Product Research · Blog post · source Apr 14, 2026
Search Query Performance Deep Analysis with Claude Cowork
Turn the Search Query Performance report into keep, grow and cut decisions
Her guide says to enable all 33 columns before exporting, because the skill needs every funnel stage to find where share drops off, and gives the prompt that runs the analysis.
Per source Self-reported by the source, not independently verified.
Copy-paste prompt
Analyze the attached Search Query Performance export (all 33 columns). 1) Split branded vs unbranded queries using my brand terms: [terms]. 2) For each query, show my share of impressions, clicks, add-to-carts and purchases vs the market, and the stage where my share drops most. 3) Sort queries into Defend, Growth, Kill and Monitor, with one line of reasoning each.
The source didn't publish its exact prompt. We wrote this starter so you can try the same workflow. Replace anything in [brackets].
What they did
An ecommerce AI consultant published a Claude Cowork skill for Brand Analytics' Search Query Performance report. It splits branded from unbranded queries, maps each query through impressions, clicks, add-to-carts and purchases, measures your share against the market at each step, and sorts queries into Defend, Growth, Kill and Monitor buckets.
How it went
Her guide says to enable all 33 columns before exporting, because the skill needs every funnel stage to find where share drops off, and gives the prompt that runs the analysis.
We only report what the source says. Results are self-reported unless noted.
Worth knowing
A how-to guide, not a reported result. The buckets are a sensible way to turn SQP into actions; check Kill candidates against ad data before cutting spend on them.