When you take over a multi-million pound project just a few weeks before the first module is due to enter development, you need to be certain what the project is required to do.

So, the last thing you want to hear is that the team responsible for ensuring the application will do what the users want have not been entirely successful in answering this question.

Not only is this disturbing, prompting a "can I go back to my previous project" moment, but also rather perplexing. The project team, including various business analysts and project managers, did in general have a significant amount of business knowledge amongst them. And yet in taking over this project, the feeling I got, and this was shared among both my team and the business at large was that they had made a complete pig’s ear of understanding the application requirements.

The reason was actually quite simple. At the start of the project an outside consultant was drafted in to run it. This is, after all, a large, high-profile project and needed someone with experience to look after it. But sometimes all the experience in the world is no substitute for looking at what is right in front of you. This consultant was so sure that his way was right that he failed to make any use of the knowledge within his team. He forced his methods and ideas on a team without considering the culture and the capabilities of the people involved. He also constantly changed the focus of the application, trying to bring new functions and business areas into the scope, which, in turn, created constant change and confusion in the project. Neither is good when you have a very fixed deadline.

Fortunately for the business, they realised this before too long. Unfortunately for me, they got me to agree to transfer across before I realised just what I was getting into. But the team was still there, and their many years of business knowledge was still in place. I could see that while things were far from ideal, it was also far from unsalvageable. In my next blog I’ll explain how I managed to rescue the project and get the first couple of modules into development without losing too much time.