"We determine on the fly if a viewer should see an ad," says Joe Zawadzki, Poindexter's founder and chairman. "We amalgamate a huge amount of anonymous viewer data all at once and use statistical modeling to cluster viewers based on their responsiveness. Our technology determines whether a new viewer shares the characteristics of a responsive cluster and should be shown the advertiser's message."
Working with data is the backbone of Poindexter's operation and Zawadzki says the company relies on anonymous cookie data plus information from third party data providers and client databases to generate the data used to compile media optimization reports for clients. The reports list responsive segments, the percentage of total media each segment represents and the percentage of total responses the segment is responsible for driving. For each segment, Poindexter identifies characteristics such as media placement, day part, geography and frequency exposure. Clients can then better target their media buys or request automated media optimization from Poindexter, which serves the ads to targeted viewers.
Zawadzki explains that a company would use Media Optimization by first sending out all versions of its ads to the entire viewing audience. Then the media optimization report is generated, which analyzes the different segments (for example, x viewers saw the ad on Yahoo! between 9 am and 12 pm in Los Angeles; they make up y percent of the viewing audience and deliver z percent of the responses).
Using this information, advertisers can target their ads to the best audience by eliminating sites with fewer responsive segments and negotiating more impressions on the best responsive segments. For automated media optimization, Poindexter instantly gathers information on the viewer and decides whether the viewer should see the ad. If not, Poindexter returns the impression to the publisher.
Zawadzki says Media Optimization works with Creative Optimization, enabling advertisers to determine which prospects and creative executions work best.
Bill McCloskey, founder/CEO of Emerging Interest, says the problem with optimization has been that it required a heavy manpower commitment and wasn't cost efficient. "It was difficult to swap things out in a timely fashion and gather data to make a judgement. It took too long," he says. He lauds Poindexter for developing software that "makes it cost effective and you don't need a huge team to manage it."
Poindexter will sell Media Optimization on a monthly or quarterly basis, with prices reflecting the amount of media a company buys. Companies can save money by buying less media, if the optimization reports determine fewer ads should be sent. Or they can buy the same amount of impressions, but target them to the best prospects to boost profits.