Preparing for the inevitable

If you deliver an application over the web, downtime is inevitable. Much as we try to prevent downtime — through redundancy, bug fixing, monitoring — it still happens. The recent — okay, a couple of weeks ago, this isn’t a real-time blog — fiasco at 365 Main shows that even the "world’s finest datacenter" can still have its problems.

It’s been interesting to see how companies with web apps are using Twitter (and Jaiku, etc.) to provide status reports when their sites crash.

And talking about disasters in the physical world, people are starting to hear about them first on Twitter, etc. Here’s a story about how the Los Angeles Fire Department uses Twitter. Here’s one about the Minneapolis bridge collapse and Twitter.

People can debate Twitter’s usefulness for everyday events. But no one can deny that Twitter’s becoming the killer app for disaster management.

Which brings me to my main point: how long will it take companies that drive revenue through web apps to setup alternate communications channels using Twitter and other microblogging services?

For example: suppose customers can’t place an order via your website. The sooner you can tell customers that you’re aware of the problem and trying to get back online better. The sooner you can post an estimate of when you’ll be back online, the better. The sooner you can inform that you are back online, the better.

I’m no expert on disaster management, but it seems like one of the things that separates good disaster handling from bad is the quantity and timeliness of communications.

It seems natural that companies will eventually post their Twitter URL on their "Contact Us" page. And perhaps a Jaiku URL, too, in case Twitter goes down! This seems to be one of those standard requirements that product managers should prescribe for any Internet web app.

If companies adopt Twitter for disaster management, this essentially makes it a mission-critical application. Could Twitter charge companies for this service? Perhaps. Every company wants to look as competent as possible during a service outage — fumbling a disaster could cost a CEO or CIO their job — and Twitter would help here.

It will be interesting to see whether Twitter starts to offer something to address this market need.

Note I’m talking about the possibility of Twitter charging companies, not individual users. I can’t see any reason for Twitter to charge individual users. What I’m talking about is Twitter possibly going with a freemium pricing model.