Industry analysts have been pointing out for years that roughly 60 – 70% of all software development/implementations fail for at least one of three major success criteria: on-time, within budget or correct functionality.The reasons for this are wellknown. The most common ones are scope creep or feature creep that is not properly managed ? especially the management of the expectations on the part of the stake holders. The impact of good changes on delivery dates, budget or over-all functionality has to be made aware to stake holders on a regular basis. Knowing when a change is good or bad requires skill in making sure there is a good sound business driver for changes and not just a ?good idea? that will cost money with no real pay back. Another obvious but regularly overlooked problem is that ?customizations? you are paying for are built-in features in competitive products or worse yet – are actually new features in a future release of the package or tools that are being used today and would be available anyway. Ongoing ?build versus buy? comparisons and close software package/tool vendor relationships are necessary to make sure that limited resources are not wasted in a duplicated effort. Regularly syncing up software vendor road maps, with internal implementation road maps is key. Also important is that you need to know who to talk to at a vendor to validate claims that a feature really exists or is coming soon. It is the responsibility of the CIO/CTO/VP/Director to use their experience and contacts to find out the truth. Finally, the most frequent problem in software implementations and development these days seems to be the projects that are missing pilots and prototypes. Of the 30 ? 40% of software package implementations or custom development efforts that succeeds every one of them has gone through at least one and sometimes several prototypes and/or pilots.