Synthetic Market Research Brief Template

Copy the brief below, fill in your decision, audience and alternatives, and send it to the providers you want to compare. The blank, filled example and supplier response sheet help you request the same work and inspect the scope before comparing fees.

If you want a team to build the audience, run the study and help interpret the result, we recommend Sapien. You can bring an unfinished brief: we can discuss the design, the segment comparisons and the follow-up questions needed for your next product, price or campaign decision.

Published by Sapien. Prepared on 21 October 2026. The worked example is fictional and contains no study results.

Copy this research brief

1. Decision and owner

• Project and decision owner: [Name, role and contact.]
• Decision this study will inform: [Choose, revise, prioritize or investigate what?]
• Action after the readout: [What could change? What requires additional evidence before action?]
• Decision deadline: [Date and reason.]

2. Audience and buying situation

• Include: [Markets, roles, eligibility, relevant purchase or use situation, and period.]
• Exclude: [People, companies or situations outside the decision.]
• Compare separately: [Only groups whose differences could change the action.]
• Current alternative: [What this audience does or buys today, including doing nothing.]
• Uncertain coverage: [Groups or situations where you need the provider's advice.]

3. Offers or materials to show

• Alternative A: [Full offer, price, terms, image or copy, and version/date.]
• Alternative B: [Equivalent detail. Add alternatives only if needed.]
• Choice context: [Where and when the audience encounters the offer.]
• Keep consistent: [Information or presentation that should remain the same.]
• Deliberate differences: [What is changing, and whether you need to isolate its effect.]

Include a route to reject all offers when that reflects the real decision. If two offers differ in both price and features, their comparison will address the combined offer unless the design separates those effects.

4. Existing evidence and permitted use

• Available material: [Source, collection date, population, relevant finding and known gaps.]
• Access and permissions: [What may be shared, used to construct the audience, retained or reproduced.]
• Restrictions: [Identifying or confidential material to exclude; required handling terms.]
• Held-back reference, if validation is required: [Evidence reserved for evaluating the output rather than informing it.]

5. Questions the study must resolve

• Primary question: [The one answer needed for the decision.]
• Supporting questions: [Comprehension, preference, objections or other specified questions.]
• Design help requested: [What should the supplier propose or improve?]
• Out of scope: [Questions this engagement will not answer.]

6. Required outputs

• Quantitative output, if needed: [One row per named study condition × audience group × question × answer option. Columns: condition/group/question/answer identifiers, response source, eligible modeled base, share (%), human reference and eligible base if any, comparison and its unit, and uncertainty or limitation.]
• Qualitative output, if needed: [Themes and supporting responses tied to the audience, question and stimulus. Label human and simulated responses separately.]
• Recommendation: [Decision, evidence, dependencies and unresolved questions.]
• Inspection and format: [Report, response access, segment views, presentation, export or other deliverables your team needs.]

Do not request a percentage when the intended result is a rank, price estimate or qualitative explanation. State the output and unit you will actually use. Ask the provider to explain uncertainty for its method; do not derive a human-survey margin of error from a generated response count.

7. Evidence needed to act

• Relevant validation: [Audience, task, human or behavioral reference, scoring measure and errors to inspect. Ask for examples if undecided.]
• Decision rule: [What difference or uncertainty would change the action? Agree this before interpreting results.]
• Follow-on evidence: [Customer interviews, usability observation, transaction test or other work needed before the final commitment.]

8. Work, timing and commercial terms

• Responsibilities to assign: [Audience construction, questions, stimuli, approvals, analysis and presentation.]
• Revision allowance: [Changes and follow-up questions needed after the first readout.]
• Schedule: [Inputs ready, approvals, research start, readout and final delivery.]
• Quote required: [Total project fee and currency, inclusions, optional charges, taxes, access period, renewal or minimum commitment.]
• Budget or constraints: [Disclose if useful.]

A short filled example

Project: Linden Rooms, a fictional UK hotel group, is choosing which weekend rate to take into a limited booking test.

Decision: Develop a flexible rate alongside its current non-refundable rate, revise the proposed offer, or gather more evidence before proceeding. The research will help select an offer for testing; bookings from the test will provide behavioral evidence.

Audience: UK adults who personally booked a paid leisure hotel stay in the past twelve months. Exclude business travel booked by an employer. Compare people who usually book flexible rates with those who usually choose non-refundable rates.

Offers: The same room for one Saturday night. A costs £129, paid on booking and non-refundable. B costs £149, paid on booking and refundable until 48 hours before arrival. Both prices include the same taxes and room amenities; breakfast and parking are excluded. Show the same photographs and a choice to keep searching. This compares complete offers, rather than isolating cancellation flexibility from its price.

Available evidence: Existing booking summaries and recent customer feedback, with their dates, coverage and permitted use documented before sharing.

Questions and outputs: Check what buyers understand about payment and cancellation. Compare modeled choice shares for A, B and continued searching within the two audience groups, with the eligible base beside each share. Return objections tied to the offer shown, plus a recommendation and remaining uncertainties.

Work requested: Supplier to propose the audience, instrument and relevant validation. Quote the full study, one review round, the readout format and the cost of any additional testing. Agree the decision rule before viewing results.

The example gives suppliers the same commercial choice. It leaves population construction and instrument design available for their advice. If a proposed return offers one overall intent number for all travelers, it misses the specified UK leisure bookers, the A/B/keep-searching choices and the booking-style groups. Request a corrected scope and deliverable before comparing fees.

Ask each supplier to return this sheet

Copy one sheet per supplier. “Unknown,” “not offered” and “separately priced” are useful answers; a blank should not be interpreted as inclusion.

• Supplier, product and proposal date: [ ]
• Scope accepted and exceptions: [Which parts of the brief will you answer?]
• Who supplies the answers: [Actual participants, simulated respondents, a combination or another source. Identify generated versus observed data.]
• Population and grounding: [Eligibility, groups, source coverage, freshness, customer inputs and material gaps.]
• Research plan: [Questions, conditions, comparisons and approvals. Explain departures from the supplied brief.]
• Validation and uncertainty: [Relevant evidence, metric, reference, discrepancies and limitations.]
• Deliverables and access: [Exact outputs, underlying response access, formats and access period.]
• Named responsibilities: [What your team does, what ours does, and who interprets the result.]
• Revisions: [Included changes, follow-ups, exclusions and additional charges.]
• Timing: [What starts the delivery clock, dependencies and committed dates.]
• Data terms: [Permitted uses, retention, deletion, confidentiality and any training use.]
• Total cost: [Currency, full scope, fee unit, optional charges, minimum term and renewal.]

Compare the returns, then choose the next conversation

First check whether the proposals use the same audience and return the required output. A lower price for a different population or a summary without the needed response access is a different purchase. Resolve those differences before scoring suppliers.

For the service responsibilities behind a quote, use our market research services guide. Our human-versus-synthetic guide explains which evidence supports different claims, and our accuracy article shows specific comparisons with human references.

Use the brief for the provider decision in front of you

• Sapien vs Aaru: connect a pricing comparison to a contribution threshold.
• Sapien vs Evidenza: commission a managed study that resolves a buying-group message decision.
• Sapien vs Simile: decide how grounding and uncertainty should change the next research step.
• Sapien vs Lakmoos: decide who should own ongoing research and how to assess a private-model pilot.
• Sapien vs Qualtrics Edge: scope a decision study within a broader human and synthetic research program.

Three further comparisons help refine the assignment: Artificial Societies for buyer choice and changing social reactions, Synthetic Users for turning interview themes into comparable onboarding routes, and Fairgen for separating respondent twins, survey augmentation and new-audience coverage.

Turn the brief into a Sapien study

Our audience-building guide shows how the decision, relevant data and research connect. The useful outcome is a recommendation your team can inspect and act on, with open questions clearly identified.

Book a Sapien introduction with your alternatives, target audience and next decision. Bring a completed brief or the few sections you know; we can discuss the scope needed to turn them into research.