John Potter, chief technology officer, Purch Most publishers are slow to adopt server-to-server connections because they don’t have the technical know-how to integrate and scale these products. But under the direction of John Potter, tech network Purch — which runs websites like Live Science and Top Ten Reviews — built its own server-side product and has sold 100 percent of its display inventory through it since November. Potter says going to server-to-server was the only way Purch could bring on enough demand partners to get the true value of its inventory. Purch has already integrated 30 demand partners into its server-side product without slowing its load times. By the end of the year, it will likely add another 10 partners, Potter says. — Ross Benes
By building its own solution, Purch can see every scrap of bidding data, including both winning and losing bids, and the spread between them. Potter considers data key to publishers controlling their inventory and understanding its true value.
While Purch’s move to the server side made sense for the company, Potter doesn’t expect many other publishers will take the same path. For one, publishers must have enough scale to entice their ad tech partners to build their end of the connection. And most publishers don’t employ ad tech engineering teams to build these connections.
“Publishers will have to wait for someone to productize it for the most part,” Potter predicted. “I think this is going to be slower than header bidding because it’s harder to implement.”
That said, Potter says Google’s exchange bidding product could see swift adoption and publishers will realize the gains of speed that come with server-to-server connections.
“When exchange bidding is rolled out as a full-fledged product to people,” he said, “it will go quickly.”