Synthetic Research for Agencies: Methods & Creative Tests
Use synthetic research to compare creative routes, understand buyer reactions and recommend the next edit. Connect the client brief, evidence and approval decision.
Synthetic research for agencies compares simulated audience reactions to creative routes before the team recommends a revision. A useful study links the material shown, the audience’s interpretation and the exact edit to develop. Use it to challenge the preferred idea, explain buying-group differences and prepare a client recommendation with clear evidence. When pricing client work, use our AI market research tools comparison to identify the setup, recruitment and interpretation your agency would still handle.
Example: two routes for a commuter bag
Morrow is a hypothetical commuter-bag brand working with an agency. Its bag has a removable work organizer. The client favors an emotional route; creative worries the product will be unclear. The team must choose what to develop before production approval. No audience findings have been collected.
Route A opens with “Leave work behind” and shows the closed bag beside a restaurant table. Route B opens with “Lift out work. Make room for the evening” and shows the organizer being removed. Both carry the same product name and a line explaining that the commuter bag has a removable work organizer.
The disagreement is useful. A leads with an occasion and an emotional benefit. B makes the mechanism visible. The research question is whether the intended audience understands the product and sees a reason to consider it, and which part of either route needs development.
Name the development decision in the brief. Approval may release a revised storyboard, not the whole campaign. Early appeal does not approve later casting, editing, media or product-performance claims.
Also establish what remains fixed: the bag, its verified features, the intended audience and the next development commitment. If the client changes the product halfway through, record a new offer version rather than combining responses to different bags.
Related guides: a brief template for synthetic studies.
Both routes need the same truthful product
Before involving an audience, the agency and client need to agree what the organizer actually does. Where can it be placed after removal? Which items does it hold? What stays in the bag? Use verified product information and suitable demonstrations to answer those questions.
The closed-bag restaurant scene does not show the removable mechanism. That is visible in the creative itself; it does not require a research result. The team can decide whether the omission is intentional or whether the route needs one additional moment before testing.
Avoid solving the problem by adding an unsupported benefit. “No more back pain” needs different evidence from a demonstration of an organizer being lifted out. Audience enthusiasm cannot establish physical comfort or durability. Keep the research about the promise the product can support.
Match the level of finish. A polished film against a rough sentence mixes idea and execution quality. A and B change headline and image together, so compare them as complete routes, not isolated headline effects.
Ask for an interpretation before a verdict
Randomly assign target respondents to one route first in the intended setting. Ask what is offered, what use changes and which detail explains that view. Then explore routine fit, doubts and what they need to see next.
Keep those questions neutral. “Did the removable organizer look useful?” teaches the feature to someone who may not have noticed it. AAPOR's questionnaire guidance recommends avoiding leading wording and checking how questions and their order influence answers. Our message testing guide shows how to distinguish accurate understanding from enthusiasm for an imagined offer.
The relevant audience is defined by the situation, not just a label such as urban professional. Someone who leaves the organizer at a fixed desk has a different practical question from someone who works wherever a seat is available. Include the routines that could make the proposition useful or unsuitable. A memorable response from an irrelevant routine should not decide the route.
Showing the alternative later can reveal a relative preference, but keep the first reaction separately. B may explain the organizer and make A easier to understand afterward. That later understanding does not show what A communicated on its own.
Interpretation determines the next edit
Different reactions would lead the agency to different work. If A is read as a restaurant bag with no work function, the route needs the product connection made visible. If B is understood but leaves people asking where the removed organizer goes, the next task is to show a credible transition, provided the product can support it.
Someone who already carries a separate work pouch might understand both routes and still prefer their current arrangement. More explanation would miss that objection. Investigate what the complete Morrow bag improves, and whether the proposed audience actually needs it.
These are branches to investigate, not findings to place in the pitch. Keep contradictory accounts and identify which routines they describe. A route can be promising for one buying situation while remaining a poor fit for another.
When the team revises the storyboard, test the changed version. Do not combine the liked line from A with the understood demonstration from B and report the combination as a tested winner. The new sequence may explain the bag differently or introduce another expectation.
Make the client recommendation easy to challenge
Show the exact creative, supporting interpretation and proposed edit. Include response source and audience definition. Name the opposing account that could change the choice instead of dismissing it with a favorable average.
Agree the report with its users. Strategy may need to defend the choice, research to check the evidence and the client to approve development. Include the necessary presentation and revision work in the scope.
A short approval sheet can keep the discussion connected to the work:
- Creative reviewed: The exact route and version shown.
- Recommended next edit: The change to develop and the interpretation supporting it.
- Response source: Whether the supporting responses are modeled or came from people, with the relevant audience and question.
- Audience difference: The routine or group whose opposing account would change the edit.
- Approval sought: The next commitment and the person who can authorize it.
- Evidence still needed: The remaining question, who will resolve it and when.
For Morrow, keep the two work routines visible in that readout. If leaving the organizer at a desk supports one recommendation while carrying it between locations supports another, the client needs to see that difference before approving the next storyboard.
For simulated research, say that the responses came from a simulated audience. Do not turn a generated comment into a customer quotation or imply that people handled the bag. If the client needs to know whether the organizer works during a real commute, that question needs suitable evidence with the physical product.
Keep creative approval separate from product substantiation. A client may approve a clearer storyboard while still withholding a comfort claim until the evidence exists. It is useful research if it narrows the production decision and reveals that dependency before money is committed.
The same discipline applies in a competitive pitch. Research can support the agency's argument for a route, but it should not be described as a completed client engagement or a guaranteed campaign result. If any output is only an illustration, say so plainly.
The recommendation should name the next creative move
At Sapien, our team builds data-grounded simulated audiences, runs the agreed study and helps interpret the findings. We can compare concepts and creative, examine differences between audience groups and explore individual responses through interactive findings and follow-up questions. Our market research services guide explains the managed engagement.
For Morrow's decision, the agency would bring the client-approved product facts, the actual creative routes, the intended routines and the choice the research must inform. Agree the audience grounding and validation relevant to the task. The agency continues to own the creative work and the client relationship; the client must confirm what the product can truthfully promise.
Once a route is clear enough to progress, assess the finished execution in its actual campaign context. Our ad testing guide connects creative findings to the elements a team can change and the later campaign test. A simulated preference is not a prediction of sales, and an approved storyboard is not proof that the media execution works.
For Morrow, a useful recommendation would identify the moment where the organizer needs to be shown, the interpretation that supports the edit and the proof still needed before production. Book a demo to discuss how Sapien could structure that comparison for the brief your agency is developing.
Related guides: synthetic vs human respondents; ad testing platforms.
Questions to ask a research provider
- Who exactly will answer, and how are they recruited or built?
- What will you hand back: raw records, tables, a report or a recommendation?
- What validation have you published, for which tasks, and where can we inspect it?
- What does the whole project cost, including recruitment and analysis, and how long does it take?
- Who does the work on our side, and who on yours?
- What would make you tell us the study cannot answer our question?
