Circana has introduced AI-powered predictive commerce signals for its Complete Audiences advertising solution, aiming to help marketers identify consumers whose purchasing behavior may be about to change and activate campaigns around those expected shifts.
Announced on October 28, 2026, the new capabilities are designed to identify households likely to enter a product category, increase spending, switch brands or disengage. Rather than relying primarily on historical purchases, advertisers can use these signals to build audiences around anticipated demand and align media spending with objectives such as sales growth, customer acquisition and retention. Circana
The launch reflects a broader shift in audience strategy: using purchase data and predictive modeling to estimate what consumers may do next, rather than targeting them only on the basis of past activity. For advertisers, that could help identify potential customers earlier in the buying journey or recognize existing customers at risk of switching to a competitor.
Circana says the signals draw on 129 million modeled U.S. households, nearly 2.6 million universal product codes (UPCs) and 500 billion deterministic signals. These data inputs provide the foundation for its commerce intelligence and household-level audience modeling. Circana+1
Predictive audiences move beyond past purchases
Purchase-based advertising audiences have traditionally helped brands reach consumers who bought a product, shop within a category or fit a defined customer segment. Predictive signals add another layer by estimating the likelihood of a future behavior.
For example, a consumer expected to enter a category could represent a potential acquisition opportunity. A household predicted to increase spending might be relevant to a premium product campaign, while a likely brand switcher could be a candidate for retention messaging. These are potential applications of the signal categories described by Circana, not guaranteed outcomes for individual consumers.
The approach could help marketers make audience selection more closely reflect business priorities. However, predictive accuracy depends on the quality, coverage and timeliness of the underlying data, as well as the way audiences are modeled and activated across media platforms.
Circana says the signals connect planning, activation and measurement, allowing advertisers to assess campaigns against commercial objectives rather than relying exclusively on media delivery metrics. The announcement does not specify all activation partners or the full measurement methodology for the new capabilities.
Early tests report performance gains
Circana reports that a recent test involving 13 brands showed improved performance in every campaign using its commerce signals. Gains ranged from 11% to 36%, with a median uplift of 22%, according to the company. Circana+1
The results provide an initial indication that predictive commerce audiences may improve campaign outcomes, but the announcement does not disclose the sample sizes, campaign objectives, comparison groups, test duration or exact performance metric used to calculate the uplift. The findings should therefore be treated as company-reported test results, not as a universal benchmark for predictive targeting.
Tinuiti, an agency partner, said the signals complement its Adaptive Commerce approach by helping inform strategy and responsive activation. That endorsement points to a potential role for predictive audiences in agency planning, although broader adoption will depend on results across different categories, media channels and advertiser objectives.
Data governance remains central
The expansion of predictive audience targeting also raises questions about how consumer purchase data is modeled, used and shared with advertising partners. The UK Information Commissioner’s Office says online advertising that uses tracking technologies or profiling to predict personal preferences and behavior may require consent under applicable data protection rules. Requirements vary by jurisdiction and implementation. ICO
For advertisers, predictive signals need to be evaluated not only for campaign performance but also for data provenance, permitted uses, audience transparency and privacy controls. Circana’s announcement describes the data scale behind the solution but does not detail the privacy mechanisms or governance controls specific to the new signals.
The commercial test will be whether advertisers can reproduce the reported gains across campaigns and demonstrate incremental sales or retention improvements. If validated at scale, predictive commerce signals could give marketers a more forward-looking way to allocate budgets. Until more methodology and results are disclosed, the reported uplift remains an encouraging but limited evidence point.
Market Landscape
Predictive audience technology uses data and statistical modeling to estimate the likelihood that consumers will take a future action. The Interactive Advertising Bureau has described predictive audiences as a way to combine multiple signals and build segments around the likelihood of a desired behavior. iab.com
Circana’s approach emphasizes purchase and household intelligence, using retail-related signals to inform advertising decisions. The wider market challenge is to turn predictive scores into measurable business outcomes while maintaining data quality, interoperability and privacy safeguards.
Strategic Outlook
Circana is positioning Complete Audiences around anticipated consumer behavior rather than historical purchasing alone. The reported results from 13 brands provide an early performance signal, but advertisers will need more detail about testing methodology, incrementality and repeatability before treating the reported uplift as a planning benchmark.
Future adoption will likely depend on how easily the signals integrate with media activation platforms, how well they perform across categories, and whether measurement can distinguish predictive targeting’s contribution from other campaign factors.
Top Insights
- Future-oriented targeting: Signals identify households likely to enter a category, spend more, switch brands or disengage.
- Large data foundation: Circana reports coverage of 129 million modeled U.S. households and nearly 2.6 million UPCs.
- Initial performance results: Tests across 13 brands reportedly showed a median uplift of 22%, with gains ranging from 11% to 36%.
- Measurement details remain limited: The announcement does not disclose the precise performance metric or full test methodology.
- Privacy remains important: Audience modeling and activation require attention to data provenance, consent and applicable privacy rules.
Get in touch with our Adtech experts
