What to localize next, from your own numbers
Speaker counts size a market. A product's own MAU, DAU and revenue say how much of it that product already serves. Load per-market usage and name the languages shipped.
An upload gives every market a penetration and headroom figure, and apportions MAU across each market's language mix, a ranking signal for the next translation target. The active product then appears in the explorer, the coverage map, the launch planner, and every country page.
- →
- →
- →
Drop your export here, or browse files
CSV or JSON · country + MAU required · DAU, revenue (monthly USD) optional · up to 400 markets
or paste CSV / JSON
Parsed in this browser. Nothing goes to a server.
How the derived metrics are computed
Penetration is MAU ÷ addressable online population (World Bank internet users × CLDR population). It can exceed 100% when a product counts accounts instead of people. Headroom is the remainder: online users the product does not reach. Stickiness is DAU ÷ MAU.
The what-to-localize-next ranking scores each candidate on the share of a market's online population that speaks it and is not covered by a shipped language, times that market's current conversion rate, times a conversion uplift the reader sets, a scenario dial rather than a measured constant. Cost comes from the same model the budget estimator runs.
The language rollup apportions each market's MAU across its languages by normalized CLDR L1+L2 share. It assumes the product's users mirror the territory's language mix, so use it to rank translation targets rather than to report audience sizes.
Uploaded data stays in this browser (localStorage); nothing is stored server-side. Pipelines can call POST /api/atlas/usage for the same join as JSON or CSV.