Strategy
Part of Market research surveys beyond the obvious
Market Research Surveys Checklist: Pass-or-Block Sign-Off
A market research surveys checklist run as a release gate: six rows, one owner each, worked pass and block examples, and when to run the gate again.
What to take away
- Treat release as a decision with a named owner, a record and a pass or block outcome.
- Run the rows in order; a row that blocks stops everything after it.
- Attach evidence to each row, so a reviewer sees the record instead of hearing assurances.
- Send a blocked row back to its owner with a named gap, then run the gate again.
- Have an independent analyst reproduce one published estimate before leaders act.
A market research surveys study is ready to release when a reviewer can see why each row passes. A completed gate supports review; it does not prove a valid survey. Adapt the rows to jurisdiction, sector, population, mode and decision risk.
One gate, six rows, six owners
Release is a sequence of rows, and each row ends in pass or block. Work through them in order, because a failed row early on makes the later rows moot. Every row has one owner who signs only that row.
| Row | Owner | Evidence to attach | Block release when |
|---|---|---|---|
| Fit | Decision owner | Decision, claim, alternatives considered | The survey cannot answer the claim |
| Coverage | Sample owner | Population, frame, sample, exclusions | Missing groups are hidden |
| Measure | Questionnaire owner | Questions, tests, versions, mode | Meaning is unstable |
| Privacy | Privacy owner | Applicable law, consent text, retention | No lawful basis is recorded |
| Supplier | Vendor owner | Standards followed, contract terms | The supplier's standards are unknown |
| Estimate | Analysis owner | Edits, weights, uncertainty, limits | The number cannot be reproduced |
A filled Fit row that passes
A Fit row names the decision, the owner and the claim the survey must support. A Fit row that passes reads like this.
Decision: launch a new pricing tier. Owner: VP Product. Reviewer sees: a claim of intent to buy, and a list of alternatives that includes a price test. Status: pass.
The row passes because the claim and the alternatives are written down. Compare the alternatives with the options in a methods mix brief, so a survey is chosen over interviews, observation or behavioral data on purpose.
A filled Coverage row that blocks
Coverage blocks release when groups the frame cannot reach are hidden from the reader. A Coverage row that blocks reads like this.
Frame: the customer list. Sample: customers who opened the invitation. Exclusion not shown: people who never bought. Claim: demand in the whole market. Status: block.
The claim reaches further than the frame, so the row goes back to the sample owner. Either the frame widens to include non-buyers, or the claim narrows to customers. The population tables used to choose the market show who is missing.
What a block does next
Blocking is not failure. It sends the study back to the row's owner with a named gap, and the gap is closed before the gate is run again.
- Write the gap in one sentence on the row.
- Name the owner who closes it.
- Attach the new evidence to the same row.
- Re-run the blocked row, then every row after it.
Rows that passed earlier stay passed unless the fix changed their evidence. A wider frame, for instance, reopens the Measure row if new groups need different wording.
Rows that cite outside rules
The Measure row also needs the scoring rules fixed in advance, because scoring written beforehand prevents post hoc stories. The U.K. Government Analysis Function's survey inclusivity and accessibility guidance links accessible formats, understandable language and question validity. Its public-sector scope does not replace testing with your own population.
For the Privacy row, record whether the GDPR, the CCPA as amended by the CPRA, or COPPA applies, even when the answer is no. The FTC's COPPA guidance says the Rule was designed to protect children under age 13, and its primary goal is to put parents in control of what is collected online.
For the Supplier row, record which professional code the supplier follows, such as AAPOR's Code of Professional Ethics and Practices or the ICC/ESOMAR code. The AAPOR standards for US teams cover the US side.
The Estimate row
An independent analyst should reproduce one published estimate from the questionnaire, raw data, edits, weights and code. If the number cannot be reproduced, the Estimate row blocks release.
Public wording of that estimate needs a check against FTC consumer research claims. Retention and deletion rules for the raw data belong in the strategy checklist. Confirm that people excluded by the frame and mode remain visible in the limitation before leaders act on the finding.
When to run the gate again
Repeat the gate after a material change in any of these:
- population or frame
- question or mode
- panel, tool or incentive
- data flow or analysis
- the claim you intend to make
Gate sign-off checklist
- Fitdecision, owner, claim and alternatives recorded
- Coverageframe, sample and exclusions shown
- Measurequestions, tests and version attached
- Privacyapplicable law and consent text recorded
- Supplierstandards and contract terms recorded
- Estimateone figure reproduced by an independent analyst
Common questions
Does a completed gate prove survey validity?
No. Qualified people must still judge method fit, coverage, measurement, execution, participant protection, analysis and interpretation.
Who should sign off?
Assign an owner to each row: the decision, the sample, the questionnaire, privacy, the supplier and the analysis. Each owner signs only the row they hold.
Do all six rows run again after a block?
Re-run the blocked row and every row after it. Earlier rows stay passed unless the fix changed their evidence.
