10things#2 - Something will usually go wrong

It doesn't matter how hard you try, how much planning you do, how skilled you think you are, without fail something will go wrong when you're building a website. You'll miss some critical IE6 (spit) bug that only the boss will notice, you'll forget to add in the Analytics code until a week after the first traffic peak, some 12-year old script kiddie will inject something malicious into something you thought was secure. If it doesn't happen, or hasn't yet, enjoy it while it lasts, because when you get that just-kicked-in-the-balls (or ovaries) feeling, and you realise the backup didn't run as scheduled, I'll feel the pain with you.

What happens next is important.

You can either bury your head in the sand, sick your fingers in your ears, and/or go cry into a beer/McDonalds, or you suck it up, hunt down the problem, fix it, and then be transparent with everyone about why/how it happened. People are much more likely to be sympathetic if the problem was out of your hands, or understanding if you take responsibility for the mistake, than if you lie your arse off about it, and whine like a 9 year old. Campaign Monitor were recently attacked, their systems infiltrated for a short time and users accounts were affected. The outpouring of support and well-wishers on their blog in the days afterwards were a perfect case of why you should hold your hands up and be honest with people when something goes wrong.

That said, sometimes it's OK, if you fucked something up and no-one has noticed, to fix it and keep shtum.

ShortURL: http://project14.co.uk/b/3