Most product organisations measure activity because activity is legible. They rarely measure judgement, because judgement is not.
The consequence is a scoreboard that rewards the work the company is already doing well and quietly ignores the work that would actually change the trajectory.
The two categories of work
**Legible work** is work that fits inside a dashboard. Feature ships, tickets closed, releases cut, experiments run. All of this is real work. None of it is a proxy for whether the organisation is building the right thing.
**Consequential work** is the work that changes the shape of the next decision. It shows up as fewer meetings about the same topic, faster time-to-clarity on new bets, stakeholders who stop asking a question they used to have to ask every quarter. It is nearly impossible to put on a slide.
A metrics stack that only measures legible work will, eventually, produce an organisation that only does legible work.
Three ways to instrument consequential work
**Decisions changed per cycle.** At the end of each quarter, name three decisions the organisation made differently than it would have last quarter, and the evidence that changed them. If the answer is zero, the organisation is running the same operating system it ran a year ago.
**Time-to-clarity on new questions.** How long does it take, from the moment a new question enters the room, for the organisation to have a defensible answer? This number should shrink as the operating system improves. If it does not, the operating system is not improving.
**Postmortem density.** Not the count of postmortems. The percentage of postmortems that produced an operating change that persisted for at least two quarters. That number is the honest measure of whether the organisation is learning.
What to do with this
Do not replace the legible dashboard. Add the consequential one alongside it, and read them together. The gap between them is where the actual work is.