• by Mark Naples , September 30, 2004

Or, more to the point, in measuring the reach of interactive, does the 80/20 rule really apply to our medium more than we care to admit? And if it does, how do we accommodate that? After all, if the study released this week by comScore Networks and DoubleClick, "Internet Audience Dynamics" is accurate, then a jagged edged star may better represent our medium than a flush plume of coverage, compared to television.

The comScore Networks/DoubleClick study analyzed the dynamics of online audience accumulation (or the rate at which a site attracts incremental unique visitors over time) and its implications for scheduling and packaging online media. Basically, it stated that since such a high percentage of site traffic comes from such a low percentage of visitors, media planners must take into account that actual reach and frequency data is comprised much differently online than it is in other media.

This may seem intuitive to those of us in the industry - may being the operative word here. The fact is that any uniform comparison between interactive and other media will probably not tell the whole story without consideration of the differences between and among the media in question.

Either way, I really applaud DoubleClick and comScore Networks for their methodology and rigor here, since this study may actually assist decision makers buying across media. If nothing else, it should make sense for media buyers frustrated by inconsistent results.

The keystone fact is intuitively apparent enough - the vast majority of Web page views come from a minority of Web viewers. This is not exactly news, nor has it ever been in the decade or so I've been in this business. It does seem, however, that the numbers are getting a little bit closer.

According to this report, heavy Internet users (39 percent), described as those who go online 19 days or more in a month, account for 73 percent of all page views. In broad terms, two out of five Internet users pull down three out of four Web pages. Data like this is important to advertisers who, the study suggests, may not be using marketing dollars efficiently. (I'm wiling to bet that those 39 percent of users also are responsible for 80 percent or so of all Web commerce. Maybe that will be in the next study.)

Few advertisers or Web publishers take advantage of data that shows site audience levels vary throughout the day, which often is easily explained by the content. For instance, entertainment sites will see higher usage on weekend evenings, while business sites see peaks in the morning and at the end of day.

Anyone remember the industry focus on dayparts a few years ago? Forbes, CBS MarketWatch, USA Today and The New York Times joined to form an alliance of sites that was designed to provide inventory targeted to those business-centric users. I'm sure one of you will tell me.

"In both television and print advertising, audience accumulation is always a factor in assessing reach, and the same principle can be applied to the Web," said Lynn Bolger, executive vice president, agency development at comScore. "Total monthly audience data is just one way to determine how a site can deliver reach -- online audience accumulation must also be analyzed to determine which day of week, times of day, and types of content perform best given advertiser goals."

The study makes strong recommendations for buyers interested in accommodating vagaries in audience accumulation. Among them:

· Buyers should balance delivery among heavy and light Internet users;

· Since high reach is not always consistent with high composition, placement on high-reach sites should be complemented with high-composition sites;

· Scheduling tactics should be directed by audience dynamics, accommodating dayparts and usage patterns as precisely as possible; and

· 'Audience build' rates vary by site and should drive scheduling options.

In other words, branded sites will generally drive the highest mindshare due to higher penetration and recognition, which should lead to better audience build. One of the things I like about our business is that its ability to quantify results justifies a claim we've all made at different points - that marketing is more about science than art. Whenever someone comes up with a study as strategic and direct as this one, I think it goes a long way toward building our industry. So good for comScore and DoubleClick.