OPINION: Stop data spin doctors with standardized traffic counts

By Published on .

Most Popular

The recent firestorm surrounding DoubleClick's use of Media Metrix audience numbers, and its resulting stock spike, drives home the need for standards and increased scrutiny in reporting Web site traffic.

Two weeks ago, in this section, we reported that Media Metrix would begin selling an ad sales network report, which would include traffic for DoubleClick, LinkExchange, 24/7 Media, Zulu Media and other ad networks. Until now, Media Metrix has reported only on individual site traffic, although it's done custom reporting for some ad network clients, including DoubleClick.

NUMBERS CRITICAL TO SALES

As online advertising buyers and sellers know, Web site traffic numbers often make or break deals, and monthly reports from the leading measurement companies are a critical element of the sales process. If your site doesn't rank on the monthly lists, you may have some explaining to do. Because for some advertisers (and investors), traffic is one of the only seemingly reliable ways of gauging how well an online property is doing.

Seemingly, because these numbers often are not what they seem to be. Every measurement company has a different way of counting traffic, and reports can vary widely. The DoubleClick controversy is just one more, very visible example of what can happen when sites try to manipulate numbers to their advantage.

As our story reported, and as DoubleClick stated in a press release issued June 30 (still posted at its site www.doubleclick.net), based on Media Metrix data, DoubleClick would have the third-largest reach behind AOL.com and Yahoo! if it were an individual site.

REACHING FOR REACH

For its release, and during interviews, DoubleClick took data from a custom report Media Metrix had created for it, showing its reach in April as 29.4% of the Internet audience. That was compared to the reach of AOL.com (46.3%) and Yahoo! (42.4%). That argument was supported by Media Metrix in our story.

However, in its release and in interviews, DoubleClick neglected to include the traffic for LinkExchange, a competing network that had a 32.8% reach in April, according to Media Metrix, which would make that network No. 3 if it were a site. The LinkExchange data were also taken from a custom report, and were not part of a published ranking list.

Following publication of this report and DoubleClick's press release, DoubleClick's stock price jumped more than 60%, demonstrating just how much clout these traffic numbers can have.

DAMAGE CONTROL

LinkExchange, meanwhile, brought out its damage control fleet, demanding retractions and letters of explanation from Media Metrix, which wrote a letter to clients stating only its methodology for the new report and emphasizing that data could be interpreted in different ways.

This episode underscores just how critical it is for the industry to develop not only standard ways of counting Web site traffic but also to create accepted practices on how to report and use data coming out of measurement services, particularly when it comes to comparing online properties. Without guidelines, sites and online properties are free to take data from measurement companies and spin them any way they can.

The Internet Advertising Bureau has already started this effort with a move to create voluntary guidelines for audience measurement on the Internet, and it will publish a document in September with its preliminary conclusions.

Beyond that, however, is the issue of reporting numbers. For publications, including Ad Age, the question is when should measurement data be reported, and just how reliable are these data?

VERIFICATION ESSENTIAL

It's one thing to demonstrate a site's growth by reporting traffic gains, with verification from a measurement service. But in reports that compare properties, including lists or other types of rankings, there is a very real danger of misrepresenting data unless an independent third party can verify it.

Publishing companies are neither equipped nor authorized to serve as auditors; therefore, the industry needs to have some mechanism in place to ensure that data it is providing for publication are accredited. Bringing third-party auditing into lists and rankings would help the industry's efforts to prove the Internet is indeed a reputable, accountable advertising vehicle.

Copyright July 1998, Crain Communications Inc.

In this article: