Which language should you add next?

Atlas maps where the world's languages are spoken and what those markets are worth. Pick the languages a product already speaks, from your own to one of the apps below, and everything follows from that list: how much of the online world it reaches, which markets it misses, and what each new language would add.

250territories
5.8Bpeople online
581languages
Unicode CLDR · World Bank Open Data
Languages you supportStart from a real app, or build your own list.
English

English reaches 21% of the world's online population.

That is 1.2B people across 145 markets, and 33% of world GDP.

Share of each country's people who speak one of your languages. Green is covered, red is a gap.

Loading map…
0% covered100%
No datahover a country · click to open

Languages worth adding

What each language adds on top of what you cover today. “Can pay” is the share of adults with a bank or mobile-money account in those markets (World Bank Findex).

  • Chinese zh+1.2Bonline$30.1T GDP · can pay: 89% · mostly China, Malaysia
  • Hindi hi+407.2Monline$5.4T GDP · can pay: 89% · mostly India, South Africa
  • Spanish es+379.8Monline$11.2T GDP · can pay: 68% · mostly Mexico, Argentina
  • Arabic ar+224Monline$6T GDP · can pay: 48% · mostly Egypt, Saudi Arabia
  • Portuguese pt+190.8Monline$4.3T GDP · can pay: 83% · mostly Brazil, Angola
  • Russian ru+184.8Monline$7.5T GDP · can pay: 81% · mostly Russia, Kazakhstan

Speaker shares count first and second languages, so one person can be covered twice. Per-country totals are capped at 100%. These numbers order markets well. They are not forecasts.

Take it further

The workbench shows you the gaps. These three turn that into a ranked plan, a side-by-side comparison, and a document you can hand to a decision-maker.

Also here: language profiles (what one locale is worth across every country that speaks it), regional groupings, and your own usage data, which stays in your browser and adds a penetration view to every page.

Look up any market

The full dataset, two ways to read it. Both open on demand, so they cost nothing until you ask.

Languages and population come from Unicode CLDR. Internet use and income come from World Bank Open Data. Speaker percentages count first and second languages, so they sum past 100% and one bilingual person is counted twice. That orders markets well and should not be read as a headcount. Every figure is traced to its source.