Choosing what to measure, reading the result honestly, and forecasting without pretending to certainty.
Measurement goes wrong in two directions, being metrics that track activity rather than value, and results read with more confidence than the data supports. These cover both, from picking the metric through to the statistics of running experiments at scale.
Which metric matters at which stage, broken down by business model, with real benchmark numbers rather than principles alone.
The shortest statement of why measuring what you shipped tells you nothing about whether it worked.
The reference on A/B testing done properly, including the traps that make a result look better than it is.
Forecasting from cycle time rather than estimates, and the charts that make flow visible. The most practical delivery metrics book.
How goals and measures connect, and the difference between a committed and an aspirational target.
Four delivery metrics with the research behind them published in full, which is rare enough to be worth the read on its own.
Having measured it, this is how you present it so the decision actually gets made.
The background on why you will over read a small sample and under adjust a forecast.
The operating detail of running OKRs, covering the cadence, the scoring and the common ways a rollout stalls. Read it after you have decided the approach and need to actually run it.