"We don't rely on publishers or anyone else to execute the service," says Robinson, who notes that Affinity doesn't consult subscriber lists or circulation figures. Instead, they query the public. "We have a panel of about 2 million people who we're constantly screening for what titles they read. When it's time to do a study for a particular title, we just go to the readers who have been reading that title recently."
CopySource uses a large pool--subscribers, newsstand buyers, public place readers, and pass-along readers. Robinson says the company asks readers how they get their hands on a specific issue. In addition to baseline questions of issue origins, Affinity also delves into the specific details of each kind of acquisition. "We then route them to a custom questionnaire designed for that subset of readers," says Robinson. For example, CopySource asks newsstand buyers, "Did you intend to buy this particular magazine? What other magazines did you buy along with it?'" For public place readers, the service asks: "What prompted you to grab that magazine from the stack?"
The CopySource data complements Affinity's existing syndicated reports, which measure ad effectiveness across different genres and titles. Robinson says these reports are used by media agencies, including StarCom, MediaVest, and Mediaedge:cia, positioning them as a currency to measure "ad engagement."
Affinity investigates the effectiveness of 30,000-35,000 specific print ads a year, posing questions to the online panel.
Robinson says the company uses a battery of questions, called "actions taken," which were developed with agencies and publishers. They cover a host of actions that readers might take. For example: "Do you have a more favorable opinion?" "Did you visit the Web site for more information?" "Did you consider a purchase?" "Did you make a purchase?" Then we net that out for a total actions-taken score."
The addition of CopySource data allows Affinity to distinguish between the effect of ads in paid and nonpaid magazine reading. Robinson says his company can "tabulate the data and break out the differences in ad effectiveness between subscribers, newsstand buyers, pass-along and public place readers." According to him, there is a lower overall index of actions taken for nonpaid versus paid. "Public place tends to be lower," he notes, "probably because there's less overall time spent reading a particular issue."