• by October 31, 2003

On August 10, 1998, Media Metrix, the renamed PC Meter Company, released its study on the Top 50 Fastest Growing Web Sites on the Internet. The five top sites were Angelfire.com, Xoom.com, Preferences.com, TheGlobe.com, and Hotmail.com, all of which, it was specified, were community-oriented services. One of them clearly wasn't, however, which was Preferences.com. Preferences.com was the ad-serving domain of ad serving company Match Logic, owned then by Excite. A visit to www.preferences.com yielded a short course in online privacy as it related to Match Logic's ad serving technology. Telephone calls and emails to Media Metrix made at the time to address this, and other questions regarding their rankings, went largely unanswered, except for the perfunctory, "Thank you for your message. We are investigating."

Not much has changed since 1998 with regards to Media Metrix and the veracity of its rankings, notably the Top Properties report released again with considerable coverage last week. The ranking, which appears also in the "Key Measures" report in the MyMetrix media planning tool, lists the Internet's "biggest web sites," and - since this is America - the implication is probably also that the biggest are the best. "Top," one could say, is not just about quantity, but also quality. Scroll through the Top 50 U.S. Properties announced by Media Metrix last week, however, and any semi-experienced Internet user is going to question Gator.com as the ninth biggest site online; or Overture.com as number 19, or Weatherbug.com as 30. Most of the properties on the Media Metrix Top Properties list are not even properties, but groups of properties and, in some cases, the reported traffic for a property is simply the aggregation of advertising impressions they place on web sites they do not even own. Hence, it is not Gator.com, but the "Gator Network" occupying the ninth most frequently visited site position on the Top U.S. Properties list.

Is the issue of consistency with regards to Internet media measurement important? We think so. On the one hand, if we're measuring the aggregation of ad impressions across web sites why not include ad networks? What excludes them? More critically, this is not 1998 and hundreds of millions of advertising dollars are spent online today ostensibly supported by Media Metrix numbers.

We have put the question of consistency to Media Metrix regularly for the last year and a half arguing for a Top Properties definition. We have proposed, essentially, two alternatives: 1) stop calling it the "Top Properties" list (because it isn't) and call it the "Top Groups of Locations" list (because it is), and then make room for numerous other Internet media "Groups," or 2) define Top Properties in a way that after it is publicly documented will make clear to us and the rest of the world how we, and many others, fail to conform to the Top Properties requirement.

After a year and a half, Media Metrix finally presented us with a definition for the Top Properties list this month that offers the kind of clarity benefiting everyone. Top Properties, they say, must represent a "list of destinations as viewed from the consumer experience." Starting at the top with the largest, and continuing in descending order, the Top Properties list must reflect the unique places people "choose" to go online. It is a rank of web site destinations.

Write that down.

"A list of destinations as viewed from the consumer experience" is a suitable definition. It is clear and easy to understand. Who can say what some clever person might be able to squeeze from the definition in order to drive a truck through it, but we have been told what the Top Properties list ranks now: it ranks consumer destinations, not consumers. Accordingly, the Top Properties list as it appeared last week must change.

No one feels the tangle of the Top Properties list more than the comScore Media Metrix team. These are the bonds that bind and then constrict. The loose, evolving structure of the list has trapped all of us, them especially. They are - we're sure - working to find a reporting solution that is appropriately scientific and objective and resourceful enough to help Internet media compete, openly and aggressively. Definition is the first requirement. Transparency is the next, which should force other changes to the Top Properties list, two of which come to mind:

1. End roll-ups on the list that aggregate web destinations under a family corporate tree. Create a different media planning report for corporate, bundled media. Consumers don't visit Viacom Online (#12); www.viacomonline.com doesn't exist.
2. Audit the auditors - no one should have to simply accept the testimony of measured companies with regards to measured impressions. Consider the unsettling news (really, who's surprised) last month about brokered ad impressions that wound up in unintended places. Brokered impressions are random. They are not impressions that are for sale; they are impressions that are bought and have no place in our measured rankings. Independent auditors (BPA, for instance, or an independent trade group) should regularly validate what is being counted and reported to the measurement companies.

In an advertising economy that is frantic about accountability strong tools are required, and our ratings services must be the strongest among them. Sadly, endorsement, not measurement, has dominated the process for years online. And money. The Media Metrix Top Properties list of today resonates with the echoes of yesterday - venture capital, IPOs, Wall Street analysts - when the Internet was largely entrusted to the market makers. In a post correction Internet economy, however, we've put our trust back in the fundamentals. It's time for them to get put back into measurement, as well.