I had worked in international HQ of IBM since 1970, mainly SW development management and marketing . Saw three revolutions in HW , a lot of challengers disappear and even IBM in danger . If your project does n't meet breakeven point in less than three years , its assesments must be completely reviewed. In large organisations, to kill in due time a project under development is sometimes the best management decision to make , whatever the costs involved. An IT corporation must maintain their teams in a permanent technological intelligence with R&D and simultaneously have a clear LT vision . Very few companies can afford both . IT industry cemeteries are full of defunct businesses , once bright but which missed to foresee a brutal change on the road.