Prompt Library

ChatGPT Prompts for UX Designers

30 copy-paste prompts

Thirty structured, copy-paste prompts that help you plan research, build personas, write wireframe briefs, run usability evaluations, sharpen critiques, and document design systems faster.

In short: This page contains 30 copy-paste ready prompts, organized into 6 categories with a description and pro tip for each. The first 15 prompts are free instantly โ€” no signup needed. Hand-curated and tested by the AI Academy team.

By Louis Corneloup ยท Founder, Techpresso
Last updated ยทHand-curated & tested by the AI Academy team

User Research

5 prompts

Discovery Interview Guide

1/30

<context> I am a UX designer planning generative research for [PRODUCT], used by [USER] to accomplish [FLOW]. We are in early discovery and want to understand needs, mental models, and pain points without leading the participant. </context> <task> 1. Write a 45-minute semi-structured interview guide with a warm-up, 4-6 thematic sections, and a wrap-up. 2. For each section, provide 3-4 open, non-leading questions plus 2 optional follow-up probes. 3. Map every question to the underlying research objective it answers. 4. Flag any question that risks bias and suggest a neutral rewrite. 5. End with a participant consent and recording-permission script. </task>

A complete, objective-mapped interview guide ready to run with real users.

๐Ÿ’ก

Pro tip: After the first draft, paste in your actual research questions and ask ChatGPT to re-order sections by logical flow and cut redundant questions.

Screener Survey Builder

2/30

<context> I need to recruit qualified participants for research on [PRODUCT]. The ideal [USER] regularly performs [FLOW]; I must screen out professionals, competitors, and people who do not match the behavior we care about. </context> <task> 1. Draft a screener survey of 8-12 questions covering demographics, behavior frequency, and disqualifying criteria. 2. For each question, mark which answer options qualify, disqualify, or quota-balance the sample. 3. Use behavioral questions (what they did) over attitudinal ones (what they think) wherever possible. 4. Recommend target quotas across the key segments. 5. List 3 fraud-detection or attention-check questions to filter low-quality respondents. </task>

A recruitment screener with qualify/disqualify logic and anti-fraud checks.

๐Ÿ’ก

Pro tip: Ask ChatGPT to output the screener as a table with a Logic column so you can paste it straight into a survey tool.

Research Plan One-Pager

3/30

<context> I must align stakeholders before researching [PRODUCT] with [USER]. Leadership wants to know why, what, and how this research will inform [FLOW] decisions. </context> <task> 1. Write a one-page research plan with background, problem statement, and 3-5 research questions. 2. Recommend the most appropriate method (interviews, diary study, survey, usability test) and justify the choice. 3. Define participant criteria, sample size, timeline, and required tools. 4. State the decisions this research will unblock and the risk of not doing it. 5. List measurable success criteria for the study. </task>

A stakeholder-ready research plan that justifies method and ties research to decisions.

๐Ÿ’ก

Pro tip: Tell ChatGPT your team's seniority and time budget so it scopes the method realistically instead of proposing an ideal-world study.

Interview Note Synthesizer

4/30

<context> I just finished interviews about [PRODUCT] with several [USER] participants. I will paste raw notes or transcripts. I need to find patterns about [FLOW] without inventing findings. </context> <task> 1. Wait for me to paste the notes, then cluster observations into 5-8 themes. 2. For each theme, give a short name, a one-line summary, supporting verbatim quotes, and how many participants mentioned it. 3. Separate observations (what was said) from interpretations (what it might mean) and label each. 4. Surface contradictions or outliers worth investigating. 5. Propose 3 prioritized opportunity areas grounded only in the data. </task>

Thematic synthesis of raw interview notes with quotes and frequency counts.

๐Ÿ’ก

Pro tip: Paste notes in batches and ask ChatGPT to keep a running theme list so it does not lose earlier patterns across long transcripts.

Survey Question Designer

5/30

<context> I am writing a survey to measure attitudes and behaviors of [USER] toward [PRODUCT] and the [FLOW] experience. I want valid, unbiased questions and clean scales. </context> <task> 1. Draft 12-15 survey questions mixing single-select, multi-select, and Likert scales. 2. For every Likert item, use a balanced, labeled scale and avoid double-barreled wording. 3. Add 2 open-ended questions for qualitative depth. 4. Order questions to reduce priming and place sensitive items last. 5. Flag any leading, loaded, or ambiguous question and provide a corrected version. </task>

A bias-checked survey with balanced scales and proper question ordering.

๐Ÿ’ก

Pro tip: Ask ChatGPT to estimate completion time so you can trim the survey to under 5 minutes and protect response rates.

Prompts get you started. Tutorials level you up.

A growing library of 300+ hands-on AI tutorials. New tutorials added every week.

Start 7-Day Free Trial

Personas & Journey Maps

5 prompts

Evidence-Based Persona

6/30

<context> I have research insights about [USER] who use [PRODUCT]. I will paste key findings. I want a behavioral persona grounded in data, not demographic stereotypes. </context> <task> 1. Wait for my research findings, then build a single primary persona. 2. Include goals, motivations, behaviors, frustrations, context of use, and a representative quote. 3. Anchor every persona attribute to a specific finding I provided and note where evidence is thin. 4. Avoid demographic filler that does not affect design decisions. 5. Add a short 'design implications' section translating the persona into 3 actionable UX priorities. </task>

A behavioral persona tied to evidence with explicit design implications.

๐Ÿ’ก

Pro tip: Tell ChatGPT to mark any attribute it inferred without data as ASSUMPTION so you can validate it before sharing the persona.

Current-State Journey Map

7/30

<context> I am mapping how [USER] currently completes [FLOW] with [PRODUCT], including the messy parts before and after they touch our product. </context> <task> 1. Break [FLOW] into 5-8 sequential stages from trigger to outcome. 2. For each stage, describe user actions, thoughts, emotions, touchpoints, and pain points. 3. Plot an emotional curve and call out the lowest moments. 4. Identify the 3 highest-impact friction points and the evidence behind each. 5. Suggest one opportunity per top friction point. </task>

A stage-by-stage current-state journey map with an emotional curve and opportunities.

๐Ÿ’ก

Pro tip: Ask ChatGPT to format the map as a markdown table with one column per stage so it drops cleanly into FigJam or a doc.

Future-State Journey Vision

8/30

<context> I want to design an improved experience for [USER] completing [FLOW] in [PRODUCT]. I have the current-state pain points and want a credible future-state vision. </context> <task> 1. Re-imagine each stage of [FLOW] showing the ideal user actions, emotions, and touchpoints. 2. For each improvement, name the friction it removes and the design move that enables it. 3. Keep the vision realistic for a mid-size product team, not science fiction. 4. Highlight 3 quick wins versus 2 larger bets. 5. Note assumptions that must be validated before building. </task>

A future-state journey that pairs each improvement with a concrete design move.

๐Ÿ’ก

Pro tip: Provide your tech and resourcing constraints up front so ChatGPT keeps the future state buildable rather than aspirational.

Empathy Map Generator

9/30

<context> I need an empathy map for [USER] interacting with [PRODUCT] during [FLOW] to align my team on the user's perspective. </context> <task> 1. Produce a four-quadrant empathy map: Says, Thinks, Does, Feels. 2. Populate each quadrant with 4-6 specific, research-grounded entries. 3. Add a Pains and a Gains section beneath the quadrants. 4. Flag any entry that is an assumption versus an observed behavior. 5. End with 3 design questions the empathy map raises. </task>

A populated empathy map separating observed behavior from assumptions.

๐Ÿ’ก

Pro tip: Paste two or three real user quotes first so ChatGPT grounds the Says and Thinks quadrants in authentic language.

Persona Pain-to-Feature Map

10/30

<context> I have a persona for [USER] of [PRODUCT] with known pains around [FLOW]. I want to connect those pains to potential features without over-promising. </context> <task> 1. List the persona's top 5 pains in priority order. 2. For each pain, propose 1-2 feature or design ideas that directly address it. 3. Rate each idea on user impact and rough effort (high/medium/low). 4. Recommend which ideas to validate first and how. 5. Flag any feature that solves a problem the persona does not actually have. </task>

A traceable map from persona pains to prioritized feature ideas.

๐Ÿ’ก

Pro tip: Ask ChatGPT to present the result as an impact-versus-effort table so you can sort candidates into a quick prioritization grid.

Wireframe & IA Briefs

5 prompts

Wireframe Brief Writer

11/30

<context> I am about to wireframe a screen in [PRODUCT] where [USER] performs [FLOW]. I want a clear brief before I open the design tool so the layout serves the task. </context> <task> 1. Define the screen's primary goal and the single most important user action. 2. List required content and UI elements in priority order. 3. Recommend a layout structure and visual hierarchy from most to least important. 4. Note states to design: empty, loading, error, and success. 5. List 3 usability risks for this screen and how the layout can mitigate them. </task>

A pre-design wireframe brief covering hierarchy, content, and edge-case states.

๐Ÿ’ก

Pro tip: Ask ChatGPT for a low-fidelity ASCII or text layout sketch so you can sanity-check structure before drawing pixels.

Information Architecture Draft

12/30

<context> I am organizing the navigation and content structure of [PRODUCT] so [USER] can easily complete [FLOW] and find related tasks. </context> <task> 1. Propose a top-level navigation with 4-7 primary categories and rationale for each. 2. Outline a sitemap two levels deep showing where key flows live. 3. Apply clear, user-language labels and avoid internal jargon. 4. Flag any content that could belong in multiple places and recommend a home. 5. Suggest a card-sort study to validate the structure with real users. </task>

A first-pass IA with sitemap, labels, and a validation plan.

๐Ÿ’ก

Pro tip: Tell ChatGPT to use your users' vocabulary, not company terms, and to list any label it is unsure about for a quick tree test.

User Flow Mapper

13/30

<context> I need to map the step-by-step user flow for [USER] completing [FLOW] in [PRODUCT], including decision points and error paths. </context> <task> 1. Lay out the happy-path flow as a numbered sequence of screens and actions. 2. Mark every decision point and branch with its condition. 3. Add error and recovery paths for the most likely failure points. 4. Identify steps that could be removed or combined to reduce friction. 5. Output the flow in a format I can recreate in a flow-diagram tool. </task>

A complete user flow with branches, error paths, and friction-reduction notes.

๐Ÿ’ก

Pro tip: Ask ChatGPT to express the flow in Mermaid syntax so you can paste it into a diagram tool and get an instant visual.

Content Prioritization Brief

14/30

<context> I have a content-heavy screen in [PRODUCT] for [USER] doing [FLOW] and I need to decide what to show, hide, or defer. </context> <task> 1. Classify each content element as essential, supporting, or optional for the primary task. 2. Recommend what belongs above the fold versus progressively disclosed. 3. Suggest where to use scanning patterns, grouping, or summarization. 4. Identify content that adds cognitive load without aiding the task. 5. Provide a prioritized content order for both mobile and desktop. </task>

A prioritized content plan that reduces clutter across viewports.

๐Ÿ’ก

Pro tip: Paste the actual content list and ask ChatGPT to rank each item, so the prioritization reflects your real screen, not a generic one.

Responsive Layout Planner

15/30

<context> I am designing [FLOW] in [PRODUCT] for [USER] across mobile, tablet, and desktop and want a responsive layout strategy before wireframing. </context> <task> 1. Recommend a layout approach for each breakpoint and what changes between them. 2. Specify how navigation, primary actions, and content reflow per device. 3. Identify touch-target, thumb-reach, and readability considerations for mobile. 4. Note any element that should be hidden, collapsed, or reordered on small screens. 5. List the 3 most likely responsive bugs to test for. </task>

A breakpoint-by-breakpoint responsive plan with mobile usability notes.

๐Ÿ’ก

Pro tip: State your minimum supported screen width so ChatGPT tailors thumb-reach and reflow advice to your real device targets.

Usability Evaluation

5 prompts

Usability Test Script

16/30

<context> I am running a moderated usability test of [PRODUCT] where [USER] attempts [FLOW]. I need a script that produces clean, comparable observations across sessions. </context> <task> 1. Write an intro that sets a think-aloud expectation and reassures the participant we are testing the product, not them. 2. Create 4-6 realistic, scenario-based tasks for [FLOW] with clear success criteria. 3. Keep task wording neutral so it does not reveal the path or hint at the answer. 4. Add probing follow-up questions to ask after each task. 5. End with post-test rating questions and an open debrief. </task>

A moderated test script with neutral, scenario-based tasks and success criteria.

๐Ÿ’ก

Pro tip: Ask ChatGPT to rewrite any task that names a button or menu, since those leak the solution and invalidate the test.

Heuristic Evaluation Checklist

17/30

<context> I am doing an expert review of [FLOW] in [PRODUCT] used by [USER] and want to apply Nielsen's 10 heuristics systematically. </context> <task> 1. For each of the 10 usability heuristics, list 2-3 concrete checks specific to [FLOW]. 2. Provide a severity scale (0-4) with definitions for rating any issue found. 3. Suggest where each heuristic most commonly breaks in flows like this. 4. Give me a table to log issue, heuristic, severity, location, and recommendation. 5. Recommend how to convert findings into a prioritized fix list. </task>

A heuristic-evaluation framework with severity scale and logging template.

๐Ÿ’ก

Pro tip: After your review, paste your raw issue list back and ask ChatGPT to assign severities consistently and sort by priority.

Usability Findings Report

18/30

<context> I finished usability testing [PRODUCT] on [FLOW] with [USER] and have observations to turn into a credible, stakeholder-friendly report. </context> <task> 1. Wait for my raw observations, then group them into prioritized findings. 2. For each finding, state the issue, evidence, affected users, severity, and a recommended fix. 3. Lead with an executive summary of the top 3 issues and overall task success. 4. Separate must-fix issues from nice-to-haves. 5. Keep tone objective and avoid blaming users or designers. </task>

A prioritized usability report with executive summary and actionable fixes.

๐Ÿ’ก

Pro tip: Tell ChatGPT who the audience is (execs vs engineers) so it adjusts depth and language for that reader.

Accessibility Quick Audit

19/30

<context> I want a first-pass accessibility review of [FLOW] in [PRODUCT] for [USER], aligned to WCAG 2.2 AA before a formal audit. </context> <task> 1. List the most relevant WCAG 2.2 AA checks for this flow across perceivable, operable, understandable, and robust. 2. For each check, describe how to test it manually and what failure looks like. 3. Highlight common issues for forms, color contrast, focus order, and keyboard navigation. 4. Recommend assistive-tech scenarios (screen reader, keyboard-only) to walk through. 5. Output a checklist I can run in under an hour. </task>

A practical WCAG 2.2 AA self-audit checklist for a single flow.

๐Ÿ’ก

Pro tip: Ask ChatGPT to flag which checks a tool like axe can automate versus which require manual testing, so you focus your time.

A/B Test Hypothesis Designer

20/30

<context> I suspect a friction point in [FLOW] within [PRODUCT] is hurting [USER] completion and want to design a sound A/B test. </context> <task> 1. Help me turn the suspected problem into a clear, falsifiable hypothesis. 2. Define the primary metric, guardrail metrics, and what success looks like. 3. Propose a control and 1-2 variant concepts that isolate one change each. 4. Note sample-size and duration considerations at a high level. 5. List confounds or biases that could invalidate the result. </task>

A falsifiable A/B hypothesis with metrics, variants, and confound checks.

๐Ÿ’ก

Pro tip: Give ChatGPT your rough traffic numbers so its sample-size guidance is grounded instead of generic.

Design Critique

5 prompts

Structured Design Critique

21/30

<context> I will describe or paste a design for [FLOW] in [PRODUCT] used by [USER] and want a structured, constructive critique rather than vague opinions. </context> <task> 1. Wait for my design description, then critique it against clarity, hierarchy, consistency, accessibility, and task fit. 2. For each issue, explain the underlying UX principle and the likely user impact. 3. Separate critical problems from minor polish items. 4. Offer at least one concrete improvement per issue, not just criticism. 5. Note what the design does well so feedback stays balanced. </task>

A principle-based critique that pairs every issue with a fix and acknowledges strengths.

๐Ÿ’ก

Pro tip: Describe the user goal and constraints first, so ChatGPT critiques against the real objective instead of generic best practice.

Critique Feedback Reframer

22/30

<context> I need to give a teammate critique on their work for [PRODUCT] and [FLOW]. I have rough notes and want to deliver them constructively without crushing morale. </context> <task> 1. Wait for my raw feedback notes, then reframe each into clear, kind, specific language. 2. Use a what-I-observe, why-it-matters, what-I-suggest structure. 3. Tie feedback to user impact and shared goals, not personal taste. 4. Phrase subjective points as questions or options rather than mandates. 5. Keep a respectful, collaborative tone throughout. </task>

Rough critique notes rewritten into constructive, observation-based feedback.

๐Ÿ’ก

Pro tip: Ask ChatGPT to keep your most important point first and limit the list to the top 3-4 items so the critique stays focused.

Self-Critique Pre-Review

23/30

<context> Before I share my design for [FLOW] in [PRODUCT] with [USER], I want to pressure-test it myself and catch weaknesses early. </context> <task> 1. Generate a self-critique checklist covering user goal fit, hierarchy, edge cases, and accessibility. 2. Ask me pointed questions a senior designer would raise in review. 3. Identify the 3 most likely objections stakeholders will have and how to pre-empt them. 4. Surface assumptions baked into the design that I should validate. 5. Suggest what evidence to bring to defend key decisions. </task>

A self-review checklist and likely-objection list to prepare for critique.

๐Ÿ’ก

Pro tip: Answer ChatGPT's questions out loud or in text, then ask it to point out which of your answers are weakest.

Design Rationale Articulator

24/30

<context> I made specific design choices for [FLOW] in [PRODUCT] and need to explain the rationale clearly to [USER] stakeholders who may push back. </context> <task> 1. Wait for me to list my key decisions, then articulate the reasoning behind each. 2. Tie each decision to user needs, research, or a UX principle. 3. Acknowledge trade-offs and explain why the chosen path wins. 4. Prepare a concise response to the 3 most likely counterarguments. 5. Keep explanations jargon-free for non-designers. </task>

Clear, defensible rationale for design decisions with counterargument prep.

๐Ÿ’ก

Pro tip: Tell ChatGPT which stakeholder is most skeptical so it tailors the rationale to that person's priorities.

Competitor UX Teardown

25/30

<context> I want to critique how competitors handle [FLOW] to inform my design for [PRODUCT] and [USER], focusing on what to learn rather than copy. </context> <task> 1. Wait for me to describe 1-3 competitor experiences for [FLOW]. 2. Evaluate each on usability, clarity, and how well it serves the user goal. 3. Identify patterns worth adopting and anti-patterns to avoid. 4. Note where a competitor strength may not transfer to our context. 5. Summarize 3 actionable takeaways for my own design. </task>

A structured competitor teardown with transferable takeaways and warnings.

๐Ÿ’ก

Pro tip: Paste real screenshots descriptions or flow steps so ChatGPT critiques specifics instead of guessing how the competitor works.

Go from copy-pasting to actually mastering AI.

AI Academy: 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ other tools. New tutorials added every week.

Start Your Free Trial

Design Systems & Documentation

5 prompts

Component Documentation Writer

26/30

<context> I am documenting a component in [PRODUCT]'s design system so [USER] designers and engineers use it correctly within [FLOW]. </context> <task> 1. Wait for me to describe the component, then write its documentation. 2. Include purpose, anatomy, variants, states, and props or options. 3. Add do and don't usage guidelines with rationale. 4. Specify accessibility requirements (roles, labels, keyboard behavior). 5. Provide concrete example use cases and when to choose a different component. </task>

Complete component documentation covering anatomy, usage, and accessibility.

๐Ÿ’ก

Pro tip: Ask ChatGPT to mirror an existing doc template you paste in, so new component docs stay consistent with your system.

Design Token Naming Guide

27/30

<context> I am defining or auditing design tokens for [PRODUCT] and want a scalable, semantic naming convention the team can follow. </context> <task> 1. Propose a token naming structure (category, concept, property, variant) with examples. 2. Recommend semantic over literal names and explain why. 3. Cover color, spacing, typography, radius, and elevation tokens. 4. Show how tokens map from primitive to semantic to component levels. 5. Flag common naming pitfalls that cause confusion at scale. </task>

A scalable, semantic token naming convention with multi-level examples.

๐Ÿ’ก

Pro tip: Paste your current token list and ask ChatGPT to spot inconsistencies and suggest renames before you refactor the system.

Pattern Usage Guidelines

28/30

<context> I want clear usage guidelines for a recurring UX pattern in [PRODUCT] so [USER] designers apply it consistently across [FLOW]. </context> <task> 1. Wait for me to name the pattern, then define when to use it and when not to. 2. Describe the standard behavior, content rules, and interaction states. 3. Provide best-practice examples and common misuse examples. 4. Note accessibility and responsive considerations. 5. List related patterns and how to choose between them. </task>

Adoption-ready guidelines for a UX pattern, including when not to use it.

๐Ÿ’ก

Pro tip: Ask ChatGPT to phrase guidelines as testable rules so reviewers can objectively check whether a screen follows the pattern.

UX Writing Style Snippet

29/30

<context> I am defining UX writing rules for [PRODUCT] so microcopy across [FLOW] feels consistent for [USER]. I want practical guidance, not abstract brand fluff. </context> <task> 1. Define voice and tone with 3-4 concrete adjectives and what each means in practice. 2. Give rules for buttons, error messages, empty states, and confirmations with examples. 3. Provide before-and-after rewrites that show the rules applied. 4. Cover capitalization, person, tense, and punctuation conventions. 5. List words and phrases to avoid and their preferred alternatives. </task>

A practical UX writing style guide with rules and before/after examples.

๐Ÿ’ก

Pro tip: Paste a few existing strings from your product so ChatGPT calibrates the voice to what you already have rather than inventing one.

Design Handoff Spec

30/30

<context> I am handing off [FLOW] in [PRODUCT] to engineering and want a clear spec so [USER]-facing behavior is built exactly as designed. </context> <task> 1. Wait for me to describe the screens, then produce a handoff spec outline. 2. Document layout, spacing, components used, and which design tokens apply. 3. Specify all states: default, hover, focus, active, disabled, loading, empty, error. 4. Detail interaction behavior, validation rules, and edge cases. 5. List open questions and decisions engineering needs from me before building. </task>

A thorough engineering handoff spec covering states, tokens, and edge cases.

๐Ÿ’ก

Pro tip: Ask ChatGPT to generate a checklist version of the spec so you can confirm every state is covered before handing it to developers.

Frequently Asked Questions

Copy a prompt, replace the [PRODUCT], [USER], and [FLOW] placeholders with your real project details, and paste it into ChatGPT. Many prompts ask ChatGPT to wait for your notes or design description first, so follow up by pasting that context. Then refine the output with the suggested pro tip.
[PRODUCT] is the app, site, or feature you are designing. [USER] is your target audience or persona. [FLOW] is the specific task or journey you are working on, such as onboarding or checkout. Replacing them with concrete details makes ChatGPT's output far more relevant.
No. ChatGPT is excellent for planning research, drafting guides and surveys, synthesizing notes you provide, and pressure-testing your thinking. It cannot observe real users or generate genuine data. Always validate AI-assisted personas, journeys, and findings with actual research participants.
A current reasoning-capable model handles synthesis, critique, and documentation well. For long transcripts or large content lists, paste content in batches and ask ChatGPT to maintain a running summary so it does not lose earlier context.
Give ChatGPT real context: your user goals, constraints, tech limits, and actual content or quotes. Ask it to label assumptions, request a specific output format like a table or Mermaid diagram, and iterate. The more grounded detail you provide, the more deployable the output becomes.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.