Thursday 20 January 2011

When 'buy before build' goes bad

It's pretty much a universal principle in a large enterprise: If you want something new, first you try to reuse, then you try and buy something, and only as a last resort do you build something yourself.

Sound enough, right? It ought to be cheaper to buy something from a specialist vendor than work it out and build it from scratch yourself. They get the economies of scale, you get a reduced price, everyone's a winner.

The only trouble is, sometimes buy before build sucks.

The trouble is that usually when you buy, you're making an often large up front financial commitment to something. Not only that, but often we buy something before we've had a chance to really work out what it is we need. So, we end up buying the uber product - something that delivers our every whim and desire.

Very often, when it comes down to it, we buy 100%, use 20%, and wind up bespoking the living daylights out of the rest. The Vincent van Gogh of our vision becomes more like an HR Geiger Alien. It costs as much to customise as it would have cost to build, and far from being a virtual utopia becomes the treacle holding us back.

So, how do you make sure this doesn't happen to you? Simple: Only buy what you're absolutely sure you need.

How do you know what you really need? Simple: Build it and see how your users use it, rework, repeat.

Sometimes life has its little ironies...