Vanity metric

A number that looks impressive but has no clear decision behind it, like total downloads with no activation, retention, or revenue context.

a big number that changes nothinglooks good in the investor deckstats that make us feel successfuldownloads with no idea who came backpageviews for bragginga number nobody can do anything withvanaty metricthe dashboard number nobody can act on

See it

Live demo coming soon

What it is

A vanity metric looks impressive but does not tell anyone what decision to make. Total signups, downloads, followers, and pageviews usually rise with time even when the product is getting worse. They have volume and a pleasing slope, but no denominator, quality check, or clear link to the value customers received. Eric Ries drew the vanity versus actionable line in The Lean Startup, and the label stuck.

Reach for the label when a dashboard review stalls at 'is this number good?' Turn the total into a rate or cohort outcome: activated signups per eligible visitor, week-four retention by signup month, paid conversion per trial, or repeat orders per buyer. An actionable metric does not need to be small or tactical; it needs an owner, a decision, and a known behavior that can move it.

Gotcha: no metric is born vain. Pageviews can be the right operational input for an ad-supported publisher, and revenue can guide a pricing decision. The metric becomes vanity when it is detached from the question. Goodhart's law adds the second trap: once a number becomes a target, people and systems find ways to raise it without improving the thing it was supposed to represent, so pair it with guardrails.

Ask AI for it

Audit this GA4 dashboard for vanity metrics. For every chart, name the decision it supports, the owner who can act, the denominator, the customer behavior it represents, and one guardrail. Replace cumulative signups, downloads, followers, and pageviews with a metric tree rooted in the North Star metric: replace a cumulative count with cohort retention when the question is repeat behavior, with activation rate when it is first-time value, and with conversion rate when it is a funnel step. Keep a raw total only when it drives a capacity or cost decision. Show numerator over denominator beside every rate, and finish with a table of keep, replace, or delete plus the exact reason.

You might have meant

north star metricguardrail metricdashboardconversion rateactivation event