Symbian still runs on more smartphones than anything else on earth and I no longer think that number tells you who is winning.
Apple has now announced the second phone, on 3G, at a price that is roughly half what the first one cost, with a shop attached that sells software. Each of those three changes addresses something I complained about at launch. The radio was the clearest failure of the original device and it is fixed. The price puts it inside the range where a carrier subsidy makes it an ordinary purchase rather than a statement. And the store answers the question I said would decide the whole thing.
What matters about that store is not that applications exist. Symbian has had third-party software for the better part of a decade. What matters is that a developer can now reach a customer without asking a carrier’s permission, and get paid without building a payment system. The thing I said two years ago was the structural problem in this industry, the one no handset maker had an incentive to fix because the carrier subsidised the handset, has been fixed by a company that arrived with no legacy relationship to protect.
So the market share number is measuring the wrong thing. Symbian’s installed base is enormous and consists largely of devices whose owners have never installed anything and never will, because the process is unpleasant enough to deter anybody who is not determined. Counting those handsets as smartphones in the same category is like counting every car with a radio as a music platform.
The nervous position is Nokia’s, and I do not think Nokia is behaving like a company that knows it. The strategy remains convergence: put everything into the device, win the checklist, ship in volume at every price point. That has been a winning strategy for a decade and it has produced the most capable handsets anybody makes. It has not produced anything a developer wants to build for.
Google’s alliance is the other answer and it remains theoretical. An operating system given away to manufacturers, backed by carriers, with a store that anybody can publish to. If it ships and it works, it attacks Apple from underneath on price and attacks Nokia from above on software, and the fact that nobody has held working hardware yet is the only reason it is not the main story.
Microsoft is the one I would write off. Windows Mobile has had every advantage for years: a licensing model that put it on hardware from multiple manufacturers, an enterprise foothold, developer tools people already knew. It has spent that lead maintaining an interface designed for a stylus, and the best work being done on it is being done by a hardware company covering it up.
My read of where this lands: the platform question gets decided in the next two years and it gets decided by developers, and developers go where the customers and the payment mechanism are. Everything else on the specification sheet is going to matter much less than anybody currently selling phones would like.