As recently as three years ago, retailers or brand agencies buying online display ads paid for a certain number of ad impressions up front—say, $200,000 for 10 million impressions over the next three months—and crossed their fingers that they’d generate good returns, says Sheldon Gilbert, CEO of online advertising services company Proclivity Media.
But now ad sales take place as auctions in open marketplaces, like stocks on the New York Stock Exchange or the Nasdaq exchange, he says. Proclivity’s ad-buying platform automates the process for retailers, but with a catch. Rather than looking for the most impressions at the lowest price, Proclivity matches all of a retailer’s e-commerce data—everything from customer details to inventory—with the ads that are most likely to drive sales, then bids high to win them.
Proclivity uses its Consumer Valuation Platform to place cookies in consumers’ web browsers to monitor their browsing behavior around the Internet and tracks their specific interactions on a client retailer’s site using tiny pieces of embedded software code in site content. Proclivity adds data from the retailer, including the merchant’s own web analytics on shoppers’ click activity, and information on sales, merchandizing campaigns and product pricing, then scores it to determine when each customer is likely to buy and at what price point, Gilbert says. With such data on hand, clients can then use the ad-buying platform to automatically bid on targeted desktop, mobile and video ads.
“The [software] engine does the bidding automatically for them,” Gilbert says. “They put in a budget and say they want this ROI back.” The Internet-hosted bidding engine can connect to several display ad exchanges at once, he adds, including Google’s, Yahoo’s or ESPN’s. Because the method is based on predicting the actual return on investment for single ads customized for individuals, clients on average generate sales five to 15 times what they spend on online display ads, Gilbert says. Beauty products e-retailer Carol’s Daughter reports generating average sales equal to 10 times its spend on digital display ads, says Jessica Klein, the retailer’s director of e-commerce. Those sales amount to roughly 12% of the retailer’s total online revenue, she adds.
“Proclivity processes all CarolsDaughter.com clickstream data and predicts each site visitor’s individual likelihood of purchasing the products available for sale online,” Klein says. Then, by bidding to place ads tailored for individual customers, Proclivity is able to maximize ROI enough to practically eliminate all of the retailer’s risk in buying online ads, she says.
With instant trading backed by analytics on shopping behavior, a retailer might see that a customer has been looking at a $500 sweater and determine she has a 50% chance of buying it in the next five days, based on its own data on her past shopping behavior. So, after calculating its profit margins, the retailer might figure it’s worth bidding high, maybe even close to $250, for an ad shown to that shopper, Gilbert says. Not only is that ad likely to pay off with a sale, but because the trading happens in a second-auction format—meaning winners pay only one cent more than the second-highest bidding price—the retailer will probably only pay a few dollars because of the typically low bids placed in auctions, he says.
“Everyone in the ad industry, when they buy impressions on these exchanges, they’re bidding pretty low prices, $1, $1.50,” he says. “So there’s no risk for bidding really high.” And because web site publishers don’t like seeing their ad impressions trading for bargain basement prices, he says, they may invite retailers known for bidding relatively high into a private exchange to get first dibs on ads. Klein, for example, says that using the Proclivity ad-buying platform has helped Carol’s Daughter buy ad space from premium publishers like Conde Nast and Hearst Brands and generate a favorable ROI.
The platform has also led to an unexpected finding: The cost of media in overseas markets, like in Pakistan or Peru, can be around 70% cheaper despite the fact that foreign consumers are equally likely to buy, Gilbert says. Take a consumer looking at a blender, for example, who then leaves the site to read the online Huffington Post. Domestic ad buyers might value her more for seeing her on the American news site, causing higher bids for an ad targeting her on the Huffington site. But if she leaves the site for that of the Beirut Times, the cost of an equivalent ad there goes down “because no one in Beirut knows she wants to buy a blender,” he says. So an American blender-selling retailer will likely have little competition buying ads from a Lebanese publisher to target her. As a result of such activity, Proclivity has been helping clients which already ship internationally to expand their online sales in foreign markets, Gilbert says, including Peru, Bolivia, Singapore and Australia.
Proclivity offers clients a choice of pricing models, including a revenue share on sales or a price per click. A typical test of the platform runs around $20,000, he says, for which Proclivity guarantees a certain ROI. “All of our clients have increased spending once they use the platform,” he says.
Analysts Joanna O’Connell and Michael Greene from Forrester Research Inc. assert in their recent report, “The Future of Digital Media Buying,” that in the near future digital media buying will happen in one of two ways: via automated buying platforms, as with Proclivity, or through media buyers who work with publishers on behalf of advertising clients to create custom, integrated content for a brand or merchant.
Proclivity began as an e-mail optimization business about four years ago, then saw the opportunity in ad marketplaces, Gilbert says. It launched its first bidding platform at the beginning of the year and is on track to be profitable soon, he says, declining to be more specific. “Retailers aren’t aware of their power with this data,” he says. “While the rest of the market is playing checkers, retailers can play chess.”