It is bad if you don't properly measure the impact of the features.

There are some features that seem like obvious wins on paper but actually do worse when you A/B test it.

They hurt business value and add needless complexity to the development process.