I hear an awful lot online, whether it's in the trade press or in the blogosphere, about 'software project failure'. First let's be honest and say that a lot of people use project failure as selling tactic. Failure is painful and software project failure is not exception. Even if you haven't been through it before it sounds bad. So it makes for a good selling point. But if you do a Google search you'll find tens of thousands of hits. Ditto Technorati. Even if half are for the purpose of selling that still leaves a lot, and too many. So what's it all about? After many years of doing software implementation I have noticed some trends that I can share. First, projects that start poorly end poorly. There ought to be a real sense of urgency in the very beginning and part of that must be the requirement to preserve time, effort and cash. Get off to a quick start and try to set some early goals that you can achieve in the first couple of weeks. Get the team used to delivering on time and making noticeable progress. But preserving time, effort and cash pays the greatest dividend when you head into crunch time and there's a go-live data and the last of the testing right around the corner. This is not the time to run out of time, effort and cash. In fact, I would venture that better than half of software project failures happen after go-live because the team is quickly disassembled and no project, software or otherwise, can maintain itself. Some people dream of the day when software systems can diagnose and treat themselves but we're not there yet, not by a stretch. The last point to make is 'Iteration'. This is one of the most important points in any software project. You should start to setup and test the system as soon as possible, even if that means testing and working in an incomplete system. Some of my former consulting colleagues thought that I was nuts to put the system in front of the end users so early. They preferred to work in secret and then, just before the cash ran out, they would reveal the actual system and before the screaming had subsided, they were out the door and onto a different engagement. My assumption is that we are all adults here, and most people understand that systems require many iterations before you get it all right. I have never had a problem with this approach and it has turned out excellent results over time and many engagements. Now NetSuite implementations are shorter and more to the point than the Enterprise ERP and CRM systems implementation, that's for sure. But the same rules apply. Look to deliver some results in the first 5 business days. Once you get the big rock rolling you'll be surprised at how it can gather some speed. In the end, crunch time, you'll have time to train and make sure that the users are ready to take ownership of the new system. Software project success can sometimes seem like an anomaly, if you believe all the hype. But it's within the grasp of those who want to work hard upfront and sock away some time, effort and cash for when they really need it. | 
