> the McKinsey framework only measures effort or output, not outcomes and impact

Measuring outcome and impact gets messy when the developers aren’t the ones deciding what to work on.

I had a VP of Engineering who was all about measuring impact and ignoring everything else. He was also adamant that product people decided what we work on, while engineers decided how it was implemented.

This created an environment where your fate was determined almost entirely by what product team you got assigned to. If your product managers wanted you to implement a feature nobody wanted, it didn’t matter how well you did. You weren’t going to be able to show positive impact if nobody used it.

Impact is only a good measure when you let people make their own decisions.