User Research Prompts for UX and Product Teams
16 ChatGPT prompts for user interviews, usability testing, surveys, journey mapping, research synthesis, and the UX research that grounds product decisions in user reality.
In short: This page contains 16 copy-paste ready prompts, organized into 4 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Drafted with AI writing assistance and checked against our editorial policy.
Interview + Discovery
4 promptsUser Interview Script
1/16✨ What it does
ChatGPT writes a user interview script for product [describe] with research goal [discovery/validation], covering 5-minute rapport, 10-minute context, 20-minute core questions, current workflow, friction, and a close, all non-leading. Ask the questions as written; do not steer the answer you want.
Write user interview script. Product: [describe]. Research goal: [discovery/validation]. Include: opening (5 min rapport), context questions (10 min), core questions (20 min), current workflow, friction points, closing. Non-leading questions.
Pro tip: User interviews: past behavior > future opinion. "Tell me about the last time you did X" > "Would you use X?" People lie about future; truth about past.
Contextual Inquiry Plan
2/16✨ What it does
ChatGPT plans a contextual inquiry for user [describe] on task [describe], covering in-environment setup, think-aloud prompts, notes, follow-ups, respect and ethics, and a debrief. Observe in their space, then debrief before you leave.
Plan contextual inquiry. User: [describe]. Task to observe: [describe]. Include: observation setup (in their environment), "think-aloud" prompts, note-taking approach, follow-up questions, respect + ethics, debrief.
Pro tip: Contextual inquiry: watch user do task in their environment. Reveals assumptions, workarounds, emotional responses. Lab setting sanitizes. Field reveals reality.
Diary Study Design
3/16✨ What it does
ChatGPT designs a diary study on topic [describe] lasting [1-2 weeks], with participant selection, daily or weekly entry format, per-entry prompts, incentives, analysis, and the longitudinal insights you should expect. Recruit the participants, then send the first prompt on day one.
Design diary study. Topic: [describe]. Duration: [1-2 weeks]. Include: participant selection, daily/weekly entries format, prompts per entry, incentive structure, analysis approach, longitudinal insights expected.
Pro tip: Diary studies reveal patterns across time. Single interviews = snapshot. Diaries = movie. Mood, context, frequency emerge. 1-2 weeks ideal; longer = drop-off.
User Research Plan
4/16✨ What it does
ChatGPT writes a user research plan for project [describe] over timeline [X weeks], covering questions, a mix of interviews, surveys, and usability, recruiting, per-method timeline, deliverables, and team roles. Assign owners to each method, then start recruiting for your study.
Write user research plan. Project: [describe]. Timeline: [X weeks]. Include: research questions, methods mix (interviews, surveys, usability), participant recruiting, timeline per method, deliverables, team responsibilities.
Pro tip: Research plan = 1-2 pager. Research questions first; methods follow. Plan shared with stakeholders before starting — alignment prevents "why did you do that" later.
Prompts get you started. Tutorials level you up.
A growing library of hands-on, step-by-step AI tutorials. New tutorials are added regularly.
Usability + Testing
4 promptsUsability Test Script
5/16✨ What it does
ChatGPT designs a usability test for product [describe] with tasks [list 3-5], covering goal-based tasks rather than step lists, success criteria, think-aloud prompts, an observer note template, post-task questions, and a SUS adaptation. Run the tasks as goals, then score SUS after each session you host.
Design usability test. Product: [describe]. Tasks: [list 3-5]. Include: task descriptions (goals, not steps), success criteria, think-aloud prompts, observer note template, post-task questions, SUS score adaptation.
Pro tip: Usability test rule: tasks = goals, not steps. "Find a red dress under $50" (goal). Not "click Shop > Dresses > Red > Filter price." Goal reveals paths; steps reveal nothing.
Moderated vs Unmoderated Decision
6/16✨ What it does
ChatGPT helps you choose moderated vs unmoderated testing for goal [describe] and participants [describe], weighing depth and probing vs speed, scale, and cost, plus UserTesting or Maze and a hybrid option. Pick the mode that matches your timeline, then book the tool.
Moderated vs unmoderated test decision. Goal: [describe]. Participants: [describe]. Include: moderated benefits (depth, probing), unmoderated benefits (speed, scale, cost), tool options (UserTesting, Maze), hybrid approach.
Pro tip: Moderated: 5-10 participants, deep insights, qualitative. Unmoderated: 20-100 participants, quantitative + fast, less depth. Start unmoderated for quick signals; moderated for deeper issues.
Usability Issues Prioritization
7/16✨ What it does
ChatGPT prioritizes usability issues [list] by severity 0-4, frequency, core-task impact, and fix effort, then plots a severity-by-frequency matrix. Fix the high-severity, high-frequency cells on your list first.
Prioritize usability issues from research. Issues: [list]. Include: severity (0-4 scale: cosmetic to catastrophic), frequency (how many users hit), impact on core tasks, effort to fix. Matrix: severity × frequency for priority.
Pro tip: Nielsen severity scale: 0 (non-issue), 1 (cosmetic), 2 (minor), 3 (major), 4 (catastrophic). Focus on 3-4 first. Perfect usability unrealistic; fix big issues first.
A/B Test Design
8/16✨ What it does
ChatGPT designs an A/B test from hypothesis [describe] and metric [describe], with control vs variant, sample-size power calc, duration, segmentation, significance threshold, and decision criteria. Freeze the metric before you launch, then wait for the sample.
Design A/B test. Hypothesis: [describe]. Metric: [describe]. Include: variants (A = control, B = variant), sample size (power calc), duration, segmentation, significance threshold, decision criteria.
Pro tip: A/B test mistakes: too small sample, too short duration, too many tests in parallel, no hypothesis. Need proper sample size (usually 1000+ per variant) + clear hypothesis + 1-2 week minimum.
Journey + Personas
4 promptsJourney Map
9/16✨ What it does
ChatGPT builds a user journey map for user [persona] in scenario [describe], covering awareness through retention, actions, thoughts and feelings, pains, delights, touchpoints, and opportunities. Mark the worst pain, then take that opportunity to your team.
Build user journey map. User: [persona]. Scenario: [describe]. Include: stages (awareness, consideration, purchase, use, retention), actions per stage, thoughts/feelings per stage, pain points, moments of delight, touchpoints, opportunities.
Pro tip: Journey maps reveal end-to-end experience. Individual screens optimized but journey broken = users lost. Map end-to-end; find where users drop or suffer. Emotional lows = opportunities.
Service Blueprint
10/16✨ What it does
ChatGPT builds a service blueprint for [describe], covering customer actions, front-stage, back-stage, support processes, touchpoints, physical evidence, and service-time metrics. Walk one real customer through it, then fix the first back-stage break you find.
Build service blueprint. Service: [describe]. Include: customer actions, front-stage (customer sees), back-stage (employee actions), support processes, touchpoints, physical evidence, service time metrics.
Pro tip: Service blueprints extend journey maps to include back-stage. Reveals operational disconnects. "Customer sees fast checkout; back-stage hand-coded mess" = fragility waiting to break.
Persona Research Synthesis
11/16✨ What it does
ChatGPT synthesizes 2-3 personas from data sources [describe], with demographics, goals, pains, behavior, interview quotes, and anti-goals. Keep a real quote on each card, then drop any persona you cannot evidence.
Synthesize personas from research. Data sources: [describe]. Include: 2-3 persona profiles, demographic data, goals + motivations, pain points, behavioral patterns, quotes from interviews, anti-goals (what they don't want).
Pro tip: Personas grounded in research >> fabricated ones. Real quotes > made-up ones. 2-3 personas optimal (not 7). Anti-goals ("don't want salesy emails") often more actionable than goals.
Mental Model Diagram
12/16✨ What it does
ChatGPT creates a mental model for user [describe] in domain [describe], covering how they conceptualize it, key concepts, relationships, metaphors, misconceptions, and interface implications. Redesign the confusing screen to match their model, not your own.
Create mental model for user. User: [describe]. Domain: [describe]. Include: how user conceptualizes domain, key concepts, relationships, metaphors used, misconceptions, implications for product interface.
Pro tip: Mental model ≠ reality. User thinks inbox is "folder for important mail" but technically it's database query. Design to mental model, not backend reality. Users win when product matches mental model.
Like these prompts? There are full tutorials behind them.
Learn the workflows, not just the prompts. Easy-to-follow tutorials inside AI Academy, and the library keeps growing.
Synthesis + Delivery
4 promptsResearch Synthesis Workshop
13/16✨ What it does
ChatGPT plans a 2-4 hour research synthesis workshop for team [describe] using data [describe], with affinity mapping, themes, insight ranking, prioritization, and action items. Block the hours, then leave with ranked actions your team will own.
Plan research synthesis workshop. Team: [describe]. Data: [describe]. Include: affinity mapping, theme identification, insight ranking, prioritization, action items, 2-4 hour workshop structure.
Pro tip: Synthesis in team > synthesis alone. Different perspectives reveal patterns. Sticky notes + wall + 2 hours = themes emerge. Remote: Miro/FigJam work equivalent.
Insight Generation
14/16✨ What it does
ChatGPT generates insights from data [describe], separating surface observations from deep patterns, unexpected findings, the so-what, business implications, and next steps. Keep only insights with a next step, then assign each one on your team.
Generate insights from data. Data: [describe]. Include: surface observations vs deep insights, patterns across users, unexpected findings, "so what" articulation, business implications, actionable next steps.
Pro tip: Insight formula: [user group] [does X] [because Y] [because Z motivation]. "We observed X" = observation. "Users do X because Y" = insight. Push 3 "why" layers deep.
Research Findings Report
15/16✨ What it does
ChatGPT writes a research findings report for audience [stakeholders] at length [short], with executive summary, methods, 3-5 findings, quotes and evidence, recommendations, limitations, and appendix, told as a story rather than a data dump. Lead with the summary, then attach the appendix for your readers.
Write research findings report. Audience: [stakeholders]. Length: [short]. Include: executive summary, methodology, key findings (3-5), quotes + evidence, recommendations, limitations, appendix. Story arc (not data dump).
Pro tip: Research reports as stories. "We studied X. Here's what we found. Here's what it means. Here's what to do." Data dumps fail; narratives land. Tell the "so what."
Research Presentation
16/16✨ What it does
ChatGPT designs a research presentation for audience [executives/team] lasting [20 min], with a surprising opening, 3-5 insights, user quotes, clear recommendations, and supporting data, story over slides. Open with the surprise, then cut any slide that is not a recommendation you will defend.
Design research presentation. Audience: [executives/team]. Duration: [20 min]. Include: compelling opening (surprising finding), key insights (3-5), user quotes, clear recommendations, supporting data. Story > slides.
Pro tip: Research presentations: open with surprising insight (hook). Key findings middle. Recommendations end. Show user quotes (emotion). Data tables bore; stories stick. One slide = one idea.
Free tool
Prompt Optimizer
Turn a rough idea into a structured, professional AI prompt.
Frequently Asked Questions
Prompts are the starting line. Tutorials are the finish.
A growing library of hands-on tutorials on ChatGPT, Claude, Midjourney, and many other AI tools. New tutorials are added regularly.
7-day free trial. Cancel anytime.
Related guides