In 2007, the company now called "AT&T" is a shell. It's a holding company, an amalgam of shareholder interests that houses zero innovation. Zero. AT&T proves the theory that splitting the original, pre-1984 monopolistic AT&T actually stifled innovation. Oh, it might have helped competition. But did it help innovation?
Today's Wall Street Journal discussed the hiccups involved with AT&T's TV offerings. Turns out that AT&T's attempt to enter the IPTV market is dependent on technology of.... ta da! Microsoft! It's called Project Lightspeed.
AT&T, the $230B company that once built everything from their switches to their handsets, now is paying Microsoft to supply their TV offering. Now, I don't really understand this. I am a Microsoft fanboi, I suppose, but I'm also of the belief that if something is really important to your business, you want to control your destiny. You really wouldn't want to turn over technology involved with your core business to someone who might actually be a competitor in the same space.
On the other end of the spectrum, we have Google. Google will acquire companies that use incompatible platforms from what they use. Then they will spend the time to port it to the platform that Google uses. Yahoo is similar, where they use their standard BSD platform that they've been using since, like, forever. Neither of these companies can risk being on Microsoft's platform for the long term. After all, Microsoft is trying to invade their space as it is (I just saw an ad for "Microsoft AdCenter" as I wrote this paragraph).
So I've been thinking about a razor for when it is and when it is not acceptable to develop your own tools. It's not completely based on cost.
The razor is this: if the tool is integral to the core business of your company--or even just one product--for the long term, you should develop it internally almost no matter what the cost. There's a very good chance that you're going to end up spending that kind of money with vendor lock-in. There are two caveats though.
a) You are developing for that platform. For example, a Windows developer really can't avoid buying Windows and Visual Studio and being tied to those.
b) The training or outsourcing costs and risks outweigh the risk reduction and cost savings you would get from developing the software.
Back to AT&T for a second. They haven't used this razor to evaluate what to do for their IPTV solution. As everything these days, they instead wanted to rush their product to market probably to appease shareholders. And they made a deal with the company that can ultimately bury the product they're trying to make. Huge, huge mistake by a company that used to develop and own everything themselves and control their destiny.