Android was never a product, it was an insurance policy

Google bought a small company called Android Inc. four years ago and almost nobody outside the deal could explain why. A mobile operating system, from a search company, with no handset business and no carrier relationships. It read as an acqui-hire with an odd shape.

Four years on there are a handful of phones running it, the T-Mobile myTouch has just arrived, HTC has put its own Sense skin over the top on the Hero, and the second real release of the platform landed in the spring. That is a slow burn by any standard, and I think the pace is the most interesting thing about it.

Nobody at Google has ever been able to give a straight answer about the business model, which usually means there is one and it is embarrassing to say out loud. Here it is not complicated. Google does not want to sell phones or software. It wants to make sure that whatever people carry in 2015 has Google search on it by default, and it has watched what happened to anyone who let a platform owner sit between them and their users. Buying Android was insurance, bought early and cheaply, against a future where Apple or Microsoft or a carrier decides what the default search engine is.

Giving it away free to handset makers is the same insurance policy continued. It is not generosity and it is not really open source idealism either, whatever the licence says. It is a company with an enormous existing business spending money to prevent a bad outcome rather than to create a good one, and those projects tend to be funded patiently and judged loosely, which is exactly what we are seeing.

The software is still rough. The Market is thin, the design is inconsistent between one handset and the next, and HTC skinning it within months of launch tells you the manufacturers already regard the stock experience as a starting point rather than a product. Apple will use every one of those things as an argument and will be right about most of them.

None of that touches the structural point. Every manufacturer without an operating system of its own is looking at a free, competent, modern platform they do not have to build, at the exact moment their existing platform is starting to look ancient next to what Apple ships. Nokia has Symbian and a problem. Motorola has nothing and a much worse problem.

I have no idea whether Android is good enough yet. I am fairly confident it does not have to be, because the alternative on offer to most of this industry is nothing at all.