Despite the dubious moniker, a name that reflects network engineers' penchant for working dinners of pizza, media buyers were impressed by the site's traffic statistics and demographics.
But the buyers, representing 3Com through Greco Ethridge Group, Boston, ultimately decided against the buy because extracheese.com used no third-party traffic statistic auditing to back up its claims.
"If they had invested in an audit to prove to me who these people are, we would have done business with them," said Paul Marobella, chief relationship officer at Greco Ethridge. "I think the sites that use auditing have more credibility."
Auditing vs. ratings
While auditing companies such as ABC Interactive verify a traffic server's measurement, ratings services such as Media Metrix use panels that reflect the Web population to measure site traffic.
Companies can choose to collect their own traffic data and have it audited by ABC or BPA International, for example, or can use the traffic rating firms to collect data and then choose to have the members audited.
"Virtually every medium out there has some sort of measurement and verification associated with it, and with that, the audit becomes increasingly important," said Peter Black, VP of BPA International, New York. "The process just confirms that the advertiser is getting what they pay for."
Mary Beth Miles, a manager of communications for the Audit Bureau of Circulations and ABC International in Schaumburg, Ill., said: "A ratings service is what we could audit. We are basically an umbrella service for other services conducting initial measurement."
The ratings services provide a barometer of how much and what kind of traffic that top sites are getting, but they have credibility problems of their own that the interactive advertising industry is trying to remedy. The handful of companies gathering Web traffic statistics or auditing business Web site logs produce vastly different statistics.
Because traffic research companies use different methodologies, traffic numbers vary widely.
For example, CNN's sites had a May reach, or monthly unduplicated visitors, of 2.47 million users, according to Media Metrix; 5.62 million users, according to Relevant Knowledge; and 11.79 million users (a monthly number based on spring-summer data), according to @Plan.
Allegra Young, research manager at USA Today Information Network, said the differences in monthly statistics have been as much as six times greater from one survey to the next.
The statistical quagmire has confused media buyers, misled investors and confounded Web site owners, and ultimately has driven a wave of change in the community.
"A lot of different numbers are floating around there," said Rich LeFurgy, chairman of the Internet Advertising Bureau in New York. "The Association of National Advertisers did a survey [of media buyers], and they said that a lack of reliable measurement was the No. 1 reason why they didn't buy on the Internet."
Web traffic guidelines
The ongoing problems have spurred two advertising organizations, the Internet Advertising Bureau and the Advertising Research Foundation, to create voluntary Web traffic measurement guidelines, including standardized panel methodology.
Two significant flaws in the methodology are that key Web audiences, such as business users and international users, are not counted proportionately to the actual Web audience. The statistic companies are finding it difficult to sign on business-based panelists because companies are generally unwilling to install the necessary software.
The unreliable statistics and the confusion they cause resulted in Web advertising getting only an estimated 0.3% of the advertising pie in 1998, a number that would be significantly higher if the pressing questions were addressed, experts say.
Industry media buyers use the information as only one of the many indicators of a buy's potential.
"We buy advertising for reach at this point," said Bob Habeck, VP-director of new media for DDB Needham Worldwide, Chicago. "We don't put much faith in the numbers. [Traffic ratings] mean a lot to sites at this time, but we're more concerned about categories than sites."
Mr. Habeck said the top three indicators for making a media buy on the Web are past performance on ads for a specific client, Net reach based on traffic statistics, and prior relationships with sites, including preferred pricing and positioning.
Karen Anderson, media director for Modem Media Poppe Tyson in Westport, Conn., said, "What [Web traffic reports] do for media buyers is offer some perspective. It's an interesting piece of info, but it's not a decisionmaker. If we're advertising on a site, if we continue to see good click-throughs, if advertisers' results are strong -- that's what is important."
As technology matures, media buyers are getting more accurate information and are able to analyze the performance of their buys after the campaign.
"For the people that complain that Net measurement is not reliable, they're right," said Chris Charron, senior analyst at Forrester Research in Cambridge, Mass. "It's better than other media, but, relative to the ideal we have set for this media, we have fallen short. That's the bottom line."