Great job, GE! And kudos to the team who are making it happen... When we look across the industry, however, we see more ERP failures than we see successes. Consider, for example, the multi-year CSC / Oracle ECSS project for the Air Force, which has been declared dead only recently: over $1 billion spent, "negligible value received". In that case, and in many others, the better solution would have been to modernize the old applications, add SOA for integration, remove redundancy, and become more agile with a code base uniquely adapted to the requirements. The last thing that should be done in these situations is the traditional, spend-3-years-writing-requirements-and-blow-the-budget waterfall approach.

So why is the GE experience different? It has to come down to the people, to the sociology, to getting committed people in one room and making a difference. And that is why this lesson is going to be hard to duplicate and use as textbook material -- because there will be different people with different motivations in every case. Hats off to Mr. Kocsi, too, he's obviously a leader and in the end, that's the one thing that is always in short supply: leadership.

Now the question is, once the ERP is in place, how does information processing STAY agile? Because customizing the ERP going forward is a limited option...