market-research

name	market-research
description	Produces a MARKET-RESEARCH.md (synthesized user/market evidence supporting a specific product decision — method, sample, evidence-tied findings); primes the ai + compliance + accessibility gates. Use to turn evidence into a decision before the brief.

market-research

The evidence-synthesis producer of the product domain — it turns user/market evidence you have or point to (interview notes, survey data, analytics, competitive scans) into a MARKET-RESEARCH.md: the method, the sample, findings each tied to specific evidence, and the product decision those findings support. It binds the pack's market-user-research sub_target. The bar is that every finding traces to evidence and every gap is a flagged assumption — no fabricated quotes or stats (the ai gate). Verified by Attestation: a named owner attests the synthesis is sound and owns the decision. It synthesizes evidence; it does not gather external sources (that is the research domain).

What this does / does not do

Does: resolve the market-user-research sub_target, synthesize the supplied evidence into a method + sample

• evidence-tied findings + the supported decision, flag gaps and assumptions, and write a MARKET-RESEARCH.md; write a receipt. Does not: gather/fetch external sources or run web research (the research domain does that), decide the problem/persona (that is product-brief, which consumes this), build anything, issue the gate verdict (the ai/compliance/accessibility gates do — this primes them), or invent quotes/stats (gaps are flagged, never fabricated).

When NOT to use

• The work is gathering new external evidence (web, literature, primary research) — that is the research domain; this synthesizes evidence you already have or point to.
• The work is deciding the problem/persona/success metric — that is product-brief (it uses this research).
• A current MARKET-RESEARCH.md still fits and nothing changed — re-synthesizing churns a decision contract.
• The user wants the synthesis attested or a feature built — that is the Attestation owner + the build domain.

Reference Routing

Situation	Reference
The MARKET-RESEARCH.md contract template	references/market-research-template.md
Per-sub_target must-declares · overlay · vocabulary	../acceptance/sub-targets.md · ../acceptance/overlay.md · ../vocabulary.md
External evidence gathering (out of scope here)	the research domain pack

Inputs

• The user/market evidence to synthesize — interview notes, survey results, product analytics, support tickets, competitive scans — provided or pointed to.
• The product domain pack (../vocabulary.md, ../acceptance/) + state/intake.json.
• The PRODUCT-BRIEF.md if one exists — the decision context the research informs.

How to do it

1. Frame the decision: state the product decision this research must inform (which segment, which problem, build-or-not). Research that supports no decision is incomplete (it gets flagged, not shipped).
2. Declare method + sample: what the evidence is and how it was gathered (interviews / survey / analytics / competitive scan), and the sample — who, how many, how representative, over what window. A finding from an unstated or unrepresentative sample is caveated.
3. Synthesize findings: cluster the evidence into findings, each tied to a specific quote, datapoint, or source. A finding that cannot point to evidence is dropped or demoted to an open question.
4. Flag gaps + assumptions: where the evidence is thin or absent, say so as a flagged assumption (record_type: alert); do not fill the gap with a fabricated quote or stat (the ai gate).
5. State the supported decision: the decision the findings support, with the confidence the evidence warrants — and what would change it. Substantiate any market/competitive claim (compliance); keep it plain enough for a stakeholder to act on (accessibility).
6. Write MARKET-RESEARCH.md from references/market-research-template.md; name the attestation owner.

Gate (before marking complete)

• Method + sample (source, size, representativeness) are declared before the findings.
• Every finding ties to a specific quote / datapoint / source.
• Gaps and unvalidated claims are flagged assumptions; no fabricated quotes/stats (the ai gate).
• A specific product decision the evidence supports is named.
• Market/competitive claims are substantiated, not overstated (compliance); stakeholder-actionable (accessibility).
• An attestation owner is named.

A gate failure is fixed before lock, not waived — these mirror the ai/compliance/accessibility gates.

Output

• MARKET-RESEARCH.md at the product root (or product/MARKET-RESEARCH.md) — persistence: Spec-Anchored. Decision framed · method + sample · evidence-tied findings · the supported decision + confidence · flagged assumptions/gaps · attestation owner.
• receipt (<workspace>/receipts/market-research-receipt-<ts>.json): base schema + sub_target, method, sample, findings, decision_supported, assumptions_flagged, market_research_path, attested_by, gate_checks. persistence: Spec-Anchored.

Proactive Triggers

• A finding with no evidence behind it: tie it to a quote/datapoint or demote it to an open question — an unsupported finding is the fabrication failure mode.
• A gap filled with an invented quote/stat: replace it with a flagged assumption (the ai gate); thin evidence is stated as thin, not papered over.
• Research that supports no decision: name the decision it informs, or flag it incomplete — insight without a decision is noise.
• An unrepresentative sample stated as general truth: caveat the finding to the sample it came from.

Completion Criteria

Done when: a MARKET-RESEARCH.md exists with method + sample declared, findings each tied to evidence, gaps flagged as assumptions, a named supported decision, and a named attestation owner; a receipt records the gate checks.

Not done if: method/sample is unstated; a finding floats free of evidence; a gap was filled with a fabricated quote/stat; no decision is named; claims are overstated; or no owner is named.

Next step: product-brief uses the synthesis to frame the problem/persona/success metric; the named owner attests the research (Attestation); the research domain handles any external evidence still needed.

Related Skills

• product-brief: consumes this research to decide the problem/persona/success metric. market-research is the evidence upstream; product-brief is the decision. Run research first when evidence is the gap.
• research (domain): gathers/audits external sources (literature, web, primary). market-research synthesizes product evidence into a decision — different jobs; route external gathering there.
• prioritization / roadmap: downstream — the decision this research supports becomes bets to rank and sequence.