notes-groupDecisionMaking-metrics

should you introduce metrics to quantify progress towards a goal even when you can't think of a way to measure exactly what you want to optimize (only things which seem to be closely related)?

two contradictory heuristics:

it's better to measure/quantify something, even if it's wrong

vs

if you reward people for optimizing a metric, then people will optimize that metric, not what you want

responses:

• choose to measure
• choose to not measure
• these are not actually contradictory; you can measure something without tying the reward to it. if you reward people based on a flawed metric, make sure that's not the sole determinant of their reward.
• we can handle it: just be clear that the measurement is only part of the picture.