Upwave launches closed beta for surveyless Predictive Brand Lift tool

September 12, 2026 · ppc.land

Chris Kelly, chief executive of Upwave, described the closed beta for Predictive Brand Lift as opening "a new generation of Brand Lift measurement." The tool estimates campaign impact using the company's proprietary Bayesian modelling, artificial intelligence, and observed data gathered across more than 3,000 brands, arriving after more than a year of development and testing with selected customers. Traditional brand lift studies require an exposed group and a control group to each complete enough survey responses for a statistically defensible result, a bar that leaves many smaller campaigns unmeasured. Predictive Brand Lift generates an estimate from patterns in Upwave's prior studies without fielding a survey. The company has not disclosed what priors its model uses or published accuracy figures comparing predictions against observed results, and it has given no date for wider availability.

Campaigns that previously sat below measurement thresholds gain a route to brand lift scores, letting smaller budgets get evaluated where survey-based vendors would set a floor and skip them. Buyers comparing an observed number against a predicted one in the same dashboard need to know which methodology produced which figure, since a modelled estimate carries different evidentiary weight than a completed survey. IAB research from February 2026 found up to 75 percent of buy-side decision-makers rating current measurement methods as underperforming, which gives predictive approaches an opening. A model trained on 3,000 brands' worth of history can only reproduce relationships already present in that data. Rivals including LoopMe, Utiq, DISQO and Cint have lowered thresholds or sped up survey-based results, keeping the survey in place. Upwave's bet on a fully modelled alternative sets it apart from that approach.