Operation Elop: The final years of Nokia’s mobile phones (2017)
medium.com
medium.com
Nokia had all sorts of structural issues, true, and a bewildering array of self-competing products, but it also had a lot to work with and the brand recognition was absolutely enormous. He effectively flushed it all down the toilet for a punt on Windows Mobile, which was already looking like an "also-ran" by that point. When it (predictably) didn't work out, MS took the brand and some of the company, then ran it into the ground.
A sad end to what had been a huge success story.
(I'm not challenging the idea that nokia was on a downward slope by the point he came along, but he just accelerated it, see e.g. the burning platforms memo)
Around the official death of Windows Mobile, most of us just wanted them to launch hardware that wasn't a joke.
This is Android based on my eight years with them.
I won't argue with you on iPhone, it's a great "I just want it to work" option. And since there's no other alternative, it's what I carry today.
But Windows Mobile was a breath of competency in a market of basically all bad Google phones.
The first modern Windows Mobile phone launched in November 2015. The market of Google phones at that point in time was the Nexus 4, Nexus 5, Nexus 5X, and Nexus 6.
Now I'm not a huge fan of the Nexus 6 (though I certainly wouldn't call it "bad"), but the Nexus 4 and Nexus 5 are still to this date the best value smartphones ever released.
Everything else outside of the Chinese market has Google Play Services with root access to the system and as controlled by Google's MADA agreement with manufacturers which has already been found illegal in multiple jurisdictions across the globe.
Convention dictates that "Google phones" specifically means e.g., Nexus/Pixel devices.
Fortunately, if you're unwilling to distinguish between hardware and software, there's already a very convenient and unambiguous term for mobile devices that run Google's OS. In fact, you just used it in the same sentence!
I like not having apps pester me for updates or pollute my notifications, when all I care about is email, websites and phonecalls (oh and offline maps, I almost forgot)
I guess I am a very boring person!
That's the problem with anecdotes... they only represent your very narrow window of experience while ignoring literally every other person on the planet.
> but I don't know a single person that wanted a Windows phone
I only represent a single person, and I make no claims about anyone else!
...no apps. Windows Mobile just couldn't attract quality developers to its platform. After witnessing that - I fully understood why Android went with Java (the language): there was already a deep pool of seasoned Java developers.
For instance, Apple forced people to write iPhone apps in Objective-C for years and nobody avoided the platform because of that, even buying Macs and paying 99 dollars a year for the privilege of writing and selling iOS apps.
>That's the problem with anecdotes
I'm not saying those problems affected Windows phone, but that's the association customers have with Windows as a brand due to decades of facing those problems with desktop Windows.
With each new generation of Windows Phone, Microsoft not only closes the gap with iOS and Android in important ways, but it also differentiates in important ways
WINDOWS PHONE 8 REVIEW
https://www.theverge.com/2012/10/29/3570494/windows-phone-8-...
I appreciated the consistency in design and that even the cheap devices like the 635 felt quite snappy. They also seemed a bit more coherent with update support. Yes, the no WP7->8 update path was a huge mistake, but I think by the end they were updating the OS for all WP devices (capable of being updated) centrally, rather than relying on the OEMs; the WP10 update betas I tried seemed competent enough but there was a growing sense their heart wasn't in it.
On its heyday (approaching 10% in Europe), Windows Phones were seen as the alternative for those that wanted an Android alternative and didn't care about iPhone.
The discipline also extended to heavy sandboxing of the apps so a lot of functionality was simply unimplemetable, such as reminders for missed notifications.
But the phone worked, the UI was intiutive and responsive, and the only reason to move on, other than general phone obsolescence like the 5MP camera, was the lack of mainstream apps.
We had a brief experience with WP10 on another phone, and it seemed spiffier but less stable. WP8 was the high point.
I guess abandoning Symbian to switch to Microsoft should be considered a significant misstep, even if motivated by the then financial motives.
Porting Symbian UI to Qt meant that they could at least merge surface-level stuff and have core apps that ran on both Maemo/Meego and Symbian, then retire Symbian when no one would care or notice.
I think this is correct. It had a traditional market base but was already hopelessly noncompetitive except for old reputation and brand value (see my other comments here) and on a steep decline. Compared to iPhone and Android which were more modern and on a steep growth path.
(Maemo, the precursor to Meego, was originally a GTK-based platform. The Symbian faction of Nokia acquired Trolltech in order to be able to advocate for Qt to be a shared development platform for both Series 60 and Maemo, which meant throwing out large parts of the existing Maemo codebase and starting again. This delayed Maemo significantly, and somewhat ignored the fact that there were basically 0 apps on Symbian that anyone cared about in the slightest)
Around 2003:ish Nokia was the only Symbian licensee left and they had even 3 UIs: Series 60, Series 80 and Series 90. Only Series 60 survived. Series 80 had 2 products and Series 90 one public prototype. Series 90 was the only with touch input originally. Series 60 got touch after the iPhone, but that was too late.
Yes, Symbian C++ was too difficult even inside Nokia. Productivity was very low, many products got cancelled before seeing the market and those who did were massively delayed.
Agree with the ugly. Well, beauty is in the eye of the beholder. Some say it was an engineering master piece of the 1990s. But it was different from everything else and it was difficult, I don't think this can be disputed.
The no good reason is not the whole truth:
* When Symbian (then called EPOC32) was developed there was no C++ compiler that supported exceptions without major problems. "Don't use them" was typical in style guides at the time. So EPOC came up with their own macro hackeries instead, which worked when used correctly. The latter was too difficult for many developers.
* Early devices had just single digits of megabytes of RAM. So 500 times less than a phone of today. More macro trickeries to save Bytes were the answer.
That's definitely not true. I joined Symbian in 2005 and there were lots of licensees around. I think 2005/2006 were the glory years of the company.
That's not fully correct. Symbian was the leading smartphone operating system years 2000 - 2008 or so in Europe and some parts of Asia. Skype, browsers, games existed. I remember Skype offering me a refund when they discontinued the app.
I remember visiting software conferences in the US maybe 2005:ish and all the Americans had only feature phones. They seriously told me "All I want to do is call and text, I don't need anything else". The truth was that packet data coverage sucked in the US in those days.
These were Smartphones with keypad navigation (qwerty models also existed). They had packet data and the possibility to install apps, both native Symbian and Java (J2ME MIDP).
The problem was that developing for Symbian was too difficult for many programmers. Not only outside Nokia but also inside. I would claim that at least 40% of the Symbian developers did not fully understand how strings worked. Which had several types, none of which was called string and conversion between them or using the wrong one was a continuous source of bugs.
MIDP was too limited in its sandbox to build truly competitive apps.
When the iPhone came and capacitive touch started to win Nokia had nothing to offer. They had a couple of models with resisitive touch (requiring a pen or long nails), but all too little and too late.
When the Smartphone development started in the US Nokia was just obsolete. No competitive hardware and a difficult software platform. So Apple and Google conquered the world.
Depends on the time. It must have been christmas 2002 that I took my first skiing videos using the video recording app. Probably still in beta then. And I sent the app to the recipient by Email. Sending the videos went by MMS. Max size 50 KB IIRC, maybe 100 KB but I believe that came only in the next model.
Yeah, those smart phones came only with a still camera. But being able to build a video recorder show the strength of the platform. Of course there was no security model, access control or anything at the time, which made things easier.
Of course in 2010, being able to build a video recording app was no longer a competitive advantage.
Are you updating the maps manually or is it still being updated through some web app interface?
https://gigaom.com/2009/10/23/symbian-executives-rips-into-g...
But he wasn't and he didn't. Everything else is 'could be'.
In 2006, my girlfriend at the time worked in HR at Nokia's US headquarters in Irving, Texas. The writing as on the wall and the company was preparing for a major layoff because Nokia's mobile phone business in the US was tanking. Luckily for her, she was able to transfer to Nokia-Siemens Networks. This was before the iPhone was released and before Elop joined the company in September 2010.
The critics and the customers loved N9. I would have loved to own one. Instead, it was silently killed off without even being given a chance.
N9 was Nokia adapting to the smartphone/touchscreen ecosystem.
(a) N9's sales were a drop in the ocean. It is like if you have a game with 100% good customer reviews. But there are only 50 reviews. The tiny but vocal segment who loves it, loves it. That doesn't matter.
(b) There was no room for a third ecosystem. Developers were busy writing apps for Android and iPhone. Nobody would write apps for (random 3rd OS that has a couple of million phones). It wouldn't matter if the N9 had had working desktop fusion and teleportation. The platform was dead.
(I'd tend to agree with the second point that MeeGo would have had difficulties and likely would have failed later, even if they'd focused on MeeGo and the N9, but there was something to it)
The N9 had a very good reception, and with Nokia's prestige and loyal customer base it would have been possible to compete against the iPhone. The N9 built on top of the N700 series, which was launched ahead of the iPhone. It was essentially Debian running in your pocket. It already had a huge "app store", all Linux repositories. Polishing this was not that hard, given that they also controlled Qt. Qt also meant a simple migration path for Symbian applications.
People don't realize how mature the N9 was. Read some reviews. Their maps application had incredibly good offline coverage and offline routing in 2011. So good, I've been using it for that purpose till 2020 as nothing compares. Mind that Google Maps still licenses lots of mapping data from Navteq, which was Nokia's mapping spinoff.
Aside from this, even if they wanted to go with Windows Mobile, it sounds like a really bad strategic move not to have a backup. Especially because going with Windows Mobile was a bigger gamble, which ended up failing spectacularly. Going ahead with the N9 didn't use many resources. Windows Mobile devices by Nokia would eventually reuse the same hardware. Why not selling both.
Did you mean N900?
> It was essentially Debian running in your pocket. It already had a huge "app store", all Linux repositories.
They dropped Debian as a base and switched to Moblin, which was one of the problems. Also it is telling that you refer the community repos as app store, and ignore that it also had actual proper app store too ("Nokia store"). The utter failure of Nokia store was another glaring issue that they never figured out. Critically the community repos, no matter how awesome, did not support commercial apps in any way, nor did it bring any income to Nokia.
> Polishing this was not that hard, given that they also controlled Qt
Yeah, nope. 80/20 etc.
The sales of a future hypothetical Windows phone were far less at that point.
> There was no room for a third ecosystem. Developers were busy writing apps for Android and iPhone.
Yet Nokia was attempting to launch a third ecosystem, just not the one that anybody wanted.
If you agree Windows phone was never going to work, then you need a story how Meego could have worked when Nokia had a fraction of the resources to put into it.
Windows phone was not going to work because the brand is herpes, and Microsoft makes technical decisions based on business strategy rather than merit (and everybody know it.) It had nothing to do with the amount of resources put into it, which I'm sure were vast; Microsoft is happy to lose money in markets for a decade or two in order to block new entrants, waiting for the market leader to make a mistake. If people wanted a phone that would trap them into a rent-seeking ecosystem, they already had a choice in iPhones that was at least attractive and common.
Meego phones could have slotted into the nerd/serious phone category that various Android distributions have awkwardly tried to fit in. It would have also been a non-US alternative for markets like Europe and China. They also could run Android apps within a VM.
https://dominiescommunicate.wordpress.com/2012/09/01/n9-outs...
Compared to iPhone, N9 had a practically zero marketing budget, and an announcement at launch that the platform has no future as Nokia is going all-in on a Microsoft partnership.
As far as an apples-to-oranges comparison can be made, N9 did great.
However, management had reasons. There was no room for a third platform.
I also note the unfalsifiability of some of these claims. On the one hand if N9 beat the Windows phone, that shows how awesome it was. On the other hand, if it was destroyed by the iPhone, well, it was sabotaged.
For context: less than 18 months earlier the phone manufacturers had finally agreed, after a long consumer campaign, to all support combined data and power over a single cable. This allowed to get rid of custom chargers everybody hated.
Verizon's specs at the time still mandated that any phone they sold had to have separate power and data sockets. How do I remember? I spotted this brilliant piece of malice in their specs and raised it with the project lead, who then took it higher up.
Jobs was able to bring his team and technology from NeXT to become the new core at Apple, and fire everyone who didn't want to play along. Elop had no such luxury.
I wish I could blow my failed ventures so horribly. And it indeed was a failed venture by the time Elop got there.
Nokia kept the parts of their business that actually were worth something, and used the MS money to buy more marketshare in networks. It was a good deal.
That said, he did make a large number of unforced errors, mostly notably the Burning Platform memo, which accelerated the decline.
That also seems quite an oversimplification, and one that shifts the blame from leadership to the employees.
From what I’m aware, Jobs instead actually came back and axed the vast majority of projects that were being worked upon. I’m fact, there’s a relatively famous town hall where a Mac user complains about this and Jobs explains why he needed to cut nearly everything.
I call that BS.
The market was shifting to Android. Industry veterans like Jean-Louis Gassée wrote that during OPK era, he was approached by Nokia board asking for advice, and he told them to go Android. The board refused, and it was decided to bet the company on Windows Phone. Not for some technical or organizational/cultural reason, but purely for strategic reasons.
https://web.archive.org/web/20121225214448/http://www.comput...
German carmakers are waking up late to electric vehicles, but they are acting now to pivot to the winning technology, before their marketshare reaches 3%. Nokia bound themselves for a multi-year exclusive contract with Microsoft, that even Microsoft folks were surprised they would enter at such a minimal payment for MS. I remember reading reports that the Microsoft negotiations team was authorized to go considerably higher than what they ended up paying.
So it was overconfidence on Nokia's part to what extent they would be able to sell the turd that was Windows Phone, coupled with the lack of exit strategy if that didn't work out. And that's even before talking about Elop's financial incentives to run the smartphone business into the ground and sell it to Microsoft.
> On October 7, 2014, two Finnish journalists Merina Salminen and Pekka Nykänen published their book Operaatio Elop in Finnish, probing into the events that took place in Nokia’s device business under the CEO Stephen Elop’s period in 2010–2013. The authors had interviewed over 100 people for the book, most of them being current or former Nokia employees... The full English translation is now published here the first time, under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.
> For readers who prefer a Kindle or PDF version instead of this online version, we provide exports in PDF/EPUB/MOBI [1] formats from Medium.
A thread from (more or less) back then: https://news.ycombinator.com/item?id=16468645
Reading him branded as the worst CEO in the world makes me feel bad for him considering all the nice things I've heard about him. It is easy to throw around insults and labels, but let's remember that the truth can be nuanced and not so simple to explain.
Edit: fixed the political title based on comment below
I think the people who ultimately led to Nokias downfall worked at 1 Infinite Loop where the iPhone was developed. Prior to the iPhone, Symbian was the dominant "smartphone" OS in Europe, or at least competitive with Blackberry, but the carriers shut Nokia out of the high-end phone market in America. I think Elop's strategy was to win over the American carriers/market but the strategy backfired and accelerated the exodus to iPhone/Android in Europe while not gaining a foothold elsewhere.
Apple left the door open for one or more alternative smartphone vendors. Nokia and Blackberry dropped the ball and Google filled the gap with Android that Verizon marketed splendidly. I'm not sure if Elop's strategy could ever have worked but it certainly had no chance when both Nokia's employees and customer base resented the approach.
I have not read the book so I'm not sure if my views mesh with what was written there.
Agreed wholeheartedly. If you want to know more about Verizon's kingmaking process to go all in with Droid over their previous partner RIM, see the book 'Losing the Signal', an excellent read on the rise and fall of RIM.
I always found those reports about the iPhone causing Nokias downfall to be not credible. Nokia's vast majority of sales were in price segments where the iPhone didn't even compete.
The iPhone destroyed its North American competitors at that time, Blackberry, Palm, etc.
Nokia was clearly destroyed by Android. Surveys at that time found Nokia customers mass defecting to Samsung and other Android vendors.
> I think Elop's strategy was to win over the American carriers/market
A foolish plan, something Nokia never had accomplished even in the heyday of Symbian, when it dominated the rest of the world..
For which he received a $25.5m bonus, 70% of which was paid by Microsoft [1].
[1] https://www.bloomberg.com/news/articles/2013-09-19/nokia-s-e...
https://www.theglobeandmail.com/report-on-business/careers/c...
Back then, Windows had about 2% share in the mobile phone market. Betting literally Nokia's entire future on Windows Phone seemed reckless, given that the competition (most notably Samsung) was being so successful with Android.
And this exclusivity was one-way: Nokia would commit to Windows Phone (or death), but Microsoft would continue to happily supply Windows Phone to Nokia's competitors, like Samsung.
So while Samsung, HTC, and the others were building both Android and Windows Phone devices, Nokia was going to go down the Windows path exclusively. How was that ever supposed to work out?
I look forward to reading this book, because Elop's decision bet Nokia's future on 2%-market-share-Microsoft is something that I have never understood. And I may be arm-chairing, but Nokia's demise went exactly as I expected it to go from day 1 of this announcement.
WinMo was a _very_ risky bet with little upside and huge downside.
[+] Android, ofc.
"Calculating the Elop Effect: He's already destroyed a company the size of Oracle, and profits the size of Google"
https://communities-dominate.blogs.com/brands/2011/12/calcul...
Tomi Ahonen is a Finn with personal relationships to the old Nokia guard. Spinning the story as Elop's failure was a convenient way to let them off the hook for destroying $100B shareholder value in 2000-2010.
Regarding Ahonen's past, look at his accuracy over the years in all things mobile phone marketing when many other supposed experts were floundering, particularly about Windows Phone.
E.g. in 2010 when smartphone apps were already a thing, he was gung-ho about MMS: https://communities-dominate.blogs.com/brands/2010/06/an-inc...
"Compared to Facebook, MMS reaches a 3 times bigger audience - its lead is radically bigger compared to smaller social networking sites like MySpace, Twitter, Flickr etc. The same is true of the total installed base of all brands of smartphones. If you only think of apps for Apple's iPhone including the total shipments of all iPod Touch and iPad devices, MMS will reach an interactive multimedia mobile phone audience... that is 20 times larger."
If you had invested based on his analysis, you'd have got a business built around last-gen messaging technology just as it was being wiped out.
[1] https://www.macrotrends.net/stocks/charts/NOK/nokia/revenue
The problems started already years before, but Nokia management was successful for a while in hiding its dysfunction behind the immediate effects of the 2008 crisis.
It could have been salvaged, but there were issues if you know what you are looking at.
Even looking at the Nokia finance numbers you posted the issues are clear. Both revenue (-2%) and profits (-6%) are down in a growing market, and the bigger drop in profits is indicative of them losing high end market share.
Nokia was in a similar boat.
I know a lot of Windows fans have convinced themselves Windows Phone was awesome but it wasn’t.
Avoiding Windows Nokia could still have been alive by going for Android. Sure a smaller company than in the heyday but still alive.
The billions in cash that they made from selling the loss-making mobile devices albatross to Steve Ballmer allowed Nokia to purchase both Siemens’ share of the networks business as well as Alcatel.
It's true to say that Nokia was in difficulty and initially laughed at the iPhone when it was first released.
Elop may have had the right idea but how that idea was delivered should be a lession in how not to do things. I remember the Burning Platform memo appearing on the intranet homepage. Instantly thousands of people stopped caring. They just hung around waiting for Voluntary Redundancy offers. Truths were told, it got very dirty.
If anyone wants any information on how it felt on ground level I’d be happy to answer any questions.
I recently visited the building and it is no longer Nokia owned, quite sad as it was impressively beautiful inside. Now it’s segmented with lots of smaller companies taking chunks of the building. :(
(It was posted to my Reddit account).
I will try to find some good public photos that represented reality.
https://upload.wikimedia.org/wikipedia/commons/6/60/Le_centr...
"The institutions already mentioned - lively and productive as they may be - flourish in such shabby and makeshift surroundings that we might turn with relief to an institution clothed from the outset with convenience and dignity. The outer door, in bronze and glass, is placed centrally in a symmetrical facade. Polished shoes glide quietly over shining rubber to the glittering and silent elevator. The overpoweringly cultured receptionist will murmur with carmine lips into an ice-blue receiver. She will wave you into a chromium armchair, consoling you with a dazzling smile for any slight but inevitable delay. Looking up from a glossy magazine, you will observe how the wide corridors radiate toward departments A, B, and C. From behind closed doors will come the subdued noise of an ordered activity. A minute later and you are ankle deep in the director's carpet, plodding sturdily toward his distant, tidy desk. Hypnotized by the chief's unwavering stare, cowed by the Matisse hung upon his wall, you will feel that you have found real efficiency at last.
In point of fact you will have discovered nothing of the kind. It is now known that a perfection of planned layout is achieved only by institutions on the point of collapse....
The intention to found New Delhi was announced at the Imperial Durbar of 1911, King George V being at that time the Mogul's successor on what had been the Peacock Throne. Sir Edwin Lutyens then proceeded to draw up plans for a British Versailles, splendid in conception, comprehensive in detail, masterly in design, and overpowering in scale. But the stages of its progress toward completion correspond with so many steps in political collapse.... It would be possible, though tedious, to trace the whole story down to the day when the British finally withdrew, showing how each phase of the retreat was exactly paralleled with the completion of another triumph in civic design. What was finally achieved was no more and no less than a mausoleum."
I had a wonderful time, but what a broken company...
There was an anecdote being sent around about how it "was the most expensive clock application ever to exist" due to rewrites.
I agree, lots of great people working there, but I think very little leadership.
Isn't that a good sign of Finland startup ecosystem, I read that Ex. Nokia employees are the crux of leading startups in Finland.
But the fact is that Nokia had become a massive (if lucrative) black hole for tech talent. A very large part of the Finnish technology scene was working for Nokia, either directly or indirectly. There was little room for small-scale innovation, because Nokia projects just hoovered up everything.
Finnish startup scene of today is a direct beneficiary of Nokia's collapse. All that talent had to find something to do to pay the bills. They knew how technology worked, and had ideas that had never got tried out.
Slush already existed in 2011, but it got huge because from 2012 onwards there was so much desperate energy directed at making new ideas work.
Coming from country where it's mostly outsourcing, one earns scraps and jumps the ship in case of turbulences, it impressed me how employees can work in their collective interest.
I originally watched from a different source, but looks like it’s also on Coursera and I thin YouTube as well.
https://www.coursera.org/lecture/critical-management/the-ris...
American firm might buy Nokia Networks to get US back into the 5G infrastructure business. Reefshark SoC with Intel failed spectacularly and killed profit hopes from Nokia Networks.
Nokia's telecom software and services does well. Small in revenue over Huawei and Eriksson.
Nokia Technologies/Bell Labs makes money from R&D IP without a product. Selling licensing for Nokia branded phones is nice extra.
Fixed/IP/Optical/Submarine networks and other parts have little common with the rest.
If Ericsson and Nokia were to merge or Huawei would attempt to buy Nokia or Nokia Networks that would be different. I think EU would agree even to Nokia-Ericsson merger.
Suri i think is the latest example yet the fault also lies totally in the board of the company. Siilasmaa and co should have resigned alongside suri since they clearly are as mediocre as leaders as him. But it appears that status and ego is more important to a lot of nokia's leaders so it is what it is. Too bad since a similar finnish company of the same size, kone, demonstrates how well the company could be run with better culture and leadership.
I know some people think it was awesome but the rest of us could see it was not a good phone OS.
It had terrible UI choices an so was far too late to the market.
Had Nokia stayed away from Microsoft, they could have saved them self by going for Android.
Nokia married the wrong technology stack and paid dearly for it.
https://communities-dominate.blogs.com/brands/2012/07/the-su...
Nokia had to make some tough choices. They had to go with another platform made by someone else and both Google and Microsoft were really interested in having Nokia as their partner.
Nokia needed money badly and it seems like massive layoff were out of the question. They wanted Google/Microsoft to pay them to put their respective OS on their phones.
Only Microsoft was willing to do that, since they too were desperate to have a strong device maker to push Windows Phone.
In the end, both Nokia and Windows Phone failed.
If you want to play with something similar, Sailfish is derived from Meego and has been ported to various widely available devices.
The design was also great.
Nokia did a very good move in 2008 by acquiring Trolltech (Qt). And then they screwed it royally: instead of immediately started launching phones with Linux and Qtopia (which was mostly ready for mainstream market), Nokia kept launching Symbian phones.
They had this plan to use Qt as the migration platform from Symbian to Maemo. That was a good plan. Except they could have started a lot earlier: why were the Nokia N770, N800 and N810 only tablets and not phones, when what everybody was asking for were phones?
Because back then Nokia middle management would internally destroy any attempt to actually improve or replace Symbian with a serious alternative. Symbian ruled. That's why the first thing Elop did was to fire (outsource to Accenture) everybody working on Symbian. It was a blunt way of saying: "you no longer are Nokia, you will no longer decide what this company does".
Everybody was very excited about Maemo back then. While it was no iOS or Android, there were A TON of developers (and companies) looking into migrating their Symbian apps to Maemo with Qt, or even writing new applications.
And Stephen Elop decided to use Windows Phone instead of Maemo. This is what actually destroyed Nokia: Windows Phone 7.5 was released with no support for C++, which left Qt out, which meant you needed to rewrite your app from scratch. That was the single defining moment where 99.9% of app developers decided they would rather go Android or iOS, and Nokia was doomed from that moment on.
As for the comments regarding the little success of the Nokia N9 or Nokia N950, the reason was they were simply not launched in many countries. When they announced the countries where those phones would be available, it looked like a bad joke.
By the way, another very stupid decision by Nokia, which shows you the power of the Symbian people back then: the Nokia N900 was released in November 2009 with 256 MB RAM and 32 GB storage for 599 EUR. Back then, phones had 64-128 MB RAM (256 MB was rare) and 512MB-1GB storage. Do you know how expensive was 32 GB flash in 2009??? They could have sold this phone for 400 EUR with only 8 GB flash, which was still a lot in 2009, and it would have reached a lot more public.
Really, Elop made the fatal mistake of chosing Windows Phone instead of Maemo. But Nokia had inflicted a lot of damage on itself already. In fact, the only bad thing about Windows Phone was not allowing C++ (and therefore Qt), which rendered all existing apps useless. But Elop probably never knew about this.
Oh, and don't even get me started on another huge mistake: trying for many years to use Gtk instead of Qt. Nokia lost at least 2 years with that. Guys, Gtk was nowhere near Qt back then, and it's even more far apart today!
(I was very very close to Nokia, with many friends inside, in those years and was about to work for them when Elop announced Windows Phone would be the next big thing)