phone to be also a Nokia – were not willing to recommend Nokia to friends – they were, in effect, ashamed to own Nokias. This is not a recipe for market success in loyalty. Nokia has just about lost this front, of the war already. Compare that to Apple owners – who will stick their brand new iPhone in front of the noses of anyone, to show how incredibly cool and clever their latest iPhone is. That is the difference. While Apple owners proudly display their iPhones, Nokia owners hide their gadgets in shame.” 
I’ll include here analysis on what was causing it, also by Tomi Ahonen:
“Now lets go back to history. From Q1 to Q2, Nokia profit margin and ASP for smartphones declined, but its market share grew! From 37% to 39%. We know after Q2, Nokia’s market share has been in free-fall and its average sales price and its handset profit margin has not recovered in similar proportion. What does this tell us. I think I know the answer for the market share crash. I think the evidence is overwhelming and the pattern is perfect. OPK had been buying market share into early 2010, by propping up an un-sustainable level of market share – by heavy discounts and/or marketing support of carriers/operators.“ 
Thank you, Tomi. This was situation at end of 2010. Nokia was rapidly losing market share. True, sales amounts were in rise, gut the growth was lagging behind the industry so bad that Nokia was indeed losing 15% of its remaining market share – every quarter. And market share had been bought for beginning of the year by lowering the prices, meaning that probably problem was present in beginning of the year already, just well hidden.
Nevertheless, facts are facts – Nokia was rejected by customers and losing market share. But problem – as Mr. Ahonen has pointed out – was NOT in strategy. Strategy was not behind the market share collapse, it was supposed to fix it in due time. But then market share collapse hit hard to low-end (less than 150€ price point) where Symbian still had hard time to reach and where MeeGo was not aimed at all. Cheap Androids were emerging and eating Nokia’s market share alive. Nokia simply did not have competitive devices in sub-150€ price range. This reality was slammed against the faces of Nokia management.
STRATEGY AS IT WAS LAID OUT IN FEBRUARY 2011
In beginning of 2011 Nokia’s management looked at the same market numbers we have seen. For fact we know they chose to use Windows Phone for their high-end smartphones and killed MeeGo, slowly letting Symbian die away. (I don’t know why Windows Phone, but it’s not my job to judge. I assume it is for the tablet/PC/phone harmony.)
About here we need to start adding facts, since those are obviously missing from majority of the reports. What we read from news is Microsoft muppet emerging and destroying everything from the way of Windows Phone. I say something else:
WHAT WAS STEPHEN ELOP THINKING?
Let’s keep the idea of Microsoft muppet alive for a moment. And remember, Windows Phone is the OS of choice and MeeGo is killed, Symbian too. Then we can ask:
- What was Stephen Elop thinking when he ordered MeeGo device to be launched in the hot-selling N9 form it was?
He was supposed to kill MeeGo, remember? Sure, Intel deal apparently required launch of a device, but he was supposed to make sure Windows Phone will look superior to all options. He was supposed to force N9 to use some über-ugly form factor like Nokia X6 and spoil the swipe UI completely. What was he thinking when he let out iPhone killer instead of ugly duckling?
- What was Stephen Elop thinking when he ordered N9 to be supported by series of updates that make it even hotter than it already was?
We just got SW version 1.3 out. “over 1000 improvements“, they say. What was he thinking when he let company spend all that money to keep MeeGo developers continuing work when there were no more devices to come?
- What was Stephen Elop thinking when he ordered Symbian updates?
First Symbian Anna. Then Belle. Now we get leaks of Nokia Belle Feature Pack 2. Stephen Elop could have dropped Symbian Anna to begin with. Why make the “burning platform” more competitive? Why spend once agin large amount of cash to make dying platform more appealing? What was he thinking?
- What was Stephen Elop thinking when he kept Qt operational?
WP was not supposed to get Qt. Qt was to die. Why did he not sell the asset to begin with?What was he thinking when he kept the company along?
NOKIA STRATEGY FROM FEBRUARY 2011 – MISSING MEMO
Let the speculations begin! I have no visibility to Nokia internal happenings since Q3 2011, so I can only fill in so little. And my NDA won’t let me do even that. Therefore I’ll link you some stuff from the web:
The answer to all this seems to be Meltemi. Meltemi has never been admitted by Nokia. But we have leak from September 2011 saying it was “Linux-based operating system code-named Meltemi, the Greek word for dry summer winds that blow across the Aegean Sea from the north. It is being led by Mary McDowell, the handset maker’s executive vice president in charge of mobile phones” 
Let’s put this to schedule: when leak happened Elop had been in charge for a year. It is not that he had not killed Meltemi, HE WAS IN THE HOUSE WHEN IT STARTED! And it really makes sense strategy-wise:
- Make a hot high-end device, get the positive market buzz for swipe UI and profit WP a bit with the HW form factor.
- Keep the platform alive, lure in developers to port Symbian apps to MeeGo.
- Refresh the Symbian and keep it attractive. Make sure it sells past the transition period so developers keep doing apps to it (and MeeGo).
- Keep Qt alive so migration path is there.
- Take MeeGo, optimize the SW, improve the UI, drag down the HW requirements, optimize more. Win back the low-end.
Sounds sane. And it’s not conflicting with WP strategy since WP will never reach those price points. Sounds sane.
Okay, here is the point where Tomi Ahonen fan club say that Elop did not do that and propose some weird heroic story how Meltemi was kept alive under rule of a tyrannic despot. Try the leaked memo from April 27th, 2011. It says very clearly “There will also be opportunities within the Meltemi organization, for personnel working within the MeeGo teams.” 
Yes, you read it correctly: Stephen Elop himself said Meltemi team is going to be increasing headcount when MeeGo and Symbian were shut down!
SO WHAT WENT WRONG?
What next? For fact we know that Nokia’s market share started to dive after Q2 2010 and corrective actions were needed. I listed a bunch of corrective actions. Still the dive never stopped, so something went wrong. As said, my vision to Nokia inside happenings ends way too early. But I can still read the signs. What comes to unverified rumors, Meltemi was cancelled recently. That is bad. really bad. But it’s just addition to list of failures:
- Symbian Belle should have been there already during first half of 2011, both for new devices and as update to old ones. That would have slowed down the decline of market share. The strategy change messed up Symbian development and Belle came out second half 2011 for new devices and Q1 2012 for updates. Belle failed to deliver. (And let’s not forget Elop Effect contribution).
Let’s remember that Belle was (and still is) developed in Accenture. Now that Nokia strategy breaks down on Symbian side, we hear about layoffs in Accenture. And of course Nokia’s market share kept going down.
- Transition was supposed to be smooth. Communication mistake of the century blew it. Now it did not cause the crash of Nokia – as I said already – but it certainly did not help. The developer community jumped to other platforms as suddenly Symbian was dead and Qt was a dead-end. Nobody told there was still need for developers and app ecosystem. Elop failed to deliver the strategy. Nokia’s market share kept going down.
- From what we’ve heard, MeeGo was planned to launch late 2010 but was delayed.  It could have slowed the market share decline by being there in time, but it was late already when new strategy was announced. MeeGo team failed to deliver. And market share kept going down.
- Meltemi should have been ready to be announced Q3 2011 and ready for launch Q4 2011.
Edit: Some new information came up in comments and I try to reflect it here.
For what was originally going around in web, Meltemi was “MeeGo lite” – take existing platform and develop on that. What we now hear is that Meltemi was OS built from scratch. Let’s keep the focus here: surely nobody wanted to do THIRD Linux-based OS from scratch? Something went wrong during the project. We cannot know what, but comments section gives horrible image of the path that was taken during the months it took. Obviously Meltemi team failed to deliver. And it seems that is why Mary McDowell got kicked out. (Elop wanted Meltemi, remember?) Market share reaction will be seen in Q3 and Q4 2012.
- And last: Original MeeGo plans had new OS versions (and new products). Instead it was cut down to 1.0, 1.1, 1.2 and 1.3. Originally Meltemi should have been public already when N9 hit the shops, so it would have been piggy-backing on the updates. (Assuming the original plan being MeeGo based). Ironically – in this case MeeGo team DID deliver. But others did not, so it does not matter.
COULD IT STILL BE DONE?
For the life of me I cannot figure out why they killed Meltemi. It obviously was the lifeline of the strategy. Only reason I could think of is the developer ecosystem. Nokia management never really understood how fragile developer ecosystem is. They just assumed that developers will happily start doing apps to Windows Phone. After February 11th 2011, those developers dropped Symbian but they did not move to WP, they went for Android and iOS. Nokia practically threw away its developer ecosystem at that point.
So Tomi Ahonen says Nokia should return to MeeGo path. At the time Meltemi was supposedly killed – June 14th 2012 – it has been 16 months since strategy change and Qt developer community shock. What can we expect to be left of Qt developers doing something on Symbian and MeeGo? Not much, really. SW companies have had full year and more in between. Any sane company switched focus during 1H 2011, fired people, hired people… They don’t want to go back. And it cannot be pushed. Nokia was able to push WP ecosystem with Microsoft. They could put enough money together to make it happen. But Nokia can not market fourth ecosystem anymore. They just won’t have cash for that. Especially since it was burned to all previous activities that failed to deliver.
That cash would have flowed back after launch of Meltemi devices, so reason to cancel cannot really be financial reasons either. Only reason I can think of is the lack of ecosystem.
And it answers to the question:
no, MeeGo and Meltemi are dead. They died already 2011. With successful transition, perhaps they would be alive and keeping Nokia as a market lead. And AFAIK, MeeGo employees have been tossed out and Meltemi employees will be. It would need more money than Nokia has to cover the costs of “back to MeeGo” strategy. MeeGo is gone, due to many failed deliveries. Sorry to say that, but Nokia failed to open the door to low-end smart phone market.
I guess that is about it on speculations.
HOW ABOUT N950 (AND N900)?
Now let’s start with N950. If you don’t know what it is, it’s “the other MeeGo device”. Photo of it is available in e.g. Wikipedia.
N950 is basically N9 with slide-out HW QWERTY keyboard. Tomi Ahonen repeatedly demands it to be launched to all markets. Problem is that apparently non-critical N950 bug fixing was skipped to speed up N9 development.
Let’s have an example: N9 SW loses “send” button from the on-display keyboard when it is dismissed. On-display keyboard gets dismissed if HW QWERTY keyboard is activated. So if you open HW keyboard when on-display keyboard is already visible, you can’t send your SMS because there is no send-button. (Interestingly enough the send-button is visible if you enter messages with HW keyboard open)
This kind of bugs are not major issues. These can be fixed. But launching N950 would need exactly the work that was not done in order to deliver N9. And that work is needed from employees who were already laid off. So in my opinion, N950 won’t see shops. It is delivered to some specific developers and that’s it.
And I’m sure everybody knows N900. Tomi Ahonen reveals us his incompetence in technology from time to time with statements like this: “older Maemo-based handset the N900 can also run MeeGo. So while its hardware is older, its camera smaller etc, that could be sold as the ‘entry-level’ model into the MeeGo line, like the older Apple iPhone 3GS is the cheapest model in Apple’s iPhone range.” 
Okay, Tomi Ahonen is right in a sense: e.g. Mer (used by Jolla) runs in N900. But let’s remember that Tomi is pushing the wonderful Swipe UI of N9 to it. And (in his words)
“Because the N9 is currently in production and selling highly profitably (and sister phones such as N950 and N900 have been produced and could be manufactured within weeks of ramp-up time)”
Within weeks of ramp-up time? Sure, the UI used in N9 could be reconfigured to N900 and architeture could do it… but weeks of ramp-up time? Wrong display resolution breaks the UI, new graphics or UI design is definitely needed, but it’s not the worst things he ignores. Worst thing is that he forgets N900 has resistive display so it cannot detect multiple touches. So long pinch-zoom! (So long any zoom in applications where volume keys were not followed for it.) N900 used “rotate gesture” in browser, but it’s not existing in N9 MeeGo UI, so we are talking about additional rework on top of one already mentioned for N950 (N900 also has QWERTY keyboard).
Oh yes, and the fact that Tomi is so fond of the Swipe UI. (Who isn’t? All it really takes is a swipe.) But swipe gestures lose their idea as you need to press noticeably harder on the resistive touchscreen for it to recognise touch. “All it takes is a swipe” just became “all it takes is a swipe and stylus”. The claim is without base. CPU line may be same but the device is unable to run the N9 version of MeeGo (and Swipe UI) “within weeks of ramp-up time”.
N900 will not be the low-end sister product you wish for, sorry Tomi.
Tomi Ahonen drew new rock bottom on this one:
“And if some independent programmers were able to do an Android port for the N9 this fast, why could Nokia not move to Android, rather than stay with Windows, and release the N900, the N9 and the N950 on Android, like next month! Like yes, in August!” 
I wonder where he gets those. Expecting that would be August 31st, Nokia needs to release that port in… 42 days!
Now I like to have correct figures used. And this is my area of expertice. I gave (valid) estimates for SW projects the last 3 years in Nokia. I know Nokia working environment. I know this field. Here are more proper numbers. (Next time Tomi, use these):
3 months – assuming that Nokia still has all the MeeGo experts in the house (they don’t) and everybody would work on the bugfixing. You see, there WILL be mandatory test rounds.
4 months – assuming People working with MeeGo PR1.3 and Meltemi can be used instead. (Don’t know if they can)
Releasing N9 Swipe UI (and mandatory work-arounds due to resistive display) on N900:
6 months – assuming N950 was finished first. And assuming that Nokia still has all the Maemo and MeeGo experts in the house (they don’t) and everybody would work on the port.
Not less than 9 months, more likely 12 – assuming People working with MeeGo PR1.3 and Meltemi can be used instead. (Don’t know if they can)
Releasing N900, the N9 and the N950 on Android:
Wow, bad one – this is first Android project for Nokia, so it starts on scratch. Now let’s assume that we can use the existing port and do the necessary bugfixing. Let’s assume we have Android experts in house. Let’s assume we have the tools.
I’ll give careful estimate of 6 months for N9 and N950, as we have the existing port.
For N900, one question: Does Jellybean run with resistive display?
Depending on answer – 12 months? 18? Really can’t tell.
I have been asked how Windows Phone fits to this picture.
It is obvious Nokia will push Microsoft to deliver low-end WP. They have to, since other plans are gone. Unless WP8 brings in new HW configurations also downwards, WP7 is the one to use. And WP7 does not support MicroSD cards (7.8 update is rumored to bring in BT file transfer), so it lacks some key elements of low-end market. And HW requirements are too high. So it’s up to WP8. If nothing comes up, no low-end smartphones from Nokia.
(Moore’s law makes HW cheaper but since also Android benefits from that, Android can always go lower than WP).
All my best to anyone who is still working in Nokia – you are the heroes. Keep the company from sinking, as hard as it may be.
 Sherlock Holmes & Hound of Nokiaville
 Wall Street Journal reports Meltemi
 Nokia cuts memo – The Register
Guideline for commenting: I hate the way Tomi Ahonen deletes criticizing comments from his blog. However, I plan to follow three of his principles: I’ll delete comments that are
- Personal insults to someone
In addition, if you wish to challenge my previous posts, please comment to those.