Landing Page Critique

Variables

Replace each placeholder before you run the prompt.

• {{page_url}}The landing page URL (or pasted markup).

  https://example.com/products/laser-markers

• {{campaign_intent}}What the ad-traffic visitor is expecting and what we want them to do.

  Visitors are searching "industrial laser marker for stainless steel". We want them to submit a quote request or book a spec review call.

Critique the landing page below for B2B paid-traffic readiness. Audience: a senior CRO + paid-media joint review.

PAGE:
{{page_url}}

CAMPAIGN INTENT:
{{campaign_intent}}

Score on these axes (0–10 each, 70+ total = solid):

1. Intent match: does the H1 align with the search query?
2. Above-the-fold clarity: in 5 seconds, can a visitor say what this is and what to do next?
3. CTA clarity & placement: is the primary CTA unambiguous and visible without scrolling?
4. Proof: does the page show specific proof (deployments, certifications, named customers) or generic claims?
5. Friction: count clicks/fields between landing and conversion. Is there hidden friction (account creation, mandatory phone)?
6. B2B vocabulary: does the copy sound like an engineer wrote it or a content marketer?
7. Mobile readability: does the page work on a phone in a manufacturing-floor wifi context?
8. Technical execution: page speed, contrast, broken links, lazy-load failures.

Output: scores per axis, 1–2 sentences of evidence per score, the total, and a Top 5 highest-impact fixes ranked by effort:impact ratio.

Do NOT recommend a redesign as a fix unless the page is genuinely unsalvageable. Find the surgical wins first.

Expected output shape

8 axis scores with evidence, total score, and a Top 5 fixes list.

Why we wrote it

Most LP critiques default to "improve the hero" with no specifics. This prompt forces 8-axis scoring with evidence so the agency can hand the result to the web team without arguing.

How to use

1. Open Claude or ChatGPT. The recommended model for this prompt is claude-sonnet-4-6 — opus when the prompt requires deep reasoning, sonnet for the rest.
2. Replace every {{variable}} with content specific to your account. The examples above are starting points, not templates to ship as-is.
3. Paste the prompt and run.
4. Read the output against the expected shape above. If the model produced a structurally different response, re-prompt rather than accept the drift.