What you need to learn before you enter
Start with the decisions the research must inform, not with a method. Before entry, the decisions are usually: which segment to target first, what to say to them, how to price and package the offer, and through which channels. The standardisation and adaptation comparison adds the decision that sits over all of them: which elements of the home marketing can travel and which need changing, element by element. Research earns its cost when it moves one of those decisions. In the international market entry playbook, this research is the step after the pilot is designed and before prices are set or any copy is localised, so its answers can still change the plan.
From the decisions, write the learning objectives as questions:
- The job and the need. What are customers in the market trying to get done, and how do they describe it? The jobs-to-be-done framework is a good frame, because a job travels better than a feature list and shows when the same product is bought for a different reason.
- The alternatives. What do people use now, including doing nothing or a different category?
- The segments. Are the home market's segmentsDividing a market or customer list into groups that need different offers or messages. present, in what form, and is there one the home market lacks? A home buyer personaA research-based profile of a typical customer type: goals, problems and buying behaviour. is a hypothesis to test here, not a fact to translate.
- The words. What language do customers use for the category, the problem and the benefit?
- The barriers. What would stop them buying from a brand they do not know, from abroad?
Desk research, trade voices and customers
Three sources answer different parts of those questions, and they work best in order.
Desk research first. The desk research guide covers finding and judging published evidence, and the PESTEL framework gives a checklist for the context. In a new market, desk researchResearch from existing sources, the business's own and published ones, done first so that interviews and surveys cover only what nobody has answered. also tells you what you do not know, which sharpens the brief for everything that follows. Search behaviour in the language is a form of desk research; the guide to keyword research for new markets shows how people actually name the category there.
Trade voices next. Distributors, consultants, retailers and trade journalists know the market's buyers and channels. The guide to interviewing industry experts and the trade covers approaching them honestly and weighing what they say, since each has a view to promote.
Customers last, and directly. Experts describe customers; only customers can say how they decide. Qualitative work, such as customer interviews, usually comes first in a new market because the team does not yet know the right questions or answer options. A survey follows when the business needs to know how common a view is, and its options can then come from what interviews heard.
| Source | Answers well | Answers badly | Language work needed |
|---|---|---|---|
| Desk research | Market context, competitors, published data, search demand | How customers decide and what they feel | Reading sources in the language; checking translations of key figures |
| Trade and expert interviews | Channels, buyer types, competitive moves, what failed before | Customers' own reasons and words | Interviews in the language or with an interpreter |
| Customer interviews | Jobs, alternatives, decision journeys, language, barriers | How common any view is | Translated guide, native-speaking moderator, translated transcripts |
| Customer survey | How common views and behaviours are within the market | Why people hold them; anything the options miss | Translated and back-translated questionnaire, pretested in the language |
Questionnaires and guides in another language
A questionnaire or discussion guide has to mean the same thing in the new language, which is a different test from reading well. Work in four steps.
Adapt before translating. Review the source for anything that only makes sense at home: the category name, brand names in answer lists, examples, units, prices, income bands, the alternatives offered. Replace them with the market's equivalents, with help from the local partner. The survey and questionnaire design guide applies as written; the adaptation is an extra pass over its output.
Translate professionally. Use a translator who is a native speaker of the target language and understands research, and give them the objectives of each question as well as its text, so they translate the meaning rather than the words.
Back-translate as a check of meaning. Have a second translator, who has not seen the source, translate the target version back into English. Compare the back-translationA translation of the target-language version back into the source language, made to check what risky lines say; it shows meaning, not how the text reads. with the source question by question. Differences show where the meaning moved: a question that became leading, a scale whose labels shifted, a term that became more specific or more general. Resolve each one with both translators and the local partner. Back-translation is a check, not a goal: a back-translation that matches the source word for word can still sit on an awkward target text.
Pretest in the language. The guide to pretesting a questionnaire covers cognitive interviews and soft launches. Run them with people in the market, in the language, led by a native speaker, and ask them to say what each question means in their own words. Pretesting catches what translation checks cannot: a question that is clear but sensitive, or an option list that misses the most common answer. A market entered on the strength of a translated, untested questionnaire that has stayed small is one of the cases the diagnosis page on a new market not gaining traction separates from the other ways an entry fails.
Recruiting in the market
The guide to recruiting research participants covers screenersThe short questionnaire that confirms each research candidate fits a study, asking about recent behaviour without revealing which answers qualify., sources and incentives; abroad, the practical questions change. Where do people in the market gather who fit the profile, and who can reach them? Panels, the local partner's networks, customer lists of a local distributor where that is permitted, communities and social platforms all differ by market, and a local recruiter usually knows which work.
Translate and pretest the screener the same way as the questionnaire, because a screener question that drifts in translation lets the wrong people in. Set incentives at a level the local partner recommends, in local form and currency. Keep a recruitment log, so the business can see who was approached, who qualified and who took part, which is what later lets you judge whether the sample describes the market. As in any study, the recruitment logThe study's restricted record of who was approached, who qualified and took part, with screener answers, consent, sessions, incentives and a deletion date. stays apart from the notes and findings, which refer to participants by code.
Moderators, interpreters and fieldwork
Interviews and groups in the customer's language are the default. A native-speaking moderator hears hesitation, politeness and humour that an interpreter relays only partly, and participants speak more freely in their own language. Brief the moderator on the objectives, the guide and the product, and hold a short rehearsal.
Where the home team needs to follow live, use simultaneous interpretation for observers rather than consecutive interpretation in the room, which halves the time and changes the conversation. Agree in advance how observers send questions to the moderator, and keep their number small.
After each session, hold a short debrief with the moderator before anyone reads a transcript. The moderator's reading of tone and emphasis is evidence too, and it is lost if nobody asks for it.
Consent and data in another jurisdiction
The principles do not change at the border. The professional codes the site's research guides describe from the MRS and ESOMAR set the standard: participation is voluntary and informed, people are told who is collecting the data and why, answers stay confidential unless people agree otherwise, and research is never a pretext for selling. The participant information and consent guide turns each into a written document.
What changes is the law and the language. Research data is Personal DataAny information relating to an identified or identifiable living person, such as a name, email address, location data or an online identifier.: recordings, contact details and free-text answers that identify someone. In the EU, the GDPR sets the framework for processing it, as the site's research guides describe; other jurisdictions have their own laws. This guide names no market's research or data rule. What applies in a given market is checked locally with whoever owns data protection in the business and with the local partner, and the participant information is written for that jurisdiction, in the language, and reviewed by them. The guide to commissioning a research agency covers agreeing in the contract who holds what data, what is returned or deleted and what the business may reuse.
Decide before fieldwork where recordings and transcripts are stored and who abroad can access them, and write the deletion date into the participant information and the recruitment log.
Comparing answers across markets
Results from two markets invite comparison, and comparisons are where cross-market research most often misleads. A score can differ between markets for reasons that have nothing to do with opinion: a question or a scale label that shifted in translation, a sample that reached different kinds of people, or differences in how people in a market tend to answer rating scales. This guide makes no claim about how any market answers; it says only that scores may not compare directly.
Four practices help. Compare a market with itself over time wherever possible, rather than with another market. Compare behaviour as well as attitude: what people did and bought is easier to compare than how strongly they agree. Look at the order of priorities within each market rather than the level of scores across them. And ask the local partner how to read the results before writing conclusions. Report the sample sizeThe number of users or sessions an experiment needs to detect a given effect reliably. and the recruitment route for each market beside every comparison, so readers can judge it.
Analysing what was said in another language
Transcripts in the language are the source; translations are working copies. Code open answers and transcripts in the original language where a native-speaking analyst is available, using the method in the guide to coding open-text answers, and translate the codeframe and the quotes, not the whole corpus. Where the team must work from translations, keep the original beside each quote used in a report, so a native speaker can check it.
Watch for words that matter. The term customers use for the category, the benefit they name first and the alternatives they mention are often the most useful findings, and they are lost when a translator smooths them into the home vocabulary. Ask translators to flag words they found hard to render and to leave key terms in the original with a gloss. Those words go into the brief for the localisation work that follows.
A coffee roaster maker researches a new market (illustrative)
Bruxeppa Roasters researches buyers in Market A
Bruxeppa Roasters, an invented maker of home coffee-roasting machines, is preparing to enter Market A. Desk research and 4 interviews with retailers suggest demand, but not why people buy. The marketing lead commissions a local agency for 16 in-home interviews with people who roast at home or have considered it, moderated in Market A's language.
The discussion guide is adapted with the agency, translated, back-translated by a second translator and pretested in 3 interviews. The back-translation shows that the home phrase for "saving on coffee" had become a phrase closer to "economising", which participants in the pretest found off-putting; the question is reworded. The agency writes the participant information for Market A's jurisdiction with Bruxeppa's data protection lead, with a deletion date for recordings 6 months after fieldwork.
The agency's analyst codes the transcripts in the language and translates the codeframe and 40 quotes. Home roasting in Market A is framed as a hobby and a craft in 11 of 16 interviews, with saving money mentioned in 3; at home, the brand's message leads on saving. The launch message for Market A is rewritten around the craft, and the words participants used for it go into the localisation brief.
Common mistakes
- Translating the home questionnaire. It carries home assumptions. Start from what the market decision needs, adapt, then translate.
- Treating a matching back-translation as proof. It checks meaning; the target text still needs a native reader and a pretest.
- Skipping the pretest because the questionnaire worked at home. The translated version is a new instrument.
- Moderating through an interpreter by default. A native-speaking moderator hears what an interpreter cannot relay.
- Using home consent wording. Participant information is written for the jurisdiction, in the language, and checked locally.
- Ranking markets on raw scores. Translation, samples and scale use can move scores. Compare within markets and on behaviour.
- Analysing only translations. Key words are smoothed away. Code in the original where possible and keep originals beside quotes.
- Researching after the plan is fixed. Research that cannot change the message, segment or price is a cost, not evidence.
