Prompt Design: A UX Approach to Getting Better Answers from AI

A young woman writing in a notebook at a white desk, symbolizing the creative process of crafting AI prompts. — Turning ideas into precise words — where creativity meets clarity.

Imagine your AI tool as a junior designer joining your team. They’re fast, patient, and capable — but they don’t know your project until you explain it.

If you simply say, “Make me a good design”, the result will likely feel vague or misaligned. But if you say, “Design a layout for a fitness app for women aged 25–35, with a friendly tone, pastel colors, and a minimalist look”, the work instantly connects to your vision.

Prompting is more than typing words into a box. It’s a way of shaping a conversation. And, just like in UX, the clarity of your input directly influences the quality of the output.

Why Prompting Feels Like UX

In UX, clarity, constraints, and context aren’t luxuries — they’re what make design usable. The same holds true with prompts.

A prompt works like a micro-brief for your AI collaborator. Instead of tossing ideas into the void, you’re giving direction that reduces guesswork. A clear prompt is the difference between wandering results and meaningful outcomes.

Think of it as designing the conversation layer:

Clarity ensures the request is understood.
Constraints focus creativity on what matters.
Context connects your intent with the system’s response.

Whether you’re briefing a teammate or an AI, you’re shaping for understanding — and that’s what turns vague requests into valuable results.

Four Ingredients of a Strong Prompt

Context — Who, what, why
“You are a senior copywriter specializing in tech products. Write a tagline for a SaaS platform that helps startups manage remote teams.”
Clear Instructions — Action verbs and steps
“List five tagline ideas and sort them from most creative to most formal.”
Desired Format — How the output should look
“Put the ideas in a table with three columns: tagline, tone, and complexity.”
Tone and Style — Emotional direction
“Friendly, with subtle humor and a sense of innovation.”

UX Principles Applied to Prompting

UX designers rely on iteration, empathy, and constraints — and those habits carry over to prompt writing.

Iteration: The best outputs rarely come from the first try. A strong workflow looks like: write → test → refine → repeat.
Empathy: An AI doesn’t know your project history. Frame prompts as if briefing a new colleague who’s walking in cold. This reduces misalignment.
Constraints: In design, systems and guidelines bring order. In prompts, specificity does the same. Paradoxically, sharper boundaries often spark more creative responses.

Prompting Tips That Work

Use examples: Show the system the tone or format you expect.
Ask for variations: Request multiple options, not just one.
Assign roles: Frame the AI as a researcher, writer, or designer.
Break it down: Use smaller prompts in sequence instead of one massive request.
Be specific: “Minimalist logo with curved blue lines” is far better than “Nice logo”.

Collaboration, Not Commands

Prompts aren’t just instructions — they’re conversations. Treating AI like a creative partner, rather than a vending machine, opens space for surprise and exploration.

A weak prompt produces generic results, just like a vague design brief does. But a thoughtful prompt — one that combines clarity, boundaries, and creative space — leads to work that feels alive and relevant.

Design isn’t about getting perfection on the first try. It’s about exploring possibilities and shaping them into something meaningful. Prompting works the same way: the value lies in iteration, in the back-and-forth, in discovering directions you hadn’t planned.

Closing Thought

Good prompting reflects design thinking. Next time you type into ChatGPT, ask yourself: “If this were a brief for a teammate, would it be clear enough to act on?”. Because at its core, prompting is design — only the medium has changed.

I’d love to hear: how do you approach prompting? Do you write them as briefs, stories, or experiments?

Thanks for reading — I’m on a journey to understand how AI is reshaping the way we design, build, and think about products.

I write to explore ideas, question assumptions, and spark better conversations.

What’s something this made you reflect on? I’d love to hear your perspective in the comments.