Month: September 2011

Google App Engine is dead, they just don’t know it yet.

That’s what I conclude after reading the tea leaves (er, blog posts).

App Engine made a radical change to the way they charge, resulting in much higher costs for typical sites. The reaction from programmers is predictable: anger. Anger at the new prices. Anger that they devoted so much time and energy to this platform. Anger that they must now spend so much time and energy porting to a LAMP stack, where they could have been all along.

Reaction from supporters, almost all of whom work for Google, is more interesting. They are are recasting App EngineĀ as an enterprise solution. If you need unlimited scalability on demand, App Engine is for you! Indeed, if I were working on a site for the Super Bowl, royal wedding, or tax day, I might give App Engine a look. Otherwise… not so much.

At this point, App Engine solves a problem that most sites don’t have: the ability to keep your site up in the face of sudden, unanticipated traffic. If your traffic increases gradually, or predictably, then there are cheaper ways to handle the load.

Supporters also argue that Google is “putting more wood behind fewer arrows” these days. In other words, management is no longer in the mood to subsidize anything. If a project doesn’t pay its’ own way, it gets cut. The price change enables App Engine to pay its’ own way, presumably.

But maybe not. App Engine surely has fixed costs. If Google fails to attract enough customers at this new pricing level, they will still lose money. Even if they eke out a profit, they might still get cut if they are not a huge revenue generator. For developers (er, enterprises), it will be extremely risky to bet on App Engine until several years of undisputed revenue success are in the books.

What do you call it when current customers are leaving and new customers are not replacing them? A death spiral.

I guess I can’t blame the folks behind App Engine for trying to save their baby. But what I think will happen is this: After all the typical web developers have fled, leaving behind a handful of enterprise customers, Google will shutdown App Engine with little fanfare. Because no one will care anymore.