Run a landing page conversion audit from GA4 and funnel data

Start by writing the landing page goal, target audience, traffic sources, offer, primary CTA, secondary CTA, and conversion event. Include what happens after conversion, such as form submission, demo booking, asset download, meeting routing, or nurture. This gives the audit a clear standard for what the page is supposed to do.

In GA4, pull sessions, users, source, medium, campaign, device, geography, landing page, engaged sessions, key events, and conversion rate for the page. Include a comparison period so you can tell whether the current performance is normal, improving, or declining.

Move the exported metrics into Google Sheets. Create columns for source, sessions, conversion rate, key event rate, bounce or engagement signal, device, conversion volume, and notes. Add flags for segments with meaningful traffic but weak conversion.

Compare the ad, email, social post, or search intent that drove traffic to the landing page promise. Look for mismatches between what the visitor expected and what the page delivers. Add notes for headline mismatch, offer mismatch, audience mismatch, CTA mismatch, and proof mismatch.

Give Claude the page goal, source data, segment table, message-match notes, and page copy. Ask it to separate traffic quality issues from page friction issues. The output should identify likely causes, confidence levels, and specific fixes.

Diagnose this B2B landing page conversion issue.

Page goal:
{{page_goal}}

Target audience:
{{target_audience}}

Offer:
{{offer}}

CTA:
{{cta}}

Traffic and conversion table:
{{traffic_conversion_table}}

Message-match notes:
{{message_match_notes}}

Landing page copy:
{{landing_page_copy}}

Output:
1. Likely issue type: traffic quality, message mismatch, page friction, offer weakness, CTA confusion, form friction, proof gap, or tracking issue
2. Evidence for each issue
3. Evidence against each issue
4. Affected segments
5. Confidence level from 1-5
6. Recommended fixes
7. Fast validation checks

Be practical and do not recommend generic best practices unless the data supports them.

Review the hero, subhead, proof, CTA, form, FAQ, objections, visual hierarchy, and mobile readability. Use Claude to identify where the page may be unclear, unsupported, or too generic. Tie every recommendation to the page goal and traffic source.

Audit this B2B landing page copy and structure.

Page goal:
{{page_goal}}

Audience:
{{audience}}

Traffic source:
{{traffic_source}}

Landing page copy:
{{landing_page_copy}}

Known conversion issue:
{{conversion_diagnosis}}

Review:
1. Hero clarity
2. Offer clarity
3. Proof strength
4. CTA clarity
5. Objection handling
6. Form or conversion friction
7. Message match
8. Mobile readability risks
9. Recommended copy changes
10. Recommended structure changes

Tie every recommendation back to the conversion goal.

Turn the diagnosis into a fix list in Airtable. Include issue, affected segment, recommended fix, owner, expected impact, effort, risk, validation metric, and deadline. Prioritize fixes that are easy to test and likely to affect the largest qualified segment.

Create a prioritized landing page fix list from this audit.

Conversion diagnosis:
{{conversion_diagnosis}}

Content audit:
{{content_audit}}

Available owners:
{{owners}}

Constraints:
{{constraints}}

Output a table with:
1. Priority
2. Issue
3. Affected segment
4. Recommended fix
5. Owner
6. Effort
7. Expected impact
8. Risk
9. Validation metric
10. Deadline

Keep the list practical and focused.

Use Looker Studio or your reporting dashboard to create a focused view for the landing page and the affected segments. Track the metric tied to each fix. After the test window, summarize what changed, what did not change, and what the next audit should check.

Summarize landing page audit results after the validation window.

Original diagnosis:
{{original_diagnosis}}

Fixes implemented:
{{fixes_implemented}}

Before and after metrics:
{{before_after_metrics}}

Quality notes:
{{quality_notes}}

Output:
1. What improved
2. What did not improve
3. Which diagnosis was supported
4. Which diagnosis was wrong or inconclusive
5. Next recommended test
6. Monitoring rule to add