Brand Metrics, a brand lift measurement vendor, now sells brand outcomes prediction to advertisers and agencies, estimating a campaign’s effect before it runs.
Brand Metrics describes the product as prediction built on a database of more than 60,000 campaigns and 13 million survey responses. It estimates the likely effect of a campaign on awareness, consideration, preference and action intent. The inputs are campaign delivery data and creative assets. According to the company, the product needs no publisher integration, no survey and no minimum campaign size.
Brand Metrics lists three uses. Before spend, a buyer compares media options and tests a plan against predicted outcomes. During a flight, the buyer compares live delivery with the prediction. After the flight, the buyer sees what the campaign achieved on all four metrics. The company frames the problem this way: brand lift measurement “relies on panels or requires publisher-side integration,” which leaves it unavailable across most of the campaigns agencies run.
Why it matters
Brand lift measurement has been tied to panels or to a publisher’s cooperation, which is why small and short campaigns often go without it. A prediction that runs on delivery data and creative alone moves an outcome estimate ahead of the spend. That puts a modeled number into planning conversations where an observed one used to arrive late or not at all.
Our read
A prediction is a model output, so the first test for a buyer is to compare predicted lift with measured lift on campaigns that also ran a survey. Ask any vendor selling predicted lift for its error ranges and for the share of its training campaigns that match your format and category. Brand Metrics’ own pitch rests on the size of its database, and that is the number to examine. For related measurement coverage, see Nielsen’s planned content measurement and the Rembrand and Omnicom Media retention study.
