7/ Search Term Mining — Google

What it does

Analyzes your search term reports across all campaigns and surfaces high-intent terms you're not bidding on yet. Groups them by theme, estimates their potential volume and CPA, and recommends match types and starting bids for each.

How it works

Claude processes your search term data over a 30-90 day window, identifies converting terms that are currently matched through broad or phrase but don't have dedicated keywords, and separates them from noise. It also finds patterns — clusters of related terms that suggest new ad group or campaign opportunities you're missing.

Practical example

Across your SaaS client's 8 search campaigns, Claude finds 67 converting search terms that don't have exact match keywords. It groups them into 5 themes: integration-specific queries (Salesforce, HubSpot), use-case queries (lead scoring, pipeline tracking), comparison queries (vs Competitor X), pricing queries, and enterprise-specific terms. Each group comes with estimated monthly volume, current CPA from broad match, and recommended starting bids.

What you get back

• Keyword opportunities grouped by theme with volume estimates
• Current performance data for each term (impressions, clicks, conversions, CPA)
• Recommended match type and starting bid for each new keyword
• New ad group or campaign structure suggestions based on term clusters
• Negative keyword recommendations to prevent overlap with existing campaigns

When to use it

• Monthly keyword expansion reviews
• When search campaigns plateau and you need new growth vectors
• After launching a new product or feature that creates new search demand
• When onboarding a new account to find what the previous team missed