Ecommerce Landing Page Conversion Audit
A structured audit for ecommerce landing pages that need clearer messaging, stronger proof, lower friction, and better conversion flow.
Browser-first guidance
If a live page URL is available, prefer OpenClaw managed browser before generating the audit.
Recommended order:
- If the user already has page copy pasted, use it as a starting point.
- If the live page is accessible, inspect it with OpenClaw managed browser.
- Use Browser Relay only when the user explicitly wants to inspect their currently attached Chrome tab.
Solves
Many landing pages underperform because:
- the headline is vague or generic;
- the offer is not easy to understand;
- proof is weak or badly placed;
- the CTA appears before trust is built;
- too many sections create friction instead of momentum.
Goal: Find the biggest conversion blockers on the page and turn them into a practical fix list.
Use when
- A landing page is getting traffic but not converting well
- The page feels busy or unclear
- The offer may be good, but the page does not explain it well
- Teams want a structured CRO review before scaling spend
Inputs
- Landing page URL or page copy
- Target audience
- Offer details
- Traffic source or intent
- Known constraints or compliance boundaries
Workflow
- Clarify the conversion goal.
- purchase?
- lead?
- quiz start?
- add to cart?
- Review headline, subhead, and above-the-fold clarity.
- Check page structure, proof placement, and CTA flow.
- Identify friction points, confusion points, and weak sections.
- Prioritize fixes based on likely conversion impact.
- Suggest improved wording or structural changes where useful.
Output
- Conversion blockers
- Priority-ranked fixes
- Copy / structure recommendations
- CTA and proof notes
- Quick-win suggestions
Quality bar
- Focus on commercial clarity, not generic design commentary.
- Recommendations should be specific enough to implement.
- Separate major blockers from minor polish issues.
- Keep the audit grounded in ecommerce conversion logic.
- If the observed page surface is partial, label that clearly.
Resource
See references/output-template.md.