• by Jim Meskauskas , June 27, 2002

And then there were two…

There was a time when there was a plethora of site traffic measurement providers at the disposal of media planners, buyers, correspondents, VCs, and anyone else interested in seeing how much traffic a particular site received on an ongoing basis. PC Data, Media Metrix, NetCount, NetRatings, comScore…

I'm sure most of you out there know that there are now only two providers of site traffic measurement of any real consequence left in the market – comScore and NetRatings. Through consolidation and attrition, the market has whittled the site metering business down to just these two contenders.

So it seems the question is begged, does site traffic measurements really matter?

I've always had an ambivalent relationship with site traffic data. Through the years, most of the media planning I’ve done was done against objectives that rarely, if ever, had communication goals dependent upon a volume. That is to say, my clients never set up a reach goal against a given target. The advertiser never aimed to speak to a certain number of people over a certain amount of time. Sure, some clients wanted to be sure that their brand received as much exposure as their budgets would allow, but the objective was almost always to get an individual to answer a call to action. When a post-buy was done (a post-buy is the report an agency will do for their client that essentially audits the campaign – delivery versus contract, response rates, cost-per-conversions, etc.), rare was it that a client ever asked how many unique individuals were touched by the advertising.

In this context, the number of uniques visiting a site was incidental at best. Who cares if Site A has more people visiting it than Site B, as long as the property delivers the audience I'm looking for and can yield the sort of results I want. Site A may generate hundreds of millions of impressions, but what does that mean to a media buyer when selecting a site? Other than providing me with a way to determine what percentage of available inventory I'm buying, which might be important to ensure I don't "burn out" the audience on that site as a ratio of the impressions being generated, there was never a compelling reason to preference a site based on the size of its audience.

Even if a client did have a branding objective, Reach, as a quantifiable communication goal, was never something that could be delivered on. This fact renders size of audience and the impressions being generated by it irrelevant.

To date, there hasn't really been much of a need to care about the size of a site's audience.

This is all about to change.

With the advent of reach and frequency tools for online advertising, the measurement of a site's audience suddenly becomes tantamount to the project of online media planning. That is, the media plan is built on a reach/frequency goal, as most other media are planned.

As more and more brand-oriented advertising makes it way onto the web, and as more and more traditional advertisers demand that their online media efforts be put into the context of their overall media mix, knowing the size of a site's audience is a must. Not only that, but knowing things like duplication of a site's audience with other sites is required. This is where online metering services once again regain their primacy as essential tools for online media planning.

The remaining site measurement services are in a unique position, along with ad servers, which can provide campaign actuals (see Dave Smith's "Optimizing R/F via Ad Serving"), to deliver on the Holy Grail du jour of Reach and Frequency for web media.

These metering services are now participating in an initiative to build Reach/Frequency curves based on criteria set by the Advertising Research Foundation (ARF). Now is the time for them to get that message out and promote this as an advantage of their use.

"Hic iacet Arthurus, rex quondam rexque futuris" – "Here lays Arthur, once and future king."