We’ve all been there – all the elements of the feature you had hoped for will be there in time for the release cut-off. There are really two choices – hold the feature until it’s done or carve back scope and release what is ready. In general, I tend to lean towards the latter strategy, but it only makes sense if you can release functionality that can stand on it’s own. Some features cannot be broken up into meaningful pieces. The key is to look at things from the user perspective and ask the following questions:

• Does this deliver value to the user as is?
• Does this feel complete even though [x] is missing?
• Is what we have in a good enough state for release?

If you can answer yes truthfully to all the above, releasing the feature, even though it’s incomplete is a net positive.