Operational innovation and the Google datacenter

There was a good article in the April Harvard Business Review about operational innovation (pay for a copy or read some highlights) which mentions some of the more well-known examples, like the Toyota Production System, the Dell business model, and Wal-Mart's cross-docking.

One system that didn't get mentioned, but should, is the Google datacenter. As Rich Skrenta reports, the Google datacenter is

a distributed computing platform that can manage web-scale datasets on 100,000 node server clusters. It includes a petabyte, distributed, fault tolerant filesystem, distributed RPC code, probably network shared memory and process migration. And a datacenter management system which lets a handful of ops engineers effectively run 100,000 servers.

This set up runs on ultra-cheap hardware and lets Google offer things like super-fast web search, Froogle and the 1 gig, fully searchable Gmail inbox. As a counter example, let me tell you about a large client moving their email services to a new corporate datacenter. In the "new world" (you'll appreciate the irony in a minute) staff will be allocated 20 megabytes for their inbox. The email servers will generate warnings as they approach their cap; as soon as they're at the max their computer will lock up until they've deleted messages and moved their attachments to a shared drive. The shared drives are searchable with the native Windows search (which blows, frankly) but there are no indexing services. And it's expensive. And it's bad for users who tend to lose things on shared drives.

If I were an IT manager I'd want to know if I could move my file/email services to Google, with Gmail as the email client. One of the great things about Gmail is it lets you use your inbox as a general purpose information and file management tool. Email is rich with native metadata--Google's labels, stars and threaded conversations make it that much better. (If you've been following the folksonomy/social classification meme, you've probably already thought of sharing labels with other Gmail users).

Getting back to the original topic, operaional innovation is a source of sustained competitive advantage. While I think Google could make a ton of money offering enterprise data services, and Jason thinks they're building a web OS, the point is that they've built a platform to do just about anything they want.