Having the Time to Fail

Another report is out, this one stating 62% of IT projects failed .

I've written before on failed IT projects in More enterprise projects gone expensive

A fresh perspective of mine is that maybe a lot of failing IT projects isn't bad, but a natural thing. Perhaps the problem is rather that they do not fail faster. Remember MySQL CEO Mårten Micklos statement fail fast, scale fast . Being entrepreneurial means quickly trying out new ideas. Technology and methodology should allow us to try new ideas quickly.

It is a matter of perspective. Let's take mp3 players as an example. Most projects to develop and produce an mp3 player is successful because an mp3 player eventually gets out of the factory. But on another level, only a few mp3 players (like iRiver and iPod) are really successful in the sense that their design and functionality generates approval and sales from the big public.

So what is your perspective? Do you compare IT projects with producing mp3 players (62% of all produced players don't work) or designing an mp3 player (62% of all mp3 player models are unsuccessful). It's the difference from designing an IT system and just producing one according to some blueprint.

How do we know if an mp3 player will be a hit? Perhaps we don't (if we're not Steve jobs). We give it a try. If we just can fail fast and cheaply enough, we can design three more from what we learned and eventually get a hit. I remember when I stared out as programmer and for the first time was (forced) to wrote a VisualBasic. I quickly realized that I had the time to try out three designs in the same amount of time as one in C or C++. It didn't grow my C skills so quickly but I got a lot better on designing systems. I had the time to try and quickly fail. I learned and ended up with improved designs. And so did my employers.