← ClaudeAtlas

google-ads-settings-auditlisted

Use for "audit my Google Ads settings", "is my account set up right", "I inherited this account, what's wrong with it", "why is my traffic showing outside my area", "check my campaign settings", "new client account review" — even when the user never says "settings". Covers search partners, Display expansion, location intent, auto-apply, ad rotation, conversion windows, budgets and bid strategies, each priced in the spend flowing through it. Most of these toggles default in Google's favour. Google Ads only.
coupler-io/skills · ★ 33 · AI & Automation · score 72
Install: claude install-skill coupler-io/skills
# Google Ads Settings Audit **Finds the configuration defects that cost money quietly, and puts a spend figure on each one.** Most of these settings default in Google's favour and never announce themselves. Search partners delivering traffic that behaves nothing like your core inventory, location targeting set to anyone merely *interested in* your area, recommendations applying themselves — none of it looks broken, and all of it spends. The account will run for years like this and the performance report will never say why. **What you get back** - **A defect list ranked by the spend flowing through each one**, not by severity in the abstract. Two accounts with identical defects get completely different reports. - **The setting's current value and what it should be**, with the evidence behind the call. - **A passed list** — what was checked and is fine, so the reader can tell that from what was never looked at. - **An explicit not-checked list**, because several of these settings aren't readable at all. **Read-only on your Google Ads account.** It reports and recommends; it never changes a setting. **This is the cheapest skill in the pack to act on.** Most findings are a toggle. ## Call budget | | Calls to a spoken answer | |---|---| | Cold | locate the data → coverage verdict (speak) → one combined query = **3** | | Warm — dataset and deliberate settings already known | coverage verdict (speak) → one query = **2** | **Already known is not re-derived** — settings p