BigCat Research
Research in emerging markets: common challenges and how credible firms mitigate them
Short answer: The main challenges are incomplete sampling frames, uneven access across modes, variable interviewer quality, social desirability bias, translation problems, data quality control and large informal sectors. Credible firms mitigate them with transparent design, pilots, supervised fieldwork, back-checks and an honest statement of what the data cannot show.
Why "emerging market" research needs its own care
The label covers very different countries, so treat it as a prompt for questions, not a method. What the markets share is that the tools borrowed from mature markets (a complete address register, a high-quality online panel, a landline directory) are often partial or skewed. A study that ignores this still produces numbers. They just describe a narrower population than the report claims.
For buyers, the useful question is never "do you have experience in emerging markets?" but "which of these seven problems apply to my study, and what exactly will you do about each?"
Challenge-to-mitigation table
| Challenge | What goes wrong | How credible firms mitigate it |
|---|---|---|
| Sampling frames | No complete list of households, businesses or consumers; informal settlements and new housing missing | Use the best available official frame, document its gaps, apply area-based or random-route selection, state coverage limits |
| Access and mode | Online panels can over-represent urban, younger, more educated people; phone and face-to-face reach different groups | Choose mode per audience; mix modes where needed (CAPI, CATI, online); report the mode and its bias |
| Interviewer quality | Inconsistent reading of questions, shortcuts, fabricated interviews | Briefing and training, written manuals, supervision, accompanied interviews, back-checks of a share of completes |
| Social desirability | Respondents give acceptable answers on income, politics, health, brands | Neutral wording, self-completion for sensitive items, indirect questions, assurance of anonymity |
| Translation | Literal translation changes meaning; dialect and register differ | Forward-and-back translation, review by a native speaker with subject knowledge, cognitive pre-testing |
| Data quality | Straight-lining, speeders, duplicates, implausible values | Logic and range checks, attention items, timing rules, audit trail, documented cleaning rules |
| Informal sector | Businesses and workers not in registries | Screeners that reach informal actors, qualitative scoping first, explicit exclusion of what cannot be sized |
A checklist to test any provider's claims
| Question to ask | A credible answer includes |
|---|---|
| What is the sampling frame and where is it weak? | Named source, known gaps, how gaps are handled |
| Which mode, and who does that mode miss? | Acknowledged bias, not "representative of everyone" |
| How are interviewers selected, trained and supervised? | Training content, supervision ratio, back-check share |
| Was the questionnaire pilot-tested? | Pilot size, what changed afterwards |
| How are sensitive questions handled? | Wording and method specifics |
| How was translation validated? | Process, not just "native speakers" |
| What data quality rules apply? | Written rules, share of removed cases reported |
| How is weighting done? | Variables, source of targets, effect on estimates |
| What are the limitations? | A written section in the report |
If an answer is vague, ask to see the methods annex of a previous report.
How to run a consumer survey in Turkey, step by step
- State the decision. What will you do differently depending on the result? This decides the questions and the sample.
- Define the population. Age, city size, product use, income band, as the decision requires. Narrow populations (for example buyers of one product category) need screening, which raises cost.
- Choose the sample frame. National population statistics from TurkStat (address-based population registration and household surveys) are the standard reference for weighting targets.
- Choose the mode. Online reaches internet users and is quick; face-to-face covers those offline; CATI sits between. Check current internet and mobile use in the target population before relying on online alone.
- Write and translate the questionnaire. Draft in the language of respondents. If an international study needs one questionnaire across countries, use forward-and-back translation and test it.
- Pilot. Test with a small group, check timing and comprehension, fix wording.
- Fieldwork. Monitor quotas daily, supervise, apply back-checks.
- Clean and weight. Apply quality rules, weight to population targets, and report effective sample size alongside the raw sample.
- Report with limits. Methods, margin of error where probability sampling applies, and what the data cannot answer.
A market-entry research playbook for Turkey
- Desk research. Official statistics, sector reports, regulator publications, trade bodies, earlier studies. It can answer many market-size, structure and competitor questions before any fieldwork is commissioned, which keeps primary research focused.
- Competitor mapping. Who sells what, to whom, at what price and through which channels.
- Pricing. Collect observed prices and offer structures; test willingness to pay only after you know the range.
- Distribution. Retail, dealer, online marketplace or direct; margins and logistics constraints.
- Consumer survey. Use the steps above to size demand, preferences and barriers.
- Regulatory check. Licences, labelling, import and data rules, ideally with a qualified local adviser.
FAQ
1. When is desk research enough? When the question is about market size, structure, competitors or regulation and credible secondary sources exist. It is not enough when you need to know what a specific group of people thinks or intends to do.
2. When do I need primary research? When no existing data answers your question: brand perception, price sensitivity, unmet needs, reasons for rejection.
3. Can online surveys represent the whole population? Usually not. Report who is covered and weight where possible; add other modes if offline groups matter to the decision.
4. How big a sample do I need? It depends on the number of segments you need to read separately, not only the total. Each reporting segment needs enough cases to be stable.
5. How do I judge a firm's local knowledge? Ask for the methods annex of a comparable local study, the fieldwork supervision process and who will manage the project day to day.
How BigCat approaches this
BigCat Research is based in Istanbul and works in Turkey. We make no claims here about other countries or about panel sizes. Relevant service pages: market research, desk research and competitive analysis. Our stated approach is to define the uncertainty first and to consider desk research before proposing fieldwork, so that primary research is used where it adds something. Documented projects, including the FESAS field operation (7,000+ respondents across 6 provinces and 25 districts, as stated in BigCat's public profile), are listed on the projects page.
Related reading: BIG 101 knowledge hub.
Sources
- TurkStat (Turkish Statistical Institute), population and ICT usage statistics. data.tuik.gov.tr
- Published methodology literature on survey sampling, mode effects and questionnaire translation (for example survey methodology handbooks from AAPOR and the Cross-Cultural Survey Guidelines).
Author: BigCat Research team. Last reviewed: 3 October 2026.