Much has been said and read about Cloud Computing in the last few months, I think my over reaching comment on the whole scene is that it’s the usual IT Jedi mind trick  – one solution/one vendor/both can solve all your IT problems. I believe it pays to retain a degree of cynicism around anything IT people say, as they usually mould a technology and then wonder how best to fit it around you, rather than the other way around.

I think there’s also the automatic reaction from CxOs that shifting IT costs under a single “outsourced” line on the profit and loss sheet makes sound fiscal sense. I don’t believe that’s true for a second. There is room for Cloud based services with no residual business value (a University’s student e-mail is a fine example, or maybe a corporate blog), but in the case of commercial data and intellectual property, I’m no more likely to park my car on someone else’s driveway than I am dropping gigabytes of data onto someone else’s data warehouse a couple of thousand miles away.

The single point I want to make in this posting is that shifting your core IT services into someone else’s domain is fraught with problems. Recently there have been high profile outages with both Amazon and Microsoft BPOS, leaving customers with a degraded (or non existent) service for hours, days and weeks. Let’s be clear about one thing – when IT services are kept “on prem” (yuck) or in house, outages happen all the time. Even with SLAs, SLSes and the like. Shit breaks, it’s a fact of life.

Now then, go back and re-read those articles I posted above. What do they both have in common? The fact that customers were not kept informed about what the issue was, when it would be fixed and what they would do post mortem to mitigate similar failures in future. I believe this is the one area where in house IT triumphs. In a previous life of running centralised, high profile IT services, I always tried to communicate what was happening during failures and how long we thought it would take to fix it. I didn’t always get it right, but I made a big effort and I’d like to think people appreciated it.

With in house services, there’s always an arse to kick. You always know specifically whose arse to kick and the buck didn’t stop with a faceless multi national company with a call centre out the back of beyond. You can call a mobile number and speak to the technical person who’s standing over a smoking server. You can offer your support (however limited that might be). You can pat them on the back when they recover the service against all odds (server rooms with 3 inches of water on the floor, power cables hot to the touch) and you can sit them down to talk over how to prevent happening again.

Much is said about communication. Less is said about the lack of it. I think the inherent strength of in house IT is the ability to directly communicate with the people who build, maintain and in some cases, love the services they provide. As soon as you sign the contract to put any of that infrastructure out of the door, it’s an automatic slap across the face for your IT people. As a result, they’re less likely to go that extra mile to love that cloud based service, not least because getting to it is obfuscated by layers of bureaucracy at the vendor end.

Yes, cloud computing has it’s advantages. We’ve been using the cloud before we started calling it that. Hotmail, GMail, Dropbox, X Drive. etc. There’s nothing new here, someone just invented a fancy name for it. Remember though that once you’ve opened the door to cloud based core services, you’re making a fundamental shift in your philosophy. Just make sure you’re fully informed before you do that. In house IT may cost that bit more, but some things are worth more than just a few hundred quid.