Skip to content
Design & CreativityAdvanced

Get a Structured Design Critique

Get specific, prioritized feedback on a screenshot or mockup against your goals, with concrete fixes rather than vague opinions.

Share on X (opens in a new tab)Share on LinkedIn (opens in a new tab)

Make it yours

Fill in each variable and watch the prompt update. Anything still in {{braces}} needs your input before you use it.

Fill in the blanks

0 of 4 filled

What you're sharing.

What the design must achieve.

Who will see it.

Anything you already suspect is wrong.

Live previewUnfilled
You are a senior product designer giving a constructive design critique.

I've attached {{design_type}}. (If you can't see an image, tell me and stop — don't guess what it looks like.)

The goal of this design: {{goal}}
Target users: {{users}}
Specific concerns I have: {{concerns}}

Critique it in this order:
1. **First impression (5-second test)** — what stands out first, and is it the right thing?
2. **What's working** — 2–3 specific strengths to keep.
3. **Issues** — a table with columns: Issue | Where | Why it matters for the goal | Severity (high/medium/low) | Suggested fix. Consider visual hierarchy, readability and contrast, spacing and alignment, consistency, accessibility, and clarity of the call to action.
4. **Top 3 changes** — the highest-impact fixes, in order.

Rules: Be specific ("the price is the same size as the body text"), not vague ("make it pop"). Base comments only on what is visible; if you're inferring something (like exact contrast ratios), say so and suggest how I can check it.
Open in Playground

Example

Example input

Filled prompt
You are a senior product designer giving a constructive design critique.

I've attached a screenshot of our pricing page (desktop). (If you can't see an image, tell me and stop — don't guess what it looks like.)

The goal of this design: get small-business visitors to start a free trial of the middle plan
Target users: busy shop owners, not very technical, often comparing us with two competitors
Specific concerns I have: the page feels crowded and nobody clicks the FAQ

Critique it in this order:
1. **First impression (5-second test)** — what stands out first, and is it the right thing?
2. **What's working** — 2–3 specific strengths to keep.
3. **Issues** — a table with columns: Issue | Where | Why it matters for the goal | Severity (high/medium/low) | Suggested fix. Consider visual hierarchy, readability and contrast, spacing and alignment, consistency, accessibility, and clarity of the call to action.
4. **Top 3 changes** — the highest-impact fixes, in order.

Rules: Be specific ("the price is the same size as the body text"), not vague ("make it pop"). Base comments only on what is visible; if you're inferring something (like exact contrast ratios), say so and suggest how I can check it.

Example output

Illustrative example response — written by Promptly, not generated live

1. First impression The large hero illustration draws the eye first; the three plan cards sit below the fold on a typical laptop screen. The thing you most want seen — the recommended plan — isn't the first thing noticed.

2. What's working

  • Plan names are short and distinct
  • Prices use a consistent format across cards
  • The free-trial message appears on every card

3. Issues

IssueWhereWhy it mattersSeverityFix
Middle plan isn't visually emphasizedPlan cardsUsers can't see which plan to chooseHighAdd a "Most popular" label and a stronger border or tint
9 features listed per cardPlan cardsHard to compare quicklyMediumShow 4 key differences; move the rest to a comparison table
Light gray body textFeature listsMay fail contrast guidelines (inferred — check with a contrast tool)MediumDarken text color
FAQ is collapsed under a small linkPage bottomExplains low FAQ clicksLowShow the top 3 questions open by default

4. Top 3 changes

  1. Shrink the hero so plan cards appear above the fold
  2. Highlight the middle plan
  3. Cut feature lists to the key differences

Why it works

This is a multimodal prompt: the image carries the design, while the text supplies what the image can't — the goal, users, and concerns. Critique judged against a goal is far more useful than general taste. The instruction to stop if the image isn't visible, and to label inferred measurements, guards against the model inventing details it can't actually see.

Techniques used