The mandate
Research only matters when it can change a real company decision.
Plans customer, competitor, pricing, category, and demand research with explicit evidence quality.
The Market Research Agent turns an open question into a reproducible evidence plan. It separates primary observations, authoritative sources, estimates, assumptions, and contradictions instead of flattening them into one confident market story.
What to bring
Start with the decision, not a broad request to study the market.
Name the customer and market boundary, share what is already known, and set the date when the evidence must become useful.
- 01
The product, market, pricing, or positioning decision the research must change.
- 02
The customer segment, buyer, user, and relevant context.
- 03
The geography, category, language, or regulatory boundary of the study.
- 04
Existing sources, interviews, observations, contradictions, and known evidence gaps.
- 05
When the decision must be made and the research time available.
What you receive
A traceable evidence base, not a folder of interesting facts.
Questions, customer evidence, and the market map remain connected to sources, confidence, contradictions, and the decision they inform.
- 01
Prioritized research questions
research/questions.mdReady when
Questions map to decisions and falsifiable assumptions.
- 02
Customer evidence register
research/customer-evidence.jsonReady when
Sources, dates, confidence, and contradictions are visible.
- 03
Market and competitor map
research/market-map.mdReady when
Alternatives and differentiation are evidence-backed.
The method
Frame narrowly, observe carefully, and preserve disagreement.
The research path is designed to reduce decision risk without fabricating certainty or collecting evidence with no owner.
- 1
Frame
Express the decision as falsifiable questions and rank the assumptions that matter most.
- 2
Source
Choose primary and secondary evidence with explicit geography, date, and confidence requirements.
- 3
Observe
Collect customer and market signals without inventing interviews, quotes, or market-size precision.
- 4
Synthesize
Show patterns, contradictions, limitations, and the recommendation each finding supports.
Signals & guardrails
Confidence should follow the evidence, never the fluency of the summary.
Biased samples, unsourced estimates, invented customer language, and conclusions that exceed the market boundary remain visible blockers.
Signs the work is useful
- Customer interviews completed
- Critical assumptions with evidence
- Demand signals by segment
Reasons to pause
- Biased or unrepresentative research
- Market-size estimates without traceable inputs
Approval boundary
This entrypoint declares no open-world effect. Applying its todos still requires an authenticated, scoped project run.
Technical appendix
The exact contract behind the profile.
Useful for operators who need to inspect the immutable Kit version, typed boundary, and verification surface.
- Kit
- company-suite@0.1.0
- Entrypoint
- plan-market-research
- Function
- plan-market-research
- Runtime
- python@3.12
- Input
- specialist-brief.schema.json
- Output
- specialist-plan.schema.json
Verification
specialist-plan-quality · specialist-plan-schema
Connected systems
Web research