Ex-Nokian comment: I got another guest writer to participate. He wished to remain anonymous too. So from now on, his text:
For quite some time now I have watched people on all sides endlessly argue over what Nokia under Stephen Elop have done, especially around the “burning platform” of Symbian. As the speed of Symbian’s decline has accelerated, the vitriol and bitterness have grown to insane levels. Some commentators are more of a Murdoch tabloid/Fox News, combining outrage, dubious (if not verifiably false) statistics, a heaping helping of circular logic, tied all together with loud volume and attacks/comment censorship on all those who disagree. While my attempts to rebut the histrionics of some bombastic analysts have been unsuccessful (see: the aforementioned censorship), the emergence of DC has given me a space to contribute without fear of censorship. So now I, either a current or ex-Nokia employee (I will not tell which or where), am going to offer my contribution to the debate:
The current dominant narrative goes like this:
- Symbian was doing fine, having just hit record sales in Q4 2010
- Stephen Elop’s “Burning Platform” memo killed Symbian sales by causing the Ratner Effect (hurting sales when the leader publicly says the product is bad) and the Osborne effect (hurting sales by revealing/promoting the future product too far in advance)
- Symbian would have been fine were it not for this, and then we point to charts that show Symbian sales going up until Feb 2011 “Elopalypse”, and then going down from there
- Therefore, Elop and the Burning Platform memo killed Symbian, and with it, gravely wounded Nokia. If it weren’t for this, Symbian sales would not be collapsing.
The argument is compelling because it seems logically sound and easy to explain and understand (and seems to have lots of empirical proof, because look how bad the Symbian sales are now!) But just because an argument seems compelling does not mean it is correct. Everyone in this current argument is having a debate confined to the walls of a windowless sauna and never looking outside. Here is an alternative:
The Nature of Change
When change happens–in nature, in society, in business, whatever–it ultimately boils down to 2 kinds: endogenous and exogenous.
- Endogenous change comes from within–something the organism or organisation does itself.
- Exogenous changes comes from the external environment and affects the organism/organisation.
If someone dies because they are old or ill, that is endogenous. If someone dies because they are eaten by a bear (or hit by an angry bird 🙂 ) that is exogenous.
Now look at the argument, which is never “how Samsung and Apple killed Nokia” but always seems to be “how Elop killed Nokia.” It is always “Elop/Nokia did this, and did that, and did that also.” In this view, Samsung, Apple, Google are like passive blobs–like liquid that will expand to fill any space but only once Nokia itself vacates that space.
Are Samsung, Apple, Google at all passive?
No! They are ferocious and aggressive competitors with billions of dollars in the bank, huge brand recognition, and global reach (all 3 are on display in India where Samsung are offering suitcases of cash to Nokia Priority Dealers that become Samsung shops). As if they weren’t unethical enough with their slavish copying of Apple’s designs…
People debate what Nokia did as though Nokia was the sole factor in its marketplace success or failure. Ironically, those who focus entirely on the endogenous “blame Elop” argument are guilty of the very same Nokia-centric myopia that got Nokia into so much trouble in the first place!
Let’s be good engineers and test the hypothesis
Simply put, my hypothesis is that Nokia’s challenges were exogenous more than endogenous, and even if Nokia had stuck with its original strategy, the sales trajectory of Symbian devices would be crashing like they are now.
How can we test this hypothesis?
- hop in the time machine and make Nokia do the original strategy
OR - observe the performance of peer companies with similar strategy
We obviously cannot do #1 (curse that Elop, he cut the Nokia time machine R&D budget 🙂 ). Although we will get a taste of it with upcoming Jolla phones and seeing if they are a mass market success or a niche product (that I hope is profitable, because even as a fan of Lumia, I love my N9 and want a Jolla 🙂 )
Can we do #2? To do it, we need a company that:
- Was in a similar strategic dilemma as Nokia
a) Strong brand and sales, seen as a big player in the market
b) Sales stagnation in Western markets masked by strong performance in emerging markets
c) Widely-used but aged OS with poor support as a platform for 3rd party app development - Pursued a strategic plan similar to Nokia’s pre-Feb 11 plan
a) Continuing to promote and sell the original OS (avoid Ratner effect and Osborne effect)
b) Develop their own next-generation platform that they control, not WP or Android
c) take sole responsibility for developing the ecosystem around it—carrier relationships, 3rd party developers (apps!), cloud services, PC connectivity etc.
And
d) do so for both the current platform and the next-gen platform simultaneously - Happened in the same market period (2010-2012)
If the hypothesis of the anti-Elop side is correct, a company that pursued such a strategy will succeed in both sales of the legacy platform and development of both ecosystems.
The best part is, we have such a company to compare to! Research in Motion, the venerable BlackBerry maker!
Let us see if they fit the hypothesis:
- Was in a similar strategic dilemma as Nokia
a) Strong brand and sales, seen as a big player in the market yes
b) Sales stagnation in Western markets masked by strong performance in emerging markets yes
c) Widely-used but ageing OS with poor support as a platform for 3rd party app development yes - Pursued a strategic plan similar to Nokia’s pre-Feb 11 plan
a) Continuing to promote and sell the original OS (avoid Ratner effect and Osborne effect) yes. To this day RIM are making and marketing new devices based on BB7
b) Develop their own next-generation platform that they control, not WP or Android yes. BB10
c) take sole responsibility for developing the ecosystem around it—carrier relationships, 3rd party developers (apps!), cloud services, PC connectivity etc yes
And
d) do so for both the current platform and the next-gen platform simultaneously yes
6) Happened in the same market period (2010-2012) yes
Heck, they even used Qt, come from a snowy country, and had Canadian leadership in the CEO office 🙂
Test results part 1 – device sales
So, if change in this market has been primarily endogenous, BB sales should not plummet like Symbian sales have. What happened?
Well that’s awkward. RIM did everything Nokia should have done to promote its old system, and yet the sales still collapsed! The strategies of Elop-Nokia and RIM were totally different. The only constant between RIM and Nokia is they were both facing the same competitors—Apple and Google/Samsung.
Even more interesting is when RIM’s sales started collapsing on this chart. Do you see which month it is? It is February 2011! The month of the “Elopocalypse”! So what is more likely to explain how sales of Nokia’s and RIM’s legacy platforms at the same time?
- Market circumstances beyond the control of any one company caused massive disruptive change and hurt all players that weren’t Apple or Google/Samsung
- Stephen Elop is so bad that he can magically destroy two companies at once with his evil powers
While everyone argues whether Elop set the platform on fire or not, they ignore that the platform was struck by a lightning storm
That is sales. The other side is apps. How has RIM done at both preserving its current ecosystem and growing its new one?
Test part 2 – app ecosystem
For the legacy BB platform, even though more BB7 devices are sold than WP7 devices in a quarter, high profile app developers are announcing the end of support for Blackberry (just like Symbian, massive unit sales do not automatically translate into thriving ecosystem). Earlier this year, visual voicemail provider YouMail announced they were dropping support for BlackBerry (and got into a public spat with RIM’s VP of developer relations in the process). Kayak, the popular airline booking site, also dropped support for BB earlier this year. This week, the New York Times ended their support for BB (by comparison, last month they just released a brand new 2.0 version of their Windows Phone app). And unless you’re on a PlayBook, you still can’t play Angry Birds on a BlackBerry device.
How is the future BB10 ecosystem development going? According to the WSJ, not so great, as RIM is “bleeding developers”. Developer interest in BB7 is declining rapidly and so is support for BB10. By comparison, WP7 interest has dropped after the WP8 announcement, but the decline is offset by interest in 8 (71% of all mobile developers surveyed said WP8 increased their interest in the platform)
Additionally, the announcements of big-time companies supporting BB10 are sparse: Trufone, Gameloft, Poynt, Mippin. Their inability to get a Netflix app for BB10 (a bellwether app or sorts for both developers and consumers looking to see if a platform is “safe”, at least in countries where Netflix has a presence) is a legendary sore spot for RIM. As it is now, Netflix won’t make a BlackBerry version even if RIM did all the work for them. Similarly, TextPlus has more faith in Windows Phone than to BB.
And Skype? Well Microsoft owns Skype, so I don’t think it would be a super high priority for them (especially when they need to do a lot of work to make Skype suck less on WP7 haha 🙂 )
Comparing 3rd party developer support for WP and BB seems to suggest that Microsoft have enough money and weight to ensure that what little oxygen there is for a “third ecosystem” will go to them. MeeGo would have been one more platform competing with MS’s dollars for developer attention for that 3rd place spot.
Some people could say “oh well of course MS will win for 3rd place, they bribe developers with their $$$!” Well, that is almost certainly true (although they do bring more to the table than just money). But that’s not some part of the equation you can choose to ignore for moral reasons or whatever. One way or another Microsoft would win. Nokia could either win with them or lose against them.
CONCLUSION
In this post I have offered a reasonable alternative explanation for why Symbian sales would have collapsed regardless of whether Elop did what he did. It wasn’t Nokia’s strategy that killed Symbian; it was better platforms that killed Symbian—the same thing that killed BlackBerry. It was an exogenous disruption. I don’t expect a lot of die-hard Elop-haters/Symbian-lovers to change their mind in this, but I hope I have made a difference to more cool-headed, open-minded people who want to see Nokia succeed.
I am not saying that Elop made no mistakes. Plenty of mistakes were made by Elop and his executives. Symbian sales would probably have tanked a few percentage points less, but make no mistake they would have tanked. Whether the CEO was willing to say the legacy platform was inferior to competitors or not, the market was willing to say it for him, with their wallets.
Dominies Communicate 
I have already mentioned before that according to Tomi Ahonen “Elop Effect” is a superpower that can alter Nokia products before Elop even was working for Nokia. So I guess I’m voting for option “Stephen Elop is so bad that he can magically destroy two companies at once with his evil powers”.
Well, of course not. But I’ll try to wrap up tomorrow all other magical things Elop Effect has done (like altering past or other companies).
-ExNokian-
They’re indeed different. Nokia already had MeeGo / N9 almost ready on Feb.11, however, RIM still doesn’t have BB10 ready till now.
Talking about apps, nobody will develop anything for a platform that no one uses, and RIM’s sales are falling, while BB10 is in the middle of nowhere. Are you willing to do anything? On the other hand, please give a look to maemo.org, tell me how active the community is! Tell me how many apps are there for N900! Oh, I forgot to mention that there was once a company trying to “port” Android virtual machine to MeeGo 🙂
Regarding Symbian, haven’t you seen Belle, which was ready long ago, and delayed severely because of Feb.11. With a 680MHz CPU and a 256MB RAM, Belly runs nicely on N8, compared to Android phones with higher spec. Of course, it doesn’t matter, because it’s declared dead before it’s released.
@xizzhu
I have Belle on my N8. Frankly the performance is still pretty terrible. Maybe you are just used to terrible performance?
And You also make the classic mistake of confusing an active hobbyist developer community with support from real companies for mass market apps. Your arguments are invalid.
@ JGsmartypants
No, I’m not saying Belle on N8 is perfect. What I was saying is, Belle on N8 is nice, compared to some Android phones with the same/similar spec, release about 1 year ago.
On the other hand, Symbian really needs more fancy animations & transitions, which is one big reason that users think it’s not fluid.
(Don’t get me wrong, I know Symbian is an dead-end, but it’s just unfair to say it’s poor performing compared to Android phones with the same spec.)
For the app ecosystem, how many major apps done by companies did Symbian miss at Feb.11? For MeeGo, again, what if Android virtual machine were “ported”?
Well put. You might like my chart here: http://www.ben-evans.com/post/27045433881/3-year-market-share – which among other things illustrates the fact that Nokia’s market share was sliding for two years before the WP announcement.
Tangentially, no-one takes the Sage of Hong Kong very seriously anymore. Shame – he used to be quite insightful
This is very well argued. The exogenous challenge was evident before Elop took charge. Nokia and RIM were both well aware of this – Nokia brought in an outside CEO with a software background precisely because of this challenge. Personally I think the fact that Nokia took a non-Finn as a CEO speaks volumes about how seriously they viewed the problem.
RIM had just cottoned on at the time too. They were a bit slower to react but were busy buying up QNX and TAT to form the basis of their new platform. I’d argue that RIM’s platform was significantly less competitive than Symbian and their replacement much further off, yet they had a slower collapse than Nokia. To me this suggests that both the exogenous and endogenous factors were important. In terms of broad market share decline, exogenous was more important than endogenous. However, I believe that when it comes to profitability and the possibility of surviving through the transition the endogenous factors tipped the balance to give Nokia a close to zero chance of recovery.
To stick with the burning platforms metaphor – the platform was already on fire, there was a giant electrical storm raging around it and then Elop showed up and poured petrol all over the flames and while he was at it spread the fire to the lifeboats. His action to save the company was to call the coastguard – Nokia now has to manage the fire as best it can while it waits for the coastguard to figure out how to save them.
This seems pretty fair to me, since Nokia’s fate is almost entirely in Microsoft’s hands now. As far as I can see, Elop basically decided against trying to fix Nokia’s software development capability (which is where the bulk of the value in the industry is now) and handed over responsibility to Microsoft instead.
I don’t see RIM’s story so different as they also announced new platform (bb10) prematurely making egsiting bb7 devices somewhow obsolete.
But the long term strategy is different (build the ecosystem vs. join) but only time will tell.
Very well argued and with good empirical evidence.
It will be interesting to see the evolution of Windows 8 with Nokia. I can’t see RIM recovering and anticipate MS/Nokia getting 3rd place within a year or two.
Why? Because the Metro UI is quite nice , feels fresh and the developer tool chain works really well. Whilst RIMs stuff is okay it won’t have the breadth of offering in comparison. And why target BB10 when you could target an OS with more potential? (win 8). As much as I personally have no love for MS and am aware of their track record over the last 10 years (lack of innovation, failure to deliver, crushing management chiefs, staff woes, quashing creativity, politics over product etc.) I think they are starting to turn this around and we may be seeing the start of an exciting resurgence.
Great article. My only point is that RIM is also seriously sufferIng from the Osborne effect with their long delayed BB10 debacle. Both the edge cases and the true BB faithful are deferring purchases due to this coming savior of the company. Every day of waiting eventually turns a group of deferrers into a group of defectors.
Seems strange for a Nokian to claim that MeeGo is the platform when [s]he should know that Qt is/was the platform – and Qt (the platform) would have been (and is) on tens of millions of Symbian devices, plus whatever MeeGo devices Nokia could sell.
To compare Nokia with RIM, Symbian with BB7 and MeeGo with BB10 is utterly specious. Nokia had a platform that is loved by developers, and was about to be deployed on several times more devices than WP has managed to sell 18 months later. Another argument for not following the original Nokia software plan is that it posed a huge threat to Microsoft, who were now calling the shots at Nokia.
Great article. I am a Canadian of Finnish background so have watched both carefully and you are spot on. In some ways they remind me of Ford (Nokia) and GM (Rim). Ford did unthinkable things like selling their logo and anything to stay alive and solvent while GM just tried the hail Mary with the Volt. And we saw how that turned out.
interesting, however, first, RIM is not dead (yet)… and second, it’s not only about symbian, meego, meltemi or WP, it’s more about RnD budget and headcounts (http://www.asymco.com/2011/02/04/nokia-employs-as-many-engineers-for-symbian-and-meego-as-apple-does-for-all-its-product-lines/).
It could have been my text. Tomi Ahonen often skips the what-if-scenario? What if Nokia had stuck with Symbian/Meego OSes glued together with a QT “bubblegum”.
What would’ve Microsoft done next? Saying “oh gosh, we’ve to quit this”. Nope, they would have chosen another mobile vendor like HTC or RIM and supported them. It’s of course difficult to see what would Nokia’s position be now in that scenario, but would it be as rosy as Tomi Ahonen often describes? Don’t think so.
In my papers Nokia’s Symbian strategy was scattered when Elop was forced to announce that Symbian 3 phones WOULD get updates longer than Nokia planned and Symbian 4 wouldn’t be a separate OS but be implemented piece by piece on Symbian 3. This happened late 2010 if somebody can’t remember.
Symbian 4 was supposed to be the completely new and rebuild version of Symbian, but carriers weren’t happy to sell Symbian 3 phones to customers that would be quickly(?) passed by a newer Symbian OS version. (Quite ironic…something same going on with WP 7.5 and WP8?)
It was then clear that Nokia would have an enormous task to mix these Symbian development lines without delays. And due to this, declining profits, high RD costs and Meego platform problems Elop and the Nokia board was forced to take a close look to another OSes available.
Hello all,
Thank you for your comments and critiques, I found them to be very valuable.
@Bob: Unfortunately there is a difference between a platform being “loved by developers” and getting the apps it needs to succeed. Having a great following of geeks who love the platform does not magically make Instagram, mint.com, kindle, netflix, evernote, etc show up. Apps are a huge part of what makes phones sell now. It just doesn’t seem to sink into Nokia’s company DNA
Nokia had failed to get sufficient interest in Qt from major app developers even before the “burning platform.” Even though they were selling tons of devices
Some more points for Bob:
“To compare Nokia with RIM, Symbian with BB7 and MeeGo with BB10 is utterly specious”
is it?
“Nokia had a platform that is loved by developers”
Some developers yes. But just as many developers hate working with Qt. Also I have heard developers say wonderful things about BB10’s development environment, but that doesn’t mean they will get the big brand apps that sell phones
“and was about to be deployed on several times more devices than WP has managed to sell 18 months later.”
Again, devices sales != apps on the platform. Symbian was already (and still is!) sold on more devices than WP, and the apps didn’t come (even before Feb-2011). The apps continue to come to WP though.
“Another argument for not following the original Nokia software plan is that it posed a huge threat to Microsoft, who were now calling the shots at Nokia.”
This is more conspiracy theory paranoia or warmed-over MS hatred from the 90s than actual argument. It is a sign of some people’s Nokia-centric arrogance if they thought Microsoft considered Nokia a theat. And if they DID consider it a threat, why did they agree to put MS Office on Symbian in 2009 and help strengthen this “threat”?
@ xizzhu:
“For the app ecosystem, how many major apps done by companies did Symbian miss at Feb.11?”
Ummm, most of them? I remember Symbian having Angry Birds and Shazam. That’s about it.
For MeeGo, again, what if Android virtual machine were “ported”?
Look what happened with IBM’s OS/2 when they made it able to run Windows 3.1 apps. All the companies stopped writing native OS/2 apps, focusing on a single Windows version that would run on OS/2 AND Windows PCs. And of course OS/2 never ran Windows apps quite the same as real Windows…and finally, that left OS/2 completely dependent on continued Windows compatibility, and screwed them when Windows 95 came out and MS no longer had compatibility with OS/2.
So to answer…it would not be very good.
Just a minor point – the “important” apps target platforms they believe to be commercially viable. In most cases that doesn’t include either Symbian/Qt or Windows Phone. Some of those apps can be persuaded to port with sufficient cash. Some won’t port even for significantly more than the financial cost of doing so because the opportunity cost of wasting their management time is much greater, or they simply don’t wish to associate their brands with an uncool platform. The combined Nokia/MS pot of cash for encouraging apps to port was about an order of magnitude greater for WP7 apps than Nokia’s alone for Qt apps (according to my sources anyway). That’s the only important difference between platforms with insignificant market share at the high end – they’ve all been doing it, just some on a much bigger scale than others. Regardless of what developers like/dislike, economics is the key driver.