Creative performance analytics

Creative metadata describes what is in an asset, such as its type, headline, image, logo or video. Performance data shows how an asset or campaign performed. Customer signals add the language people use when they describe a problem, objection, desire or reason to buy.

Google’s asset-reporting guidance says assets can be compared across ads and campaigns, but assets should be compared by asset type. That makes the input model practical: use metadata to define the comparison, performance data to observe the result, and customer signals to suggest the next hypothesis.

A pattern might show that ads with a particular message are associated with stronger results. That observation is useful, but it does not prove the message caused the result. Observational studies can identify associations but cannot establish causality without randomisation, as explained in a 2023 National Library of Medicine article.

A test decision adds discipline to the pattern: name the hypothesis, define one variable to change, retain an original version for comparison, and select the primary success metric or metrics. Google Ads describes this same original-versus-trial structure for custom experiments and allows up to 2 primary success metrics in reporting.

A team notices that customer feedback repeatedly raises a product-proof objection. The team does not conclude that proof is the cause of every conversion. Instead, it creates 2 versions of the same ad: an original version and a trial version that changes only the proof presentation.

Before launch, the team chooses up to 2 primary success metrics. After the comparison, the output is a decision: retain the original, iterate on the proof hypothesis, or scale the trial. That is creative performance analytics doing decision work, not merely reporting a pattern.

Component-level figures can be misleading when they are treated as campaign totals. Google notes that app asset reports may show higher metrics because each component used in one impression is counted, and advises relying on campaign-level totals for accurate performance and spend tracking.

Public competitor ads also have limits. Meta’s Ad Library shows ads currently running across Meta technologies and typically reflects first impressions or changes within 24 hours, but public visibility does not establish ordinary commercial advertisers’ ROAS, conversion rate or profitability. Read more in Meta Ad Library Shows Ads, Not Competitor Results.

Customer feedback can improve the quality of a hypothesis, especially when it reveals a recurring objection or desired outcome. It is not a substitute for a controlled comparison, and it should not be presented as performance proof.

The FTC’s 2024 Consumer Reviews and Testimonials Rule guidance explains that reviews featured in advertising or marketing become testimonials. Teams should therefore distinguish insight gathering from using a review as promotional proof. For a test-planning extension, see Creative Analytics Is Not a Test Plan, Here’s the Missing Layer.