Hacker News new | comments | show | ask | jobs | submit login

I worked at a place where we tried practicing the principle of the 30% demo. We had one additional point -- at 30%, you should have rough "end-to-end" progress.

This made a drastic difference in how early we were able to assess a design direction. Previously we would have only made 30% progress but that 30% would have been polished. Making it an "end-to-end" demo allowed us to get that kind of rough feedback on the whole feature. And if we needed to, tuning our direction at that point wasn't that painful.

The result was that we no longer had to throw away work that was polished but didn't deliver the goals we wanted -- or worse, continue to force the path because of sunk cost bias, ignoring the feelings that something wasn't right.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact