Skip to content
Syncfusion GTM Playbook

Ads account audit

Type
checklist
Status
draft
Owner
COE Team
Last reviewed
2026-11-07
Products
All products

Draft. Not yet reviewed. Check facts before you use it with customers.

Run quarterly per product account (and before any budget increase) to catch the common Google Ads and Microsoft Ads pitfalls; done means every item has a Pass/Fail/n/a result and an approved fix list. The Claude skill google-ads-audit runs this checklist.

Before you start

  • Read-only access to the account through MCP, with your own login. See MCP and tools.
  • The audit is read-only. Claude proposes fixes as a list; a human approves each change that alters spend or live campaigns in chat before anything is applied.
  • Have the product’s section of pricing and licensing and the Google Ads strategy open; they define what “right” looks like.
  • Date range: last 90 days, compared with the prior 90.

Checklist

Phase 1: targeting and spend leaks (paid search lead or Claude)

  • A1 Broad match without negatives. Broad match keywords run only with Smart Bidding, enough conversions and a full negative list. Flag broad keywords whose search terms are mostly irrelevant.
  • A2 Search Partners and Display bleed. Search campaigns have Search Partners and Display Network off, unless a test shows qualified pipeline from them.
  • A3 No student/free/job-seeker negatives. Shared negative list includes free, student, tutorial, course, jobs, careers, salary, internship, crack, torrent (keep “free” only where a free tier or trial is the goal).
  • A4 Geo “presence or interest”. Location option is “Presence: people in or regularly in” your targeted locations.
  • A5 Search-term waste. Top 50 search terms by spend reviewed; irrelevant ones flagged.
  • A6 Audience Network (Microsoft Ads). Reviewed or off on search campaigns.

Phase 2: automation and conversions (paid search lead + sales ops)

  • B1 Auto-applied recommendations. Off in both Google Ads and Microsoft Ads. List any applied in the period.
  • B2 PMax cannibalizing brand. Brand terms excluded from Performance Max (brand exclusions or negative keywords); brand campaign impression share stable.
  • B3 Wrong conversion goals. Primary goals are qualified pipeline stages imported from the CRM, not raw form fills, page views or scroll events. Form fills are secondary.
  • B4 Offline conversion import. GCLID/GBRAID/WBRAID captured on forms; import ran in the last 7 days without errors.
  • B5 CPA vs pricing model. Target CPA or ROAS per campaign fits the product’s model (entry tier vs enterprise deal). No single target across very different offers.
  • B6 Brand exclusion. Brand terms are negatives in non-brand campaigns.

Phase 3: ads, compliance and landing pages (product marketing)

  • C1 Competitor trademark issues. No competitor trademarks in ad text unless permitted; competitor copy is positive and factual; any trademark complaints or disapprovals resolved.
  • C2 Landing page mismatch. Each ad group’s landing page matches its query theme and intent (pricing → pricing page, evaluation → demo/contact sales, developer → docs/trial). Pages pass the website buyer-fit audit.
  • C3 Disapproved or limited ads. None open longer than 7 days.
  • C4 Claims. No unsourced claims in ads (for BoldDesk, never the 90% automation goal, TODO: source). No prices that could go stale.

Results format

Claude prints the result in chat; save it to the team’s shared results log (location TBD). If an output directory is given (for example by pbmaintain), also write the report files defined in reports/schema/README.md there. Otherwise don’t create results files in the repo. Use exactly this format:

## Ads account audit: {product slug}, {account name}, {YYYY-MM-DD}
Period: {date range} | Run by: {name or "Claude"}


| ID | Item | Result | Evidence | Est. monthly waste | Proposed fix |
|---|---|---|---|---|---|
| A1 | Broad match without negatives | Fail | {campaign, keyword, search terms} | {currency or "unknown"} | {change} |


### Score
- Pass: {n} | Fail: {n} | n/a: {n}


### Proposed changes (need human approval before applying)
1. {campaign / setting}: {current} → {proposed}. Reason: {item ID}.

Use “unknown” for waste that can’t be calculated from account data.

Report keys (for the report files): Item IDs are A1–C4 above. Metric keys: ads.spend, ads.wasted_spend_est, ads.audit_pass_rate.

Done when

  • Every item has a result with evidence.
  • The proposed change list is approved, rejected or deferred by the paid search lead in chat, and approved changes are applied.
  • Findings are summarized in the next weekly GTM review.

If something goes wrong

  • A change caused a drop in conversions or a spend spike: revert it from the account change history and tell the paid search lead the same day.
  • Trademark complaint from a competitor: pause the affected ads and tell product marketing.
  • MCP access fails: run it manually in the Ads UI, same format.