AI Prompts for UX Designers (To Think, Not Just Generate)

Most AI prompts for UX designers floating around are aimed at producing output — generate a persona, write test scripts, draft copy. Useful, but that’s not where the real friction is. As a designer, you rarely get stuck because you don’t know how to design. You get stuck deciding what to do next.

This is a smaller, sharper set of prompts. They don’t design for you. They use AI as a thinking partner — to clarify the problem, explore options, pressure-test a decision, and refine before you commit. They’re grouped by the moment you’re in, not by deliverable.

The principle behind all of them: the value isn’t the AI’s answer. It’s how fast the answer helps you see the problem clearly or make a decision. Stop once you have clarity — the goal is momentum, not a perfect response.

How to use these prompts

• Copy, then adapt. Don’t paste blindly. Replace the brackets with your real product, users, and constraints.
• One at a time. Pick the prompt that matches your situation. Don’t chain them until you’re clear on the output you need.
• Keep your judgment in charge. Use these to challenge, validate, or expand your thinking — not to override it.
• Stop at clarity. When you can name your next move, you’re done. Close the tab.

The prompts are grouped into four moments: when you’re stuck, exploring, deciding, and refining.

When you’re stuck

For when the screen is blank or the brief is vague and the direction isn’t clear.

Clarify the real problem

I’m designing [feature or flow] for a mobile app. Before I sketch anything, help me state, in plain language: the user’s primary goal, the main friction they hit right now, and what success looks like from the user’s point of view. Push back if my framing is vague. Stay concrete — real behavior, not UX theory.

Use when the brief feels fuzzy or you’re not sure what problem you’re actually solving.

Cut the scope down

This problem feels too big: [describe it]. Help me split it into what must be solved in this iteration, what can wait, and what I should deliberately ignore for now. Reason like a designer working under a real deadline, and tell me which single piece would teach me the most if I designed it first.

Use when you’re trying to solve everything at once and nothing is moving.

When you’re exploring

For opening up ideas and flows without getting lost or overthinking.

Generate genuinely different options

I’m designing [feature or flow]. Give me three structurally different approaches — not visual variations. For each, name the core idea, when it would work best, and one real drawback. Focus on UX structure and user behavior, not styling.

Use when you have one obvious idea and want alternatives before you commit to it.

Walk the journey, find the drop-offs

Walk me through the realistic user journey for [task] in a mobile app. Mark the key decision points, the moments of hesitation, and where users are most likely to drop off. Assume messy real-world usage, not an ideal path.

Use when you want to sanity-check a flow before designing screens.

Surface the edge cases now

For [feature], list the edge cases and user states I’m likely to overlook — first-time users, returning users, empty states, error states, and partial data. Keep it concise and practical for design planning, and flag which ones would force a redesign if I ignored them.

Use when you want to avoid a rework later caused by a missed scenario.

When you’re deciding

For evaluating options and moving forward with confidence.

Compare two options honestly

I’m choosing between two directions. Option A: [describe]. Option B: [describe]. Compare them on usability, clarity for the user, implementation complexity, and long-term scalability. End with a recommendation and the reasoning — and tell me what would have to be true for the other option to win.

Use when you’re stuck between two approaches and circling.

Pressure-test the decision you like

I’m leaning toward [decision]. Act as a skeptical reviewer: name the trade-offs I’m making, the risks or downsides, and what I should watch after release. Don’t reassure me — try to find the weak point.

Use when a decision feels right intuitively but you want it stress-tested before you commit.

When you’re refining

For critiquing and improving without losing momentum.

Critique with fresh eyes

Here’s a screen or flow: [describe or paste context]. From a user’s perspective, point out what might cause confusion, what’s unnecessarily complex, and where clarity could improve. Practical UX, not visual trends. Rank your points by impact.

Use when you’ve stared at a design too long and lost objectivity.

Tighten the words

Review the UX copy in this flow: [paste copy]. Make it clearer, more human, and easier to scan — without changing the meaning. Keep the tone neutral and appropriate for a product. Show me before/after for each change.

Use when labels, buttons, or messages don’t quite land.

From asking yourself to being asked

Here’s the honest limit of a prompt pack: it only works when you remember to use it, phrase it well, and stay objective about your own work. That’s exactly the part that breaks down — you know what you meant, so the gaps stay invisible to you.

That’s the gap a case study is judged on. For the structure behind a strong one, use the UX case study guide; to see the thinking in finished work, read the examples; and before you publish, run the checklist. If yours keeps getting passed over, start with why UX case studies get rejected.