Predictive Advertising Targets Consumers on Purchase Data

A network of e-commerce sites and product manufacturers intends to target ads across the Web using predictive modeling based on consumers’ past purchase behavior. For advertisers, it may be an attractive offering, but for consumers it may raise privacy concerns.

Executives from aCerno, the network, said they spent two years building the network designed to share data across retailers on a single cookie to provide continuous modeling across all the sites.

“We gather all of [an e-commerce site’s] users and behaviors, all anonymously, to then target all of those users across all of the Web to drive them back to all of the member sites to create more transactions,” said Bennet Zucker, VP of sales and market at aCerno.

Last fall the network opened to advertisers and agencies, when aCerno began using anonymous data gathered in user cookies to serve targeted ads across publisher sites and ad networks. It’s a change from the behavioral marketing model where users are targeted based on recent search behavior. ACerno instead categorizes segments based on predictive modeling and matches advertisers to an audience segment.

“We’re able to take those same audiences, package them up, customize those segments for brand advertisers, make deals with agencies, [and buy] CPA deals with target sites,” said Zucker.

Behavioral targeting competitors, such as AlmondNet, say using purchase data is too late to reach consumers through advertising. But it may have another use. “If someone has already purchased, that information may be used to upsell and offer a complimentary product,” said Roy Shkedi, CEO of AlmondNet.

ACerno spent two years building the network while operating in stealth mode. It now has 375 online multi-channel retail and product manufacturer Web sites in its network. Since last fall it’s begun buying for 450 advertisers using the purchase data from the network’s e-commerce sites. As a blind network, members are unknown even to each other. ACerno refused to publicly name retailers or advertisers.

ACerno buys ads from portals, major publishers, and ad networks. It buys on the top ad networks ranked by comScore, and matches ads in real time based on cookie data at the time the ad is called up.

Consumers concerned about privacy are likely to worry about being targeted by such a network, if they are aware they are being targeted. Zucker said privacy policies on e-commerce sites are very specific about cookies and what can be done, and consumers are becoming more savvy about Web practices. “The ad network thing is something that’s a lot newer to the sites. We are working with [members], and coaching them. We go to great pains, as do the retailers, to make sure customers realize that we are only working with anonymous bits of data,” he said.

Privacy appears to be a priority for aCerno. Tom Sperry, the company’s CEO and chief privacy officer, sits on the board of the Network Advertising Initiative. ACerno is also listed as a participating network on the NAI Web site. ACerno’s parent company, I-Behavior, is a database marketing solutions provider and behavioral targeting firm for multi-channel marketers and advertisers.