Responsive search ads assemble headlines and descriptions in different combinations using auction-time signals. The asset report can now expose detailed statistics, and the combinations report shows common assembled ads. Those views are useful, but individual asset metrics are not clean isolated experiments: several assets can receive credit for the same impression, and ratios depend on the combinations and contexts in which each asset served.
VISUAL LESSON
What you will learn
- 01Open ad-level asset and combinations reports.
- 02Explain non-summable metrics and directional ratios.
- 03Build a cautious keep, replace, or test decision.
ILLUSTRATIVE ONE-AD REPORT
Asset impressions can exceed ad impressions
GOOGLE ADS INTERFACE MAP
Read the ad from three views
Review each headline and description with impressions, clicks, cost, conversions, value, added-by status, and date range where available.
See which combinations appeared frequently without assuming a static recreation would perform the same way.
Review query themes, landing-page alignment, campaign outcomes, and conversion lag before replacing an asset.
THE RSA EVIDENCE STACK
Asset → combination → query → qualified outcome
THE LEAD ATLAS METHOD
Lead Atlas Data can produce a contact list researched for the Search campaign's specific locations, business categories, and market, while the ad team evaluates RSA assets against separately reconciled qualified leads.See how custom list research works ↗Start at the campaign outcome
Set the date range, allow conversion delay, and review search terms, cost, conversions, conversion value, qualified leads, and landing-page performance at campaign and ad-group level. Confirm tracking before making creative judgments.
An asset report cannot rescue an ad group serving irrelevant queries or sending people to a mismatched page. Diagnose the entire path first.
Read asset statistics directionally
Compare headlines with headlines and descriptions with descriptions inside the same ad and time frame. Note serving volume, added-by status, and whether an asset lacks enough evidence.
Do not sum asset impressions, clicks, cost, or conversions to reproduce the ad total. Several assets may receive an instance of the same served ad, and ratios are influenced by co-served assets and auction context.
Use combinations for context
The combinations report shows common assembled ads and their impression frequency. Use it to check repetition, meaning, policy-safe phrasing, and whether pinned or flexible assets create awkward messages.
Do not copy a common combination into a static ad and assume it will behave identically. The original RSA was selected in real time with signals the reconstruction may not share.
Connect assets to customer intent
Review search-term insights and detailed terms to see which categories of demand reached the ad. Strong assets should accurately answer valuable queries and lead to a destination that continues the same promise.
Replace an asset when it has enough directional evidence, duplicates another message, creates weak combinations, or no longer matches the offer. Preserve major concepts for a controlled experiment when causality matters.
Complete the RSA decision sheet
Export one ad's asset statistics, mark non-summable columns, review its common combinations and top search categories, then classify each asset as keep, revise, replace, or insufficient evidence.
Deliverable: annotated report, combination screenshots, query-to-message map, qualified-outcome check, replacement brief, pinning rationale where used, and the next review date.
THE TAKEAWAY
Use asset data directionally, avoid summing non-summable metrics, read combinations and search intent, and test major message changes with a fair experiment when causality matters.OFFICIAL REFERENCES