← All tools
Analytics utility · Definitions

Metrics dictionary builder.

“Retention is up” is useless if two teams mean different things. Generate a versioned definitions table your team can own.

Tick the metrics you report, set the owner and version, and copy the output into your wiki or data catalogue. Edit the definitions to match your own event schema.

Metrics to include

Add a changelog line every time a definition changes — who approved it, and which dashboards, reports and models just changed meaning.

How it's calculated

The method, in plain terms.

  1. One row per metric: definition, formula, filters, grain, owner, version
  2. Changes append to a changelog with affected dashboards and models
Worked example

“D7 retention — users active on day 7 ÷ users installed on day 0; excludes test devices; owner: Analytics; v2 (2026-08): switched from rolling to classic” tells every reader which number they are looking at.

Defaults on this page are illustrative starting points, not benchmarks for your game. Replace them with your own numbers.

The thinking behind it
Terms on this page

Metrics dictionary · KPI tree

Embed this calculator

Free to embed on your blog, wiki or course page with the credit link intact.

FAQ

Common questions.

What should a metrics dictionary include?

For each metric: a plain-language definition, the exact formula, filters and exclusions, the grain it is reported at, an owner, a version and a changelog. The changelog matters most — it is what tells people a number changed meaning.

Who should own metric definitions?

A named person or team per metric, usually analytics for core KPIs. Ownership means approving definition changes and telling the dashboards, reports and models that depend on them.

Classic or rolling retention?

Either can be right; the dictionary’s job is to say which one you use. Classic (day-N) retention counts users active exactly on day N; rolling counts users active on day N or later. They are not comparable.

How do I keep the dictionary from going stale?

Treat it like a product: review it when a metric changes, attach the change to the dictionary entry, and list which dashboards and models are affected. A dictionary written once during onboarding drifts within months.

Calculator by The Game Scientist ↗