Cognitive Dissonance in Product Development: Falling in love with the struggle

The short lesson here is: don’t keep in features just because they were hard to build.  Good and bad are not the same as hard and easy, and even if worthwhile things were always hard (which they aren’t), that doesn’t mean that all hard things are worthwhile.

The longer explanation starts with the documentary a friend shot.  When I was watching the rough cut, I was marveling at how varied the footage was: some extremely high quality and quite moving, others less so.  My favorite was a city council meeting with the world’s worst audio; periodically, a door chime would come on while people were talking that made it sound like it was filmed in a 7-11.

Yet it made it into the final cut.  And as I was talking to my friend, I began to understand why.  Getting into the city counsel meeting, and filming it, was a tremendous endeavor that involved plenty of red tape and logistical challenges.  And because it was so hard, she was experiencing cognitive dissonance: it must be valuable because why else would I have put so much work into getting it?

I’ve seen people do this product development meetings over and over again.  Either a feature was hard to imagine or hard to execute on, and so when all those resources had been put it, it was immediately imbued with tons of value.  Acquisitions are a classic case: we paid so much for this, it must be really strategic and core.

Mistakes happen.  Sometimes we built parts of our product that just don’t belong, either because of poor fit or poor execution or just plain bad roots.  By leaving them in, we compound the mistake: it is like watching idiots in survivor movies.  You want to tell them to hurry up and build a shelter already, or get some food, but instead they just keep spending a bunch of time building showers.

Don’t let your mind win.  If a feature is bad, cut it.  This also goes for team members, marketing campaigns, and bizdev deals.  The biggest mistake you can make is holding on to bad ideas.