"In order to be a successful company in this space, you have to either be a solution set for the advertiser or the publisher," Peralta said. "So we thought it would be best to focus on one thing, and that's being a solution set for the advertiser."
Peralta compared AudienceScience to enterprise software provider SAP in that it provides managed services to marketers and connects them with sellers. Asked what these new roles portend for agencies, Peralta said media shops will morph into being "system integrators."
That role is still a few years off, and Peralta acknowledged that there are a lot of other ad-tech companies who are also making the transition to an SaaS model.
For now, though the company charges clients an annual fee to use its platform, it still has to deal with the "old way" of receiving IOs from the agencies and clients.
"The majority of networks, DSPs and networks have fleets of ad salespeople who are calling media buyers all day trying to get a monthly IO," said Emily Riley, AudienceScience's VP for marketing and product. "That creates complexity – everybody has an incentive to complete the IO, whether it's good for the client or not. And the advertiser loses out in the end. We have a partnership with the advertiser directly, and with the agency. All media buying is passed through to the publisher at cost and the advertisers just pays us an annual flat fee to decide."
In other words, there is absolutely no incentive to increase the volume to get AudienceScience's margins up, or to use arbitrage in order to make money on the deal, Riley said.
Next steps: AudienceScience delivers 2.5 billion impressions every month. It also processes more than 10 billion impressions through the DSP/DMP side of the business. Video is becoming a bigger deal. In order to meet the scale, AudienceScience needs more engineers. While Peralta wouldn't discuss staffing numbers, he did acknowledge that the company had to shed employees when it closed the ad net business.
Since that time, the company has increased its engineering team by 10%. It plans to continue to ramp hiring on the client services and product side.
"By Q1 or Q2, we will probably have made up the loss of the network positions with engineering posts," Peralta said.
The 80% cut: AudienceScience's evolution started several years ago, when it changed its name from Revenue Science (yes, it used two separate words then) and created a behavioral targeting network in 2009. That was followed by its move into real-time bidding and building out its DSP and DMP business.
Ultimately, its global clients suggested combining the DSP and DMP into an integrated audience targeting and management platform. Clients also demanded that AudienceScience not take a margin. "And we want you to be transparent," Riley said clients told the company. "So we created a totally open platform, where they can buy and manage media for a flat fee."
Being that "one-stop shop" eliminates, on average, about 80% of the cost of buying media for AudienceScience's clients, Riley said.
"That's what doing away with being a 'black box' can add up to," Riley said.