The problem this solves
Every search campaign leaks. Queries that looked close enough to your keywords trigger your ads, collect clicks, and convert never - and the spend disappears into a report nobody opens. The search terms view holds the whole story, but reading it properly means scanning thousands of rows, judging intent query by query, and maintaining negative keyword lists that go stale the month after someone last groomed them.
It is classic important-but-boring work. The audit gets done thoroughly once, during onboarding or after a scare, and then entropy takes over: new queries drift in, match types loosen their grip, and the waste rebuilds a click at a time. Because no single wasted query looks alarming, the aggregate leak persists - budget that should be concentrated on converting terms quietly subsidizes traffic that was never going to buy.
How the mission runs
- Pull thirty days of search terms. The mission reads the full search terms report from Google Ads for the trailing thirty days: every query that triggered your ads, with its spend, clicks, and conversions. A read-only pull - the account is queried and left exactly as it was.
- Judge intent at the query level. The agent evaluates each meaningful query against your offer: is this someone who could buy, a job seeker, a student, a wrong industry, a competitor's customer looking for support? Spend on mismatched intent is tallied as waste with the reasoning recorded.
- Derive the negative keyword set. Wasteful queries are generalized into negative keywords and themes at the right level - exact negatives for specific offenders, phrase-level blocks where a whole pattern is bad - structured to cut waste without choking legitimate traffic.
- Publish the full analysis to Sheets. The complete audit lands in Google Sheets: every flagged query with its spend, verdict, and proposed negative, plus a summary tab totaling recoverable waste by campaign. The workbook is the reviewable record your team can filter, challenge, and act from.
- Post the top ten fixes to #performance. The mission posts the ten highest-impact fixes to your #performance channel with the recoverable spend attached to each. The team sees where the money is leaking without opening the workbook, and the fix conversation starts immediately.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
A Google Sheets workbook holding the complete thirty-day waste audit - every flagged search term with spend, intent verdict, and the negative keyword that would stop it, totaled by campaign - plus a Slack post in #performance ranking the top ten fixes by recoverable spend. Applying the negatives stays a deliberate action your team takes with the full evidence in front of them.
Make it yours
- Extend the window: audit ninety days to catch slower-bleeding patterns and seasonal query drift a single month can hide.
- Focus on one campaign type: audit only Performance Max or only brand campaigns, where waste hides in different shapes.
- Schedule it: run monthly so the negative list stays current and each audit reports the waste trend against the last one.
Frequently asked questions
Does the mission apply the negative keywords itself?
Not in this playbook - the audit reads the account and delivers the evidence, and applying negatives remains your explicit action. On this platform, reads are free and writes wait for approval, so account changes only ever happen when a human authorizes them.
How does it avoid flagging queries that actually convert?
Conversion data rides along with every query, and anything with meaningful conversions is protected from the waste list regardless of how odd it reads. The gray zone - spend without conversions on plausible intent - is flagged as judgment calls for your review rather than lumped in with clear waste.
We run broad match on purpose. Does this fight that strategy?
It complements it. Broad match strategies depend on disciplined negative lists to stay efficient - the wider the net, the more important the filter. The audit lets you keep the reach while systematically closing the intent leaks the wider matching lets in.