Rules
How to brief a market research methods mix that holds up in 2027
A step-by-step briefing process for pairing each business question with a market research method, sample and evidence standard, from first estimate to sign-off.
What to take away
- A brief holds up when every business question gets a named method, a named sample and a stated evidence standard before fieldwork starts.
- Write the decision, the deadline and the cost of being wrong on one page, then let those lines pick the technique.
- Qualitative work explains reasons, quantitative work sizes them, and most programs need both in a set order.
- Sampling sits inside the brief, because it caps what the final report can claim.
- Sign-off comes when the report settles one live decision and the data trail can be checked months later.
Start from the decision, not the technique
A method mix is defensible only when it traces back to a choice. Write the decision, the deadline and the cost of being wrong on a single page. That page becomes the spine of the brief.
Vendors reply faster when the brief names the decision. A request for consumer insight invites a generic proposal. A request to test an 8 percent price rise in two regions invites sampling, timing and a budget line.
For the wider frame, Making sense of market research strategy (2027 update) connects decisions, scope, evidence gaps, ethics, inclusive samples, quality controls and action.
Pair each question with a method
| Business question | Method | Sample | Evidence standard |
|---|---|---|---|
| Why do buyers hesitate? | Focus groups, depth interviews | 8 to 12 per segment | Themes with verbatim support |
| Which features drive choice? | Conjoint or choice modelling | 300 or more completes | Part-worth utilities with intervals |
| Is the ad understood? | Copy comprehension test | 150 per market | Unaided recall plus meaning checks |
| How many would switch? | Tracking survey | Quota panel sample | Share shift with margin of error |
Qualitative vs quantitative methods is a sequencing question, not a loyalty question. Run two or three groups to hear the language buyers use. Then measure how often that language appears at scale.
Choose a sample that can carry the claim
Sample size follows the smallest effect you must detect. Check published formulas for confidence level and margin of error, such as sample size determination, before you accept a vendor's fixed package.
Panel composition matters as much as size. Buy online panel capacity with documented recruitment sources, quota controls and attention checks. Ask how duplicate participants and professional respondents get removed.
Teams that build holdouts, panels and documentation into the plan write stronger briefs. 6 sampling choices Seattle research teams make at Amazon and Microsoft shows the specification page that keeps those decisions visible.
Pilot the instrument before the full field
- Draft questions against the decision, one question per claim you must support.
- Run a pilot of 20 to 30 completes and time the median interview.
- Review open-ended answers for ambiguity and coding load.
- Fix wording, quotas and routing, then freeze that version.
- Field the main wave and log every deviation from the frozen plan.
An unpiloted instrument produces clean-looking data and unusable answers. Budget the pilot as part of the project, never as an optional extra.
Example: briefing a pricing question
Suppose a retailer must choose between a flat price and a two-tier bundle. The brief states the decision, the two options and a tolerance of three points on share.
The method is a choice-based survey with a conjoint module. The sample is 400 current buyers, quoted to match the customer file. The evidence standard is preference share per option with confidence intervals.
The brief is finished when a stranger could run the study and reach the same recommendation.
Ethics, compliance and cost
Consumer contact rules reach recruiting calls and texts, not only sales. The vendor questions in FTC guidance on the Telemarketing Sales Rule belong in the brief before any list is dialed.
DIY tools cut cost per complete but shift sampling and quality control onto your own team. Professional fieldwork buys process, not only respondents, and the brief should say which of the two you are paying for.
Getting sign-off and closure
The report answers one real decision, names the limits of the sample and states what would change the recommendation. Primary market research: the version that survives contact with reality shows how a piloted instrument and a matched sample keep a study useful.
Persuasion testing and comprehension testing are not interchangeable across cities. Los Angeles advertising psychology compared with Chicago consumer research explains why the same stimulus can be tested in two different ways.
A study stops needing attention when the decision is recorded, the data is archived and the next brief inherits the lessons.
