Again, as I think was talked about earlier on, the original decision point of cutting back on the functionality of the system really caused a lot of the problems, and things needed to be resolved at that point. It all started to build from there.
If the project had been managed the right way all along, you would have hoped that would have been identified much earlier, rather than to say at the last minute, “well I guess we'd better not launch this”. It would have been better to find a way to continue on with the old system at that point in time, but even that would have been an 11th-hour, 59th-minute decision.
Really, to have avoided this, better decisions needed to have been made all the way along.