Substandard work, confidently delivered
Research design, questionnaires, fieldwork and reporting are flawed — but the deck still looks finished. The error surfaces after the decision has been made.
Capability 05
Quality Assurance
Speed only counts when the data can be trusted.
Back-checking on quantitative work, 100% editing of completed interviews, and verification of qualitative recruits against the approved screener — with issues escalated during fieldwork, not after.
The problem it solves
Speed and quality fail together.
When acquisition runs fast, errors enter quietly: an interview that was not fully administered, a respondent who does not match the screener, a section completed from memory. The deck still looks finished, and the error only surfaces once the decision has been taken.
Catching that after fieldwork closes is expensive and usually pointless — the remedy is another wave, by which time the moment has passed. So the checks run while the work is live, and anything that threatens the data is escalated during fieldwork rather than reported after it.
Insight that arrives too late
Timelines slip. The answer lands after the moment it was commissioned to inform has already passed.
What's included
Six checks, run as standard.
These are not an upgrade or a premium tier. They are the protocol every study runs under, including the ones where we only supply the field.
-
Back-checking
Completed quantitative interviews are back-checked, under a protocol owned by our head of quality assurance.
-
100% interview editing
Every completed interview is edited — all of them, not a sample of them.
-
Respondent verification
Qualitative recruits are verified against the approved screener before they are accepted onto the study.
-
Field monitoring
The standing instruction across the field network is that potential issues reach the project manager immediately.
-
Progress reporting
Progress reports are kept current for project managers while the wave is still open.
-
Data validation
Data processing, input, validation and control sit with our head of data analytics — over twenty years of that work specifically.
How it runs
Checks that run during the work, not after it.
Quality assurance that reports at the end is an audit. This is designed to change the fieldwork while the fieldwork can still be changed.
-
01
Outsmart
Set the standard before the field opens
Screener, questionnaire and back-check criteria are agreed as one set, so what counts as a valid interview is not decided afterwards.
-
02
Outpace
Verify the recruit
Qualitative respondents are checked against the approved screener before they take part.
-
03
Outpace
Back-check the quantitative work
Completed interviews are back-checked while the wave is live, so a problem interviewer is found in week one rather than at the debrief.
-
04
Outpace
Edit every interview
100% of completed interviews are edited before the data goes anywhere near analysis.
-
05
Outperform
Report while it still matters
Progress reports stay current for the project manager, and anything that threatens the data is escalated during fieldwork.
Related
Proof, once it can be published.
The slots below are built and wired. Nothing appears in them until Stage 3 supplies the engagement and the client agrees to it being named.
Related case study
What the back-checks found, what was rejected and re-run, and what the corrected data changed about the conclusion.
Client approval required
Related insight
Editorial writing on how to tell whether a supplier's quality process is real, and which questions expose it quickly.
Awaiting first article
The rest of the chain