Find wasted spend in five minutes

In almost every account we audit, between 8% and 20% of non-brand spend goes to search terms that will never convert: job seekers, DIY researchers, the wrong product, the wrong country. Finding them used to mean exporting a report and scrolling for an hour. With an assistant connected to Google Ads it is two tasks.

The query

SELECT campaign.name, ad_group.name, search_term_view.search_term,
       metrics.clicks, metrics.cost_micros, metrics.conversions
FROM search_term_view
WHERE segments.date DURING LAST_30_DAYS
  AND metrics.clicks >= 5
  AND metrics.conversions = 0
ORDER BY metrics.cost_micros DESC

You will never type this; the assistant does. But it helps to know what "wasted search terms" means in data: clicks, cost, zero conversions, recent.

Classification rules

Ask the assistant to sort the list into buckets before you decide:

• Irrelevant intent: jobs, salary, careers, free, DIY, how to, tutorial, template, course.
• Wrong product or audience: terms for products you don't sell, B2C when you're B2B, other locations.
• Competitor names you don't want to bid on.
• Relevant but unconverted: leave these alone; low volume is not a verdict.
• Brand misspellings: keep.

The prompt

Review, say "go", and adicionar_negativas applies them. Two tasks, one of them a report.

Where to add the negatives

• Campaign level for anything that is wrong for the whole campaign (most cases).
• Ad group level for cross-contamination between ad groups (a "kids shoes" query triggering the "trail shoes" group).
• Shared negative list for account-wide junk; create once with executar_operacao_avancada, attach to all campaigns.

Make it weekly

Wasted terms regrow. Run this every week with a 7–14 day window; it becomes a one-sentence task. For the complete weekly routine see this playbook.