Minimum Viable Localisation (MVL): Testing with Less Translation

Every company that wants to open an application to a new market runs into the same question: is it worth translating in full? Full localisation is costly, and there is no way of knowing at the outset whether that market will show any interest. This is where minimum viable localisation comes in as a clever middle way of managing the uncertainty; it makes it possible to test a market with very little translation.

Contents

This article looks at what minimum viable localisation is, at the levels of localisation, at the range that runs from the zero level to deep localisation, at deciding from download data and at the moment of moving deeper. The aim is to stop localisation being an ‘all or nothing’ decision.

What is minimum viable localisation?

Minimum viable localisation (MVL) is a level of localisation at which an application’s text is translated to the smallest possible extent and the application is trialled in selected markets (Jiménez-Crespo, 2024). Under this approach only a minimum portion of the text is translated and the application is tried out in a few chosen regions. The aim is to gauge the market’s interest without making the full investment.

The concept turns localisation into a risk management tool. Investing in full localisation for a market can be risky while it is still unclear whether that market will respond. MVL reduces the risk: the company tests the water with a small step first, and takes the larger step according to the results.

The essence of MVL is flexibility. Instead of making the same investment in every market, it becomes possible to advance in stages, in line with each market’s potential. This approach is what lets teams on limited budgets try many markets sensibly.

Levels of localisation

To understand MVL you have to see localisation as a spectrum of levels. The table below compares the three basic levels.

Level	What is translated	Purpose
Zero localisation	The store description only	To gauge interest
MVL	Minimum in-app text	To test the market
Deep localisation	All text and dedicated features	To enter the market fully

These levels work like a ladder; a company can start at the bottom and climb. Each rung offers a chance to gather more information before investing in the next. Resources are then directed to the markets that respond most.

Zero localisation

At the bottom end of the spectrum sits zero localisation. At this level the company translates only the application’s store description and distributes the application as it is (Jiménez-Crespo, 2024). The in-app text is left untouched; only the store listing is carried into the target language.

The purpose is to observe whether the number of downloads rises in that market. When the store description is in the local language, the application becomes more visible there. The company then watches the download data to see whether that visibility turns into genuine interest.

Zero localisation is the cheapest form of testing. Because no in-app translation is done, the investment falls to a minimum. Its limit is just as clear: since the in-app experience is not local, users may struggle to stay in the application even if downloads rise.

MVL and deep localisation

MVL sits one rung higher. Here a minimum portion of the in-app text is translated as well and the application is tested in selected markets. This goes beyond the zero level and gives the user a minimal local experience, so the market’s interest is measured more reliably.

At the top of the spectrum is deep localisation. At this level, much as in game localisation, dedicated features and launch campaigns are designed for the target market (Jiménez-Crespo, 2024). The application is not merely translated; it is adapted specifically for that market and distributed through local channels.

The move between these two levels is a staged investment decision. When MVL produces a positive result in a market, the company can consider moving to deep localisation. You can find more information about these tools in our knowledge centre.

Deciding from download data

The heart of the MVL approach is deciding from data. The company watches whether the number of downloads rises in any given target market (Jiménez-Crespo, 2024). This data is a concrete indicator of that market’s real potential; it puts measurement in the place of guesswork.

Download data makes the investment decision objective. If interest stays low in a market, the company stops before committing further resources there. If interest is high, there is a case for taking localisation up a level. Every decision then rests on evidence rather than instinct.

This data-driven approach makes the most efficient use of resources. Rather than fully localising dozens of markets at once, the company first tries all of them at a minimum level and then invests deeply only in those that respond. This is both an economical and a strategic method.

Which level, and when?

Choosing the right level depends on the type of application and on the company’s goals. For an uncertain, new market, the zero level or MVL makes sense; for a strategic market with proven potential, deep localisation is justified from the start. The decision calls for risk and opportunity to be weighed together.

The nature of the application matters too. An application with a great deal of text, whose experience depends on language, will not take hold if it is localised at a low level; a visually driven application with little text can circulate in a market even at the zero level. The choice of level follows the product’s dependence on language.

MVL is therefore not a strategy in itself but a decision framework. The company looks at each market through this framework to determine where to start and when to move on. Flexibility is the greatest strength of the approach.

Server-based content and MVL

One layer that is often missed in the MVL decision is server-based content. Some applications pull their content from servers: news, product listings or user-generated content. Even if the application’s interface is translated, if this flowing content is not, the user may still find the application foreign (Jiménez-Crespo, 2024).

When setting the MVL level, therefore, you have to account not only for the fixed interface text but for the flowing content as well. A stream of content that is updated continuously demands a process of its own, just as agile web localisation does. An MVL carried out without localising that content can fall short.

The decision changes with how far the application depends on content. A static application with little content can get by at a low MVL level; an application whose content changes constantly cannot offer a genuine experience without localising the flowing content. Seeing this distinction is the key to setting the MVL level correctly.

The risks and limits of MVL

Powerful as MVL is, it has limits. A minimal translation can give the user an incomplete experience; when some parts of the application are in the local language and others remain in the source language, the impression is a confused one. That confusion can make a market look less promising than it really is and lead to the wrong decision.

Another risk is that download data can mislead. Interest measured at a low level of localisation may not fully reflect the potential that would emerge under full localisation. In other words, a market can look weak under MVL and shine under deep localisation. This uncertainty means the decision has to rest not only on data but on judgement.

MVL should therefore be seen as a stage of learning rather than an end point. Its purpose is not to replace full localisation but to light the road towards it. Used with its limits in mind, MVL is a valuable tool for directing resources wisely.

The Türkiye context

Türkiye is a market that many international applications test with MVL. Companies often put the store description into Turkish first, watch the download data and, if interest is sufficient, translate the inside of the application as well. This staged approach lets them measure the potential of the market in Türkiye at low risk.

Turkish users, however, expect the in-app experience to be local too. An application whose store description is in Turkish but whose interior stays in English struggles to hold on to users even once it has been downloaded. Staying at the zero level for long therefore tends to turn into a lost opportunity in Türkiye. A user who opens the application and meets a foreign language inside can give up quickly, which produces a misleading picture of high download numbers and low usage.

Given the size of the market in Türkiye and how widespread mobile use is, MVL is generally valuable as a starting point; but the move deeper is often needed early. Teams that watch download data closely get the timing right on when to deepen localisation in Türkiye. The fact that mobile use in Türkiye long ago overtook desktop means the in-app experience has to be localised early as well; an approach that settles for the store description alone often falls behind competitors in this fast-moving market and stops the initial interest turning into loyalty.

What does this mean for you?

As you open your application to new markets, do not treat localisation as an ‘all or nothing’ decision. A staged approach reduces risk and lets you direct your resources to the markets that respond most. MVL is the framework for that measured progress. It is especially valuable for teams that want to try many markets on a limited budget; testing with small steps first and investing deeply only in the promising markets uses resources far more efficiently than making a full investment in every market at once.

1. Try a new market first with a store description or a minimum translation.
2. Watch download and usage data closely.
3. Deepen localisation step by step as interest grows.
4. Choose your starting level according to how far your application depends on language.

If you would like to talk about which level to start with in which market, you can look at our localisation services and get in touch with us for an assessment suited to your strategy. Starting at the right level lets you use your resources in the most efficient way.

Frequently asked questions

What is minimum viable localisation?

It is a level of localisation at which an application’s text is translated to the smallest possible extent and the application is trialled in selected markets. The aim is to gauge the market’s interest without making the full investment.

Why is MVL preferred?

To reduce risk. Before investing in full localisation for a market, a company can observe that market’s interest and any rise in downloads with a limited translation.

What is the difference between zero localisation and MVL?

At the zero level only the store description is translated and the application is distributed as it is. Under MVL the inside of the application is translated as well, to a minimum extent.

When do you move from MVL to deep localisation?

When downloads and interest in a market have risen sufficiently. The process is then scaled up to a full and deep localisation.

Sources

• Jiménez-Crespo, M. A. (2024). Localization in Translation. Routledge (Ch. 11).
• OneSky. Mobile Localization Guide (Sharifulin 2020; Yip 2022).
• Roturier, J. (2015). Localizing Apps. Routledge.