A Better Yardstick

Publish date:

For as long as Wall Street pundits like myself have been declaring the Web "the next mass medium," Madison Avenue advertising types have been demanding, begging and pleading for better measurement of Internet audiences.

But what I think they've really wanted all along is not so much better measurement but more uniform measurement. I would argue that the shortfalls in Web audience measurement to date are not all that different from

Nielsen Media Research's

inability to measure audiences who watch television outside the home -- be it a sporting event in a bar or the evening news in a hotel room.

Rather, the real problems dogging Web audience measurement have been a wide range of difficult-to-explain metrics -- hits, page views, unique users, click-throughs -- generated by a wide range of unproven information providers. Differences in methodology and metrics have historically yielded vastly differing results among the competing ratings firms. Further compounding the problem is that many of the Web metrics don't synch up apples-to-apples with traditional media audience measurement -- which continues to focus on frequency and reach.

These realities have provided little comfort to agencies and advertisers looking to do business on the Web. This week's merger of New York-based

Media Metrix


PC Meter

) with Atlanta-based

Relevant Knowledge

combines the two leading providers of Web audience measurement. It should hopefully go a long way to solving at least one of those problems by creating a large, unified provider of Web measurement and one that is potentially positioned to dominate Web measurement the way Nielsen has dominated television ratings,


in radio and


and the

Audit Bureau of Circulation

in the print medium.

The combined company, which will be called Media Metrix, with the tagline "The Power of Relevant Knowledge," seems to be nearing critical mass, offering coverage of more than 15,000 Web sites and online services and providing audience measurement data on a monthly, weekly and real-time basis.

Whereas each company had its own particular shortcomings -- Relevant Knowledge didn't effectively measure online services like



, while Media Metrix didn't measure use at the office -- the combined company will boast a panel of more than 15,000 Internet users. While would-be rival Nielsen has been promising an Internet measurement service for a number of years, the offering has yet to hit the market. The Media Metrix-Relevant Knowledge merger obviously hopes to preempt another player from moving into the Web measurement space in a big way.

But even more important to Madison Avenue than receiving data from one trusted vendor will be the ability to start to look at Web advertising on a competitive basis -- by comparing it on an apples-to-apples basis with other media such as broadcasting, cable TV and print and not just against itself (e.g. comparing www.news.com vs. www.yahoo.com).

Sources at both Media Metrix and Relevant Knowledge as well as at a number of ad agencies believe this will provide the true key for the Web taking its rightful place against other measured media -- and not simply being relegated to an incremental buy when budgets permit or, even worse, some type of ongoing marketing test.

The measures I expect Web sites, ratings firms, advertisers and their agencies to focus on going forward will be unique users, page views and duration, which are roughly equivalent to frequency and reach. Whereas reach measures the percentage of a particular demographic -- let's say women 18 to 34 years old -- that an advertising campaign reaches over a given period of time, unique users similarly measures the number of different people who visit a site on a daily, weekly or monthly basis. Whereas frequency measures how often a viewer will see a particular advertisement over the course of a campaign, page views similarly give a sense of how much time a unique user spends clicking around a particular site and what pages are actually pulled up. We expect this measure to be increasingly refined by looking at duration -- or the amount of time spent by a user on any given page.

Finally, better measurement is key to the long-term well-being of Internet media stocks. Without a higher level of comfort with regard to audience delivery and measurement, advertisers will never fully embrace the new medium and the Internet will not take its rightful place among measured mass media. In the absence of broad-based advertiser/agency acceptance, the ad-based Internet business model cannot be fully realized.

Paul Noglows is a senior analyst in Hambrecht & Quist's Internet Research Group. Hambrecht & Quist has a buy recommendation on CNET and has been an underwriter for the company. H&Q has a hold recommendation on Yahoo! and has not been involved in any recent underwriting for the company. H&Q has a strong buy rating on AOL and has not done any recent underwriting for the company. Noglows welcomes your comments, which you can send to