The other day at lunch I popped some leftovers into our toaster oven and set it to 10 minutes. Well the timer started ticking away, so I left to take care of something. When I came back the timer was just dinging but I noticed that the oven was cold. Seems that I had set the timer for the oven but the machine was set to toast, and I had not turned on the toaster. Ok, start over. Another time I set the time for the toaster and set the toaster switch to Dark. The ticking started. Again I came back later to find a cold, lonely, untoasted lunch. The machine was not plugged in! Still another time I had my lunch ready to go, set the oven to 400 and accidentally turned the toaster to dark. The machine will not work when both the oven and the toaster are turned on, evidently. Same result. Ok, all of these examples are my fault. But it still makes me laugh when I consider the time I spent to find a new toaster oven for the office. I looked at 25 models over the course of 3 weeks. Not because I am that interested in toaster ovens, but because I have difficulty when I have to purchase something I know nothing about. My requirements, I thought, were very simple. It should be big enough to heat a lunch, but small enough to make toast. Safety and an attractive look are also worth something. What I never really considered, and would not even think of considering, was using the toaster oven and getting good results. It never occurred to me that an electric toaster oven would have a manual timer. This has been the brunt of my trouble. It also never occurred to me that an electric toaster oven would have 2 timers, one manual and one electric. So it follows that I have had to learn the hard way that there are two separate processes for toasting and baking, and never the twain shall meet. Frankly I never thought about the processes in nearly the level of detail that I now understand was necessary. For small and medium enterprises, SMEs, custom software development holds the same challenges. You will not know what you missed in the requirements until you have a finished product and then it's too late. And unless you are completely different from every other SME I know, you are not spending a full 50 hours a week on trying to figure out the detailed requirements. Are you going to meet the most pressing issues of your business today in the new software? The only way to do it is to work hard at it through a number of iterations. This should not be an outsourced task. The people who know your business best are the ones who have to really roll up their sleeves and get into it. The other big question is does the software meet the challenges of your business tomorrow? This is much more difficult to answer, obviously. We just don't know what the challenges are going to be tomorrow. I have had several custom developers over the years answer this question by saying that they will add a few undefined fields to the new system's records. The idea I guess is that as requirements evolve some additional fields will meet the new challenge. Even people who are not versed in systems know that this is a 'useless gesture.' The future will not be so kind. Bottom line is that development will continue into the future - far into the future. Either that or you will shape your business to meet every new challenge within the constraints of the custom built system. These are not good choices. The alternative? Find a system from a successful and growing vendor that enables you to scale and add new functions as you do. You may not get 100% of the bells and whistles that your employees want, but you'll be able to meet today's challenges, and tomorrow's, with the same software. It will also not cost you an arm and a leg in ongoing development and maintenance.The time that you will spend on custom software is better spent on buying and selling profitably, both now and in the years to come. Get out of the custom software business; almost everyone else has when you think about it. What was the last company that made it big in custom software development? IBM is probably the last of the old dinosaurs, and the only reason that they are still around is that they have moved in several other directions. Custom development is not even listed prominently on their website. The current IT landscape features a lot of small custom software firms, but they never seem to grow. Back in the day when custom software was all the rage, because it was the only option, firms grew large quickly and many were publicly traded. Those days are over. Buying packaged software is an SME's best bet. It gives you the flexibility to keep challenging the markeplace while offering a robust set of functions and ongoing R&D. Your company and all of the others who subscribe have the opportunity to move the software to adopt new ideas. Think about how costly and difficult it was for a company that developed a custom software solution in 1990 to then find out that they had to add Web Front End in 1999, and remedy the Y2K problem at the same time. I've seen my share of custom software over the years, both as a user and as a builder, so excuse my lack of enthusiasm. I guess that there are those dire situations that still require a custom solution, but I would think long and hard before I took that plunge. | 
