Your Lighthouse score is 94. Your PageSpeed report glows green. And your bounce rate is still 63%.
That gap exists because technical audits measure what machines can verify: load times, accessibility tags, image compression. They measure zero design quality dimensions. Not whether your headline communicates value in five seconds. Not whether visitors can find your call to action. Not whether your typography is readable on a phone screen.
A website critique fills that gap. It evaluates the design layer that actually shapes how people feel about your site, and whether they stay or leave. According to Stanford's Web Credibility Research, 75% of users judge a company's credibility based on visual design alone.
The good news: you don't need a design degree to run one. You need a scored rubric, 30 minutes, and a willingness to be honest about what you see. Below is the exact framework, dimension by dimension, step by step.
Why a Website Critique Catches What Technical Audits Miss
Technical audits and website critiques answer fundamentally different questions. Audits ask: does this site function correctly? Critiques ask: does this site communicate effectively?
Google Lighthouse evaluates five categories: Performance, Accessibility, Best Practices, SEO, and Progressive Web App compliance. Every one of those is machine-verifiable. None of them tells you whether your hero section makes sense, whether your visual hierarchy guides the eye toward a conversion action, or whether your site looks trustworthy enough to hand over a credit card number.
Nielsen Norman Group's research shows that users leave web pages within 10 seconds if they don't quickly perceive value. That perception is shaped by design dimensions: messaging clarity, visual weight distribution, contrast, typography. A perfect Lighthouse score cannot save a confusing layout.
If you want a deeper look at what a website critique includes and how it differs from a technical audit, we covered that distinction in detail previously. This guide picks up where definitions leave off: it gives you the scoring rubric and the process to actually do the critique yourself.
The 8 Dimensions of a Website Critique
A website critique is not a matter of taste. It evaluates design against measurable thresholds. Each of the following eight dimensions gets a score from 1 to 10, where 1 means "fundamentally broken" and 10 means "nothing to fix."
| Dimension | What to Evaluate | Pass/Fail Threshold |
|---|---|---|
| Visual Hierarchy | Does the page guide your eye in a clear sequence? Can you identify the primary, secondary, and tertiary elements instantly? | Fail: no clear focal point on the page; three or more elements compete for primary attention |
| Typography | Is body text legible? Are heading sizes distinct enough to signal structure? Is line height comfortable? | Fail: body text below 16px; fewer than 2 distinct heading sizes; line height below 1.4 |
| Color & Contrast | Do text and background meet contrast requirements? Does the palette feel cohesive rather than random? | Fail: contrast ratio below 4.5:1 for body text (WCAG AA standard); more than 4 primary brand colors |
| CTA Clarity | Is it obvious what action visitors should take? Does the primary CTA stand out from surrounding elements? | Fail: no CTA visible above the fold; CTA button blends into the background; CTA text is vague ("Submit," "Click Here") |
| Above-the-Fold Messaging | Can a stranger understand what you do, who it's for, and why it matters within 5 seconds of landing? | Fail: headline requires scrolling to find; no subheadline explaining the value proposition; jargon-heavy language |
| Navigation | Can users find what they need in two clicks or fewer? Is the nav structure predictable? | Fail: more than 7 top-level nav items; no mobile hamburger or equivalent; broken or orphaned links |
| Trust Signals | Are there social proof elements (testimonials, logos, case studies, security badges) visible without scrolling far? | Fail: zero trust signals on the homepage; testimonials without names or photos; no security indicators near payment flows |
| Motion & Interaction | Do animations serve a purpose (directing attention, confirming actions) or are they decorative noise? | Fail: animations that delay content visibility by more than 300ms; elements that move without user initiation on mobile; no hover/focus states on interactive elements |
This rubric borrows principles from Nielsen Norman Group's usability heuristics, adapted specifically for evaluating website design quality rather than application usability.
For a deeper look at the first dimension, which is often the root cause when a site "feels off," see our guide on visual hierarchy and how to fix it.
How to Critique Your Website in 5 Steps
Now that you have the rubric, here is how to use it. The entire process takes about 30 minutes. You'll need a laptop, a phone (for mobile testing), and a willingness to evaluate your site as a stranger would.
Step 1: Run the 5-Second Test
Open your homepage in an incognito window (no cached assets, no personalization). Stare at it for exactly five seconds. Then close the tab.
Write down your answers to three questions:
- What does this company do?
- Who is it for?
- What should I do next?
If you can answer all three clearly, your above-the-fold messaging is working. If you hesitated on any answer, that dimension starts with a low score.
For a more honest result, ask someone who has never seen your site to take the test. Founders are terrible at evaluating their own messaging because they already know what the product does.
Step 2: Score Each of the 8 Dimensions
Walk through the rubric above, one dimension at a time. For each:
- Open the relevant page (homepage for most dimensions, pricing page for trust signals and CTAs, blog for typography)
- Check the specific criteria in the "What to Evaluate" column
- Apply the pass/fail threshold first. If it fails, score 1 to 4. If it passes, score 5 to 10 based on quality
- Write a one-sentence note explaining your score ("CTA says 'Get Started' but it's the same blue as body links, so it doesn't stand out. Score: 4.")
Score on mobile separately. A responsive layout that looks balanced on desktop can collapse into an unreadable wall of text on a 375px screen.
Step 3: Identify Your 2 to 3 Lowest-Scoring Dimensions
Sort your scores. Circle the bottom two or three. These are your high-impact targets.
Most non-designer-built sites cluster their weaknesses in predictable spots: above-the-fold messaging (too vague), CTA clarity (too hidden), and visual hierarchy (too flat). If yours breaks the pattern, that's useful information too.
Step 4: Map Each Low Score to a Specific, Fixable Element
A score is a diagnosis. Now you need a prescription.
For each low-scoring dimension, identify the single element most responsible for the low score. Not "the whole page feels wrong." Something specific:
- "The hero headline says 'Welcome to Our Platform' instead of explaining the product" (above-the-fold messaging)
- "The primary CTA button is 12px gray text at the bottom of a 900px hero section" (CTA clarity)
- "Body text is 14px on a 1440px monitor with no max-width constraint, so lines run 140+ characters" (typography)
Specificity is what separates a useful site critique from a vague opinion. If you can't point to a single element, zoom in until you can.
Step 5: Prioritize by Impact and Effort
Not every fix matters equally. A two-word CTA rewrite that doubles click-through rates beats a three-day typography overhaul every time.
Rank your fixes on two axes:
- Impact: How much does this dimension affect conversion? Above-the-fold messaging and CTA clarity almost always rank highest. According to Unbounce's Conversion Benchmark Report, the median SaaS landing page converts at 3.8%, while top-quartile pages hit 11.6%. The gap between those numbers is almost entirely explained by clarity, trust, and CTA design, not load speed.
- Effort: How long will this fix take? Changing a headline is 5 minutes. Restructuring navigation is 5 hours. Start with high-impact, low-effort fixes.
SiteCritic automates Steps 1 through 4 of this process. Paste your URL and get a scored, timestamped critique across all 8 dimensions with specific observations tied to exact page elements. The manual process above works well, but it takes 30 minutes per page and requires you to evaluate your own work objectively (which is genuinely difficult).
DIY Site Critique vs. Expert Review vs. AI Design Critique
Three approaches exist for critiquing a website. Each fits different budgets, timelines, and depth requirements.
| Factor | DIY (This Guide) | Expert Designer Review | AI Design Critique |
|---|---|---|---|
| Cost | Free | $2,000 to $15,000+ | $0 to $50 per critique |
| Time to Complete | 30 minutes per page | 1 to 3 weeks | Under 2 minutes |
| Depth | Surface to moderate (depends on your design eye) | Deep (includes strategic recommendations, competitor benchmarks, redesign priorities) | Moderate to deep (scored dimensions, specific observations, timestamped feedback) |
| Objectivity | Low (hard to evaluate your own work) | High (external perspective) | High (algorithmic, not influenced by your pitch deck) |
| Best For | Early-stage founders validating before spending money | Pre-Series A companies investing in a major redesign | Founders who need fast, scored feedback on live pages or iterative design changes |
The DIY approach from this guide gives you a solid baseline. Expert reviews are justified when you're planning a significant redesign and need strategic direction, not just a list of issues. AI-powered critique tools like SiteCritic sit in the middle: faster and cheaper than hiring a designer, more structured and objective than doing it yourself.
If you want to see how different website design audit tools compare for startups, we reviewed the major options side by side.
Common Mistakes That Weaken a Website Critique
A critique is only as useful as the method behind it. These four mistakes turn a potentially valuable exercise into wasted time.
Critiquing taste instead of principles. "I don't like the blue" is not a critique finding. "The blue CTA button has a 2.1:1 contrast ratio against the background, which fails WCAG AA and makes it hard to read" is. The rubric above exists specifically to prevent taste-based evaluation. Stick to the dimensions. Stick to the thresholds.
Skipping mobile. Over 60% of web traffic is mobile. A site that scores 8/10 on desktop and 3/10 on mobile has a mobile problem, not a "minor responsive issue." Critique mobile separately, every time.
Ignoring messaging. Designers focus on visual polish. Founders focus on feature lists. Neither addresses the most common conversion killer: unclear messaging. If your five-second test fails, no amount of visual refinement will compensate. What makes a website look professional starts with clarity, not aesthetics.
Treating critique as a one-time event. Your website changes. Your audience evolves. Your competitors ship redesigns. A critique from six months ago is stale. Schedule a critique at every major milestone: before launch, after a redesign, at the start of each quarter, and whenever your conversion rate shifts more than 15%. A pre-launch design review checklist can help you build critique into your process rather than treating it as an afterthought.
What to Do After Your Website Critique
Scores without follow-through are just numbers on a page. Here is how to turn your critique into shipped improvements.
Fix the lowest-scoring dimension first. Not the easiest. Not the most interesting. The lowest-scoring one. A CTA that scores 2/10 is dragging your entire site's performance down more than a navigation structure that scores 6/10.
Focus above the fold. If both your above-the-fold messaging and your trust signals scored poorly, fix the messaging first. Nothing below the fold matters if visitors leave in the first five seconds.
Re-critique after every round of fixes. This closes the loop. You scored a 3 on typography, you bumped the body text to 18px and constrained line length to 75 characters. What does the dimension score now? Track your scores over time to see whether your site is trending toward the design quality that supports conversion.
Translate scores into design feedback. If you're handing critique findings to a designer or developer, raw scores need context. Our guide on translating critique findings into actionable design feedback walks through how to communicate findings so they actually get implemented correctly.
Benchmark against competitors. Once you have your own scores, critique 2 to 3 competitor sites using the same rubric. This gives you relative positioning: if every competitor scores 7+ on trust signals and you're at 4, that dimension jumps to the top of your fix list. We covered the full process for benchmarking your design against competitors if you want to formalize this step.
For ongoing iteration, SiteCritic generates scored critiques you can compare across versions of your site. Paste your URL after each round of changes, and you get a timestamped record of what improved and what still needs work. It turns a one-time exercise into a measurable design quality practice.
Start With One Critique
You don't need to overhaul your entire site today. You need 30 minutes, the rubric above, and honesty about what you see.
Score the 8 dimensions. Find the 2 or 3 that are dragging your site down. Fix the worst one this week. Then re-score.
That single loop (critique, fix, re-critique) will improve your site's design quality more than any amount of Lighthouse optimization or plugin installation. Design is what visitors judge. Scores are how you stop guessing about it.