QuestionProAI Lens

QuestionProAI Lens is an automated pre-publish check for your survey. It scans your survey for logic errors, accessibility issues, respondent-experience risk, methodology problems, and compliance gaps, then scores the survey with a single Readiness Score. Two AI-powered checks also review question wording and overall survey coherence, surfacing additional suggestions alongside the score.

Every finding can be resolved without leaving the review panel: some findings come with a one-click fix, some with an AI-suggested rewrite you approve before applying, and others with an AI-proposed change you review first. Once you've made changes, re-run AI Lens to confirm the score has improved before you publish.

• Open your survey in the Survey Builder and click Review in the toolbar.



• You can also open AI Lens from the publish flow. If findings are still outstanding when you try to publish, you'll be offered a Publish anyway option if you'd rather not resolve them first.

AI Lens checks your survey across five categories: logic (dangling branches, unreachable questions, loops), accessibility (WCAG-related issues like empty question text or oversized matrices), respondent experience (survey length and excessive open-text burden), methodology (unbalanced scales, missing validation, missing neutral options, inconsistent scale lengths across a battery), and compliance (missing consent or privacy disclosure when personal information is being collected).

Alongside these, two AI-powered checks run automatically: one reviews individual question wording for issues like double-barrelled questions, vague terms, or overly complex phrasing; the other reviews the survey as a whole for questions that fall outside its apparent theme, or a survey name that promises something no question actually measures.

• Every survey starts at a score of 100.
• Points are subtracted for each finding, based on severity:
  • Blocker (the survey is broken for the respondent)
  • Warning (the survey runs, but data quality will suffer)
  • Advisory (a suggestion, safe to launch without acting on it)
• If your survey has even one Blocker, the score cannot go above required quality threshold, so a serious issue never gets masked by an otherwise clean score.
• A survey with no questions at all scores 0, regardless of any other findings.

The score is shown with a readiness band: Not ready (0–49), Needs work (50–74), Good (75–89), or Ready (90–100).

Only findings from the logic, accessibility, experience, methodology, and compliance checks affect the score. Findings from the two AI wording and coherence checks are advisory only and never change the score.

• Open a finding in the review panel to see what it caught and why it matters.
• Fix this: for issues with one clear correction, AI Lens shows you an exact before/after preview. Click to apply it directly, no AI involved.
• Suggest a fix: for wording-quality findings, AI Lens proposes a rewrite (which may split one question into several). Review the before/after, then apply it if you're happy with it.
• AI-proposed fix: for findings without one obvious correction, AI Lens proposes a change and shows you a preview before it's made. Nothing is applied until you confirm it.

Fixes apply immediately once confirmed and can't be undone, so review the before/after preview carefully before applying. If someone else has edited the survey since you opened the preview, the fix will be blocked rather than applied against out-of-date content.

Yes. Dismissing a finding hides it from your worklist so you can focus on what's left, and it can be restored later if you change your mind. Dismissing does not resolve the underlying issue and does not improve your Readiness Score — the score only improves once a finding is actually fixed.

Reopening AI Lens on a survey you haven't changed shows you the same result instantly, without recomputing anything. Once you edit the survey — including applying a fix — the Refresh control re-enables so you can re-run the AI checks and confirm your score. If you're collaborating with someone else on the same survey, Refresh also re-enables automatically once their edits come through.

Large surveys can occasionally take longer to review. If a check is simply taking longer than usual, AI Lens will let you know and suggest checking back shortly — your score and the standard checks are never affected. If a check genuinely fails to complete, AI Lens will tell you the AI checks couldn't run while noting the rest of your review is unaffected.

• AI Lens does not change your survey's data structure or any responses already collected — it only evaluates the survey, and edits question wording, logic, or settings when you apply a fix.
• Restoring an earlier version of your survey clears the record of previously applied fixes for that survey, since a restored draft can leave a fix pointing at a finding that no longer exists.
• Publish still requires at least one question in your survey, independent of your Readiness Score.
• Findings and applied fixes are saved with your survey, so reopening AI Lens later shows the same results you left, until you make further changes.