The choice of which languages to target will vary for each company, but following is a good rule of thumb to use in putting together your list. This methodology can also be used to sequence language releases which will be discussed later in the article.
Languages Ranked By Population
The first place I start is to look at which languages can unlock the largest populations. Spanish, for example, is spoken in over 30 countries (and large parts of the US). Just by adding Spanish, you can unlock much of Central and South America along with Iberia. Chinese and French are two other widely spoken languages. It’s also important to note that regions like Latin America the rate of English proficiency is relatively low, so lack of language support can be a blocker to many users. Following is a list of the top non-English languages by population.
| Language | Speakers | Primary Region(s) | English Proficiency |
|---|---|---|---|
| Chinese | 1.18 Billion | China, Taiwan, Hong Kong, APAC Diaspora | Varies |
| Hindi | 602 Million | India | High, due to history |
| Spanish | 548 Million | Central and South America, Iberia, southern USA | Varies |
| French | 280 Million | France, large parts of Africa | Varies |
| Arabic | 274 Million | Middle East, northern Africa | Varies, but higher with business users |
| Bengali | 272 Million | India | High |
| Russian | 258 Million | Russia, Eastern Europe | Low |
| Portuguese | 257 Million | Brazil, Portugal | Low |
| Urdu | 231 Million | Pakistan | Low |
| Indonesian | 199 Million | Indonesia | Varies |
| German | 134 Million | Germany, Austria and Switzerland | High |
| Japanese | 125 Million | Japan | Low |
Countries Ranked By Per Capita GDP
Another way to rank countries and languages is by per capita GDP for the regions where they are spoken. Japan and the Nordics score well on this basis, and indeed, they are important regions for many companies. The table below compares per capita GDP for some of the largest economies.
Note: a comparison of on the basis of purchasing power parity may also be useful here.
| Country | Per Capita GDP | Languages Spoken | English Proficiency |
| USA | $85,000 | English, plus Spanish in southern regions | High |
| DACH Region | $63,000 | German, English as a bridge | High, but prefers German |
| France | $46,000 | French, English as a bridge | Medium |
| Japan | $37,000 | Japanese | Low |
| South Korea | $35,000 | Korean | Low |
| China | $14,000 | Chinese (Simplified) | Low/Medium |
| Brazil | $10,000 | Portuguese | Low/Medium |
One other thing to consider in ranking opportunities is price sensitivity which may make a seemingly large market less attractive. Lower per capita GDP is generally correlated with increasing price sensitivity. India is good example, with a per capita GDP of $2,700. There the combination of high English proficiency and low GDP suggests that the return on localization may not be as high as raw population numbers suggest.
Languages Ranked By TAM (Total Addressable Market)
This maps more directly to your company and product. Here you want to rank countries and regions in terms of the total addressable market for your industry segment. For each country, identify the language(s) spoken there. This will give you a better idea of which countries to target, and from there, you can identify the languages you’ll need to support. AI based research tools can be helpful in assessing the relevant TAM metrics for your industry sector.
Languages Ranked By User Activity
Lastly, if you are seeing organic growth in a particular region, it will only help to add language support. At Notion, we saw huge growth in Japan and Korea. Adding support for Japanese and Korean only accelerated our growth, and today, these are two of Notion’s most important regions.
Scoring Languages Based On Accessibility
Another thing to do with each country you target is to access the average level of fluency in widely used languages such as English, Chinese, French and Spanish. In much of Scandinavia, the level of English proficiency is quite high, so localizing to a language like Norwegian is more for boosting brand perception and visibility than necessity. In Brazil, and Latin America in general, the rate of English proficiency is low except among people who went to university in English speaking countries.
Consumer Versus Business To Business Services
Another factor to consider is whether your product or service is consumer or business to business. Consumer facing services generally do better if the are localized for local markets, even where the level of English proficiency is high (games are a great example of this). Business facing services can often get by with supporting major international languages like Chinese, French and Spanish (caveat: if you are selling to enterprise customers, decision makers will often include language support as a prerequisite).
| Segment | Language Accessibility |
| B2B Enterprise | Often Required |
| B2B Mid Size | Required In Low English Areas |
| B2B Small Business | Not Required, Recommended In Low English Areas |
| Consumer | Often Required |
Costs & Barriers To Entry
Some of these markets will be difficult to enter due to implementation costs, regulatory barriers and other factors. China, for example, is a difficult market to enter because of both the implementation cost as well as trade barriers that favor domestic companies.
In other regions, such as the Middle East, the cost of implementing right-to-left scripts will often cause companies to rate languages like Arabic and Hebrew as a lower priority.
Japan is also tough in terms of technical implementation, with the need to support Japanese scripts, significant QA requirements, and high quality expectations. However, it is often worth the investment since Japan is a wealthy country with a low rate of English proficiency.
Combining These Lists
I generally recommend developing a target list using a combination of these lists. Spanish and Chinese unlock such large user populations that it is a no-brainer to add them (the relatively low English proficiency in Latin America is also a blocker for many users). Then add languages based on a combination of TAM and GDP to go after higher income regions. And, if you have strong organic growth in specific regions, add the languages relevant to those. Most companies will land on somewhere between 5 and 20 languages.
Sequencing
Now that you have a ranked list of languages, it is a good time to think about sequencing.
Phase 1 : Initial Launch
With the first release, you will be demonstrating that the translation tooling and infrastructure is built out and operating correctly. I generally recommend launching one or two languages at most, This is also the logical time to launch a secondary language for use in a domestic market (e.g. Spanish for the US market, Turkish for the German market). The primary goal here is to prove out the tooling to support the next launches.
Phase 2 : Adding More Languages
In this phase, you will typically launch a batch of 2-4 languages, mostly likely targeting regions where you are seeing organic demand, since these regions have demonstrated potential if language barriers and other barriers to usage are reduced. At Notion, our first four languages were Korean, Japanese, French and German. At this point, most of the engineering work and spend to support adding languages is done, so the main costs at this stage are translation related work plus product and engineering work to fix functional issues that are specific to the recently launched languages.
RTL issues for Arabic, date format problems for Japanese, sort order bugs for certain character sets are examples of the type of functional issues you can expect. This also underscores that localization is never “done” and will require support from product and engineering long-term.
Phase 3 & Beyond : Scaling
In phase 3 and beyond, localization is operationalized. Launching new languages is largely incremental, although you will continue to encounter functional issues (the good news is that most of these will affect multiple languages, so the rate at which these are surfaced should drop as more languages are offered). This is also a good time to think about themed releases. At Notion, we launched Spanish and Brazilian Portuguese to target LatAm users, followed by Dutch and Nordic languages to target the Nordics as a region. They have since gone on to similar languages targeting APAC and Middle Eastern users and are now in 19 languages and dialects.
Grouped regional launches allow you to concentrate marketing and support resources, generate regional PR, and build momentum in a specific geography rather than spreading effort too thin.
Rollbacks
Does it ever make sense to rollback language releases? Generally there are few cases where this make sense. The money to generate the translations has already been spent. The one case where it makes sense is when the cost of customer support is high relative to the revenue coming in.
In this scenario, it may be worth looking at disabling this language(s) for customer support and other contact points while leaving product localization on. This may also make sense if the languages are for a high English proficiency like the Nordics.
That said, it is usually better to focus on fixing whatever issues are holding users back than to abandon languages. Most likely this is a marketing issue and not a functional or technical issues.
Related Reading
Get Out Of The User’s Way – Before you even start the process of formally translating and localizing your product, start by fixing functional and design issues that are blocking early adopters. Early adopters can also be a great source for insights about their region and how well your product works for them.
Choosing A Global Ready Content Management System (CMS) – Before you start translating your website, help center and other user touch points, you will need to make sure that your CMS can support multilingual operation and use cases. This article explains what questions you should be asking potential vendors and other issues to be aware of.
Translation Management Systems : Which One Is Best? – One of the decisions you will need to make before rolling out language support is which TMS (translation management system) is a good fit for your use case.
Budgeting For Localization – When you are ranking and sequencing languages, this is also a good time to scope out your budget so you know what to expect.
Spanish As A Prelude For International Expansion – As discussed, launching Spanish for the US market is a good way to decouple language support from international expansion. This is especially useful if you have exposure to regulatory and compliance issues when operating in other countries.
Get In Touch
Want to learn more or discuss your international expansion plans?