Market insights: pick the method by the question, not by the budget
Most studies that disappoint were chosen before the question was written. A survey answers "how many"; a group answers "why"; neither answers the other well. This page orders the methods available today — including synthetic simulation — by what each actually answers, with the cost and the lead time, for teams looking into Spain, Mexico or Chile from somewhere else.
Two families of question
Size questions — how many, what share, how much more than last year — need a sample designed to project. Comprehension questions — why, compared to what, what stops them — need conversation.
If you cannot tell which one you have, write down the decision the result will drive. "How much budget do we commit" is size. "Which message do we run" is comprehension.
What each method answers
None of them is better. They answer different things and they get in each other's way when used out of place.
- Online panel survey: incidence, preference, size. Projectable with a sound sample. Days to weeks.
- Focus group, in person or online: comprehension, vocabulary, group dynamics. Not projectable. Weeks, and thousands.
- Depth interviews: individual journeys and topics a group suppresses.
- Your own behavioural data: what people actually do. Unbeatable, and only available if you already sell there.
- Synthetic simulation: hypotheses and exploration in minutes, near-zero marginal cost, zero projectability.
The order that saves money
Cheap exploration first, expensive measurement second. Cut ten hypotheses to three where being wrong costs nothing, tighten the wording, and then measure what is left with the sample it deserves.
A questionnaire that reaches the panel already pruned costs the same and returns far more, because every unnecessary question is both spend and drop-off.
Entering a market you have never sold in
The hardest part is not the data, it is knowing what to ask. Prices, retail structure, how people talk about credit and instalments, and which references are meaningful all differ across the three markets — and a question that assumes the wrong one produces a confident, useless answer.
Running the first version of the guide against a grounded synthetic panel at €7.99 exposes those assumptions before you pay a local agency to discover them for you.
What the report has to say
Whatever the method, the report must state what it was run with, on how many, and with what limit — that is what lets the reader know how much weight to put on each sentence.
For synthetic audiences the label is explicit and it travels with the export: synthetic participants, directional signal, not a replacement for research with real people.
Methods by question, lead time and indicative cost
| Method | Answers | Lead time | Indicative cost |
|---|---|---|---|
| Online panel survey | How many, what share | 3–10 days | €600–6,000 |
| Online focus group | Why, in whose words | 2–4 weeks | €3,000–9,000 |
| Depth interviews | Individual journeys | 2–3 weeks | €2,000–6,000 |
| Synthetic room | Hypotheses, vocabulary, reaction | 2 minutes | €7.99 for 60 answers |
Frequently asked questions
Run one session per market rather than one averaged session. Spain, Mexico and Chile answer differently on price, retail and register, and a panhispanic average describes nobody. The setup takes a minute to duplicate.
Keep reading
Last reviewed: 2026-09-14. Participants are AI-generated synthetic archetypes, not real people.