Attempts have been made over the past few years to move to audience-based measurement of media campaigns instead of the impression-based system we have now. The idea is not new to Web advertising. Check the archives of many of the long-running discussion lists on the online ad industry, and you’ll find suggestions for moving away from impressions dating back to 1997.
I’ve resisted the urge to move Internet advertising toward the measurement models some of its more mature media brethren routinely use. For the most part, I don’t want to return to the concepts of reach, frequency, and gross rating points (GRPs), which pretty much dominated any discussion of effectiveness measurement in my traditional-media days. It used to upset me that a client could sum up the effectiveness of a media plan merely by saying, “We’re reaching x percent of our target audience an average of y times during the average month.” Part of what attracted me to the Web was the notion of knowing exactly how many targeted impressions were delivered without having to rely on estimates and educated guesses by “media research.”
Many would argue that the impression-based system is what holds the Internet from tapping into traditional media budgets. Those same people would say that the inability of the Internet to utilize the same metrics as traditional media keeps Web advertising out of serious budget discussions. I think there are a number of other reasons why our beloved medium hasn’t captured large percentages of ad budgets from brand advertisers, but I’m willing to concede that the impression-based system does make things a bit more difficult from the perspective of a client struggling to understand how Internet ads can contribute to an integrated communications plan. We should be looking for ways to make understanding the value of Internet advertising easier.
We’re about to take a nice step forward. This week, I saw an announcement that The New York Times Web site will debut a new advertising method. The Times will allow its advertisers to convey messages to a given user over a period in an exclusive fashion. Delivering up to three messages to a specific user per page over the course of a minimum of five page views constitutes a “session.” During a session, the advertiser’s message is the only commercial presence on the page. Should a user not reach five page views during any visit to the Times site, the session will not count.
Not only does this new measurement present some interesting possibilities for pricing (clicks per session?), but it also introduces advertisers to an audience-based model. The Times is known in the industry for having a particularly good grasp of its user base, as evidenced by its advanced targeting capabilities. No doubt it will soon be able to deliver targeted rating points (TRPs) against specific demographic audiences.
What’s really compelling about the new idea from the Times is that we’re not talking about impressions — we’re talking about reaching a site’s audience. The session format almost guarantees prospects will see a given advertiser’s online ad. Divide the number of sessions served by the number of unique users reached by those sessions, and you get an average frequency. Suddenly, we’re speaking a language traditional media planners, marketing directors, and brand managers understand. This is a step in the right direction. Implementing the concept of selling by sessions won’t preclude the Times from continuing to sell impressions to agencies that still want to buy that way. Smart thinking. It could help get some traditionally oriented advertisers online, which would only be a good thing.