Run AI message-match QA across ads, landing pages, and CTAs

Create one row per paid journey in Google Sheets, not one row per landing page. For each row, capture channel, campaign, ad group or audience, keyword or targeting logic, ad headline, ad body, destination URL, UTM parameters, landing page headline, primary CTA, form or conversion event, and offer. Pull the ad details directly from Google Ads or LinkedIn Ads so the QA reflects what buyers will actually see, not a summarized media plan. Add owner, launch date, review status, and severity columns so issues can be assigned instead of discussed loosely. This table becomes the control center for every downstream message-match and tracking check.

Use Claude to turn each ad into a structured promise inventory. Ask it to identify the primary promise, pain point, implied audience, expected proof, expected next step, funnel stage, and any wording that could create an unsupported expectation. Keep the output close to the ad language so the page audit does not drift into general campaign strategy. Add the promise inventory back into Google Sheets next to the original ad copy. The landing page should now be judged against what the ad made the buyer expect, not what the team intended internally.

Extract the buyer promise from these ads.

Ads:
{{ad_copy}}

Campaign context:
{{campaign_context}}

Target audience:
{{target_audience}}

For each ad, return:
1. Primary promise
2. Target buyer or role implied by the ad
3. Pain point or job-to-be-done
4. Expected proof the landing page should provide
5. Expected CTA or next step
6. Implied funnel stage
7. Any risk of overclaiming or unclear language

Be specific. Do not rewrite the ads yet.

For each destination URL, capture the visible landing page copy, first-screen headline and subhead, CTA text, form fields, proof sections, FAQ sections, thank-you path, and mobile view notes. Use Webflow for page source or draft content when available, but also review the live or staged page exactly as a buyer would see it. If the page changes by UTM, device, geo, or personalization rule, capture the variant used by that campaign row. Add screenshots or links to the QA tracker so reviewers can verify the exact page state later. This prevents the common mistake of auditing outdated copy or a desktop-only experience.

Feed Claude the ad promise inventory, landing page evidence, target audience, offer, and conversion goal. Score each journey on headline match, offer match, proof match, CTA match, audience specificity, objection coverage, form friction, and above-the-fold clarity. Require Claude to explain every score below four with the exact mismatch and the specific buyer confusion it may create. Add the score, severity, recommended fix, and owner to Google Sheets. The output should separate true launch blockers from small copy improvements so the team can act quickly.

Audit message match between these ads and landing pages.

Ad promise inventory:
{{ad_promises}}

Landing page evidence:
{{landing_page_evidence}}

Audience:
{{audience}}

Conversion goal:
{{conversion_goal}}

Score each journey from 1-5 on:
1. Headline match
2. Offer match
3. Proof match
4. CTA match
5. Audience specificity
6. Objection coverage
7. Form friction
8. Above-the-fold clarity

For every score below 4, return:
- The mismatch
- Why it matters to the buyer
- Severity: blocker, high, medium, or low
- Recommended fix
- Whether the fix needs PMM, web, paid media, analytics, or legal review

Do not invent new proof points or claims.

Before rewriting copy, verify that the measurement path is sound. In GA4, confirm the relevant event or key event fires for the form, demo request, download, or thank-you page. Check that Google Ads and LinkedIn Ads destination URLs include the intended tracking parameters and that the CRM or form capture keeps campaign source, medium, campaign, content, and term fields where needed. Add every issue to the QA tracker with severity and owner, including broken UTMs, duplicate events, missing thank-you pages, form errors, mobile layout breaks, and CRM routing gaps. A page with strong message match can still fail as a campaign asset if the conversion signal is broken.

Use Claude to turn the audit into a prioritized fix package. Separate items into must-fix before launch, should-fix this week, and test later. For each must-fix copy issue, ask for one best replacement and one safer alternative for the headline, subhead, proof block, CTA, FAQ answer, or form microcopy. Route claim-sensitive changes to PMM or legal before publishing, and route tracking issues to analytics or ops. Update the Webflow page only after the issue owner marks the fix approved in the tracker.

Create prioritized landing page fixes from this message-match and tracking audit.

Audit findings:
{{audit_findings}}

Current page copy:
{{current_page_copy}}

Brand rules:
{{brand_rules}}

Approved proof points:
{{approved_proof_points}}

Create:
1. Must-fix before launch
2. Should-fix this week
3. Test later
4. Replacement headline and subhead for each must-fix journey
5. Proof block rewrite
6. CTA rewrite
7. FAQ or objection copy
8. Form microcopy recommendation
9. Tracking or handoff fixes
10. Approval owner for each fix

Rules:
- Do not invent proof points or claims
- Keep copy aligned to the original ad promise
- Flag anything that needs PMM, legal, analytics, or web approval

After approved changes are published in Webflow, click through each journey using the exact ad destination URL and tracking parameters. Confirm desktop and mobile layout, CTA links, form behavior, thank-you page, conversion firing, and CRM capture. During the first one to two weeks after launch, monitor CTR, bounce or engagement rate, scroll depth, CTA clicks, form conversion, sales acceptance, and qualitative comments from sales. Use Claude to summarize remaining mismatch signals and recommend whether to adjust the ad, page, offer, or audience. Save the post-launch review in the same Google Sheets tracker so the next campaign starts with evidence instead of memory.

Analyze post-launch message-match performance.

Journey map:
{{journey_map}}

Pre-launch QA findings:
{{pre_launch_qa_findings}}

Performance data:
{{performance_data}}

Sales or SDR feedback:
{{sales_feedback}}

For each journey, return:
1. Likely issue: ad, page, offer, audience, tracking, or unclear
2. Evidence from the data
3. Remaining mismatch signal
4. Recommended next action
5. Whether to change copy, targeting, CTA, proof, form, or tracking
6. Priority for this week

Focus on practical changes, not generic CRO advice.