01Overview
Suggestibility is the vulnerability of memory and report to implantation through leading questions, repeated suggestions, and authoritative framing. Users "remember" events that were proposed, rate pain that was labelled, and prefer options that were primed — then experience those suggestions as endogenous preference.
Design research, surveys, onboarding defaults, and AI copilots all suggest. Suggestibility is not bad faith from users — it is how memory works. The ethical and methodological line runs between scaffolding genuine recall and manufacturing false consensus. Products that suggest heavily produce users who defend implanted narratives as insight.
02Detailed explanation
Suggestion enters product work through familiar channels:
- Moderator: "Was checkout confusing?" — participants remember confusion.
- Survey scale anchored "how frustrated were you" — frustration remembered.
- Default pre-selected plan — later recalled as chosen.
- AI summary inserting plausible detail user did not say — accepted in next session.
Reduce suggestion in research when seeking discovery; use openly when nudging beneficial behaviour — but label influence and measure without leading where stakes are high.
03Why it exists
Memory is reconstructive — gaps filled with plausible suggested detail, especially from authority or repeated exposure.
Growth and research pressure favours leading instruments that produce clear answers — suggestibility makes answers clear, not true.
The answer you suggested is the answer you will get — and later, the answer they will remember.
04Effects on users
Users remember agreeing to terms suggested by pre-checked framing — suggestibility plus default effect.
Community lore incorporates suggested interpretations from influential posts — social suggestion at scale.
05Effects on designers & teams
Teams lead without documenting lead:
- Biased survey wording. Confirmation disguised as feedback.
- Interview scripts with embedded conclusions. Synthesis writes itself.
- AI note-takers hallucinating quotes. Becomes research record.
- Onboarding copy suggesting problems. "Struggling with X?" creates X memory.
06Practical takeaways
- Use open-ended prompts first. Suggest only after free recall.
- Neutral language in research. "Tell me about checkout" not "how bad was checkout."
- Audit defaults as suggestions. Track opt-out, not only opt-in.
- Validate AI-generated summaries. Human check before repository.
- Separate priming from preference tests. Different sessions or counterbalance.
- Document suggestion in report. What we asked influences what we heard.
07Design examples
Confusing by suggestion
Study asks "how confusing was nav?" Post-study, participants recall confusion. Re-run with neutral prompt; confusion themes drop — suggestibility drove prior synthesis.
Remembered as choice
Users interviewed about plan selection describe weighing options. Logs show default accept. Suggestibility plus default — false agency narrative.
Quote that wasn't
Auto-transcript adds "I hate billing." Stakeholder cites in roadmap. Video review shows user never said it — suggestion via faulty tool became memory for org.
Frustration scale
CSAT follow-up assumes frustration. Scores and verbatim align. Neutral cohort study shows lower frustration — question suggested affect.
08Ethical risks
Suggesting false memories in legal, medical, or safety contexts — through copy or AI — is harm with accountability.
Research suggestibility that validates harmful features exploits users' reconstructive memory for internal politics.
Self-test: Where does your product or research script suggest the answer before the user supplies it — and would answers change without the suggestion?
10Suggested reading
Suggested reading is temporarily unavailable. Please check back later.