Use this when: You have metrics and dashboards but nobody looks at them regularly. The data exists. The habit doesn't.
You're done when: The team has run three consecutive reviews without someone asking "why are we doing this?" and at least one decision has changed because of what came up in the meeting.
The Sequence
Template
The Three Anti-Patterns That Kill Metrics Reviews
Every team that tries a regular metrics review eventually falls into one of three traps. Knowing them in advance is the only way to avoid them.
The "Stare at Dashboard" meeting. Everyone opens the same dashboard. Someone shares their screen. The group reads numbers in silence. Nobody prepared an explanation for why anything changed. The meeting ends with "looks fine" and everyone goes back to what they were doing. This happens when there's no prep owner and no narrative. Numbers without context are decoration, not information.
The Blame meeting. A metric dipped. The meeting becomes a courtroom: whose feature caused it? Who pushed the bad deploy? Who approved the campaign that didn't convert? Once this happens twice, people start gaming the metrics or avoiding the meeting. The review becomes something people dread instead of something they use. The fix is simple: focus on "what happened and what should we do" instead of "who caused this."
The "Everything is Fine" meeting. Every metric is presented without comparison periods, without targets, and without anomaly flags. Everything looks normal because nobody did the work to check if it actually is. The prep owner just pulled the current numbers without comparing them to last week, last month, or the target. This one is structural: the template needs to force comparison. Current value alone is meaningless.
Example
A 30-person SaaS company started weekly metrics reviews after their Series A. The first three meetings were disasters: the CEO asked questions nobody could answer, the engineering lead felt attacked when conversion dipped, and the meeting ran 90 minutes because everyone had opinions about every number.
They restructured. One prep owner (the analytics engineer) populated the template every Monday morning. The meeting shrank to 30 minutes. Only five metrics: new trials, activation rate, weekly active users, expansion revenue, and churn. Each had last week's number, this week's number, and a one-sentence "why" prepared in advance. If nobody could explain a change, it went on a list for investigation, not debate.
The breakthrough came in week 6. The prep owner flagged that activation rate had dropped 8% over three weeks, a slow decline that nobody noticed because each weekly dip was small. They traced it to a pricing page change that confused new users about which plan included the key feature. Rolled it back. Activation recovered in 10 days. Without the cadence, that 8% decline would have compounded for months before anyone connected it to the pricing page.
Written with ❤️ by a human (still)