Not only did the ranking place a competitor well above First USA on the charts--the result conflicted sharply with reports from her company's Web data provider, Nielsen/NetRatings.
"It came down to differences in the methodology," said Wing, who had previously stopped buying Media Metrix's Web measurement reports in favor of Nielsen/NetRatings. "I'm just more comfortable with NetRatings' methodology over Jupiter Media Metrix's."
First USA isn't alone. Just over a year after Jupiter Communications and Media Metrix announced a merger aimed at creating an online research powerhouse, some of its customers are succumbing to the dot-com bust while others are favoring rival services such as Nielsen/NetRatings.
To be sure, customer loyalty is fleeting in the much-maligned Web traffic measurement business, and Jupiter executives are quick to offer their own list of clients wooed from Nielsen/NetRatings, including eBay and CBS MarketWatch. But analysts say Jupiter is more vulnerable than ever to deep-pocketed rivals as it struggles with integrating its research and data businesses at a time when demand for its services appears to be waning.
In an interview, Jupiter's interim chief executive, Tod Johnson, defended the company and its methods, arguing that it remains very competitive. "We think that we're doing well; we're retaining our dominant market share," he said.
Nevertheless, with its stock price slipping below US$2 a share, the economic downturn has hurt Jupiter as much as many of the Net companies that are the primary focus of its research, some analysts say. Harsher critics add that stiffening competition and lingering questions about the relevance of its traffic research have helped the company squander some of its advantages as runaway leader in the Internet measurement market.
According to sources close to the company, management missed a self-imposed June deadline to find a full-time replacement for interim chief executive Johnson, who has been dividing his time for the past year between Jupiter and his other job as CEO of NPD Group--a privately held research company that spun off Media Metrix in 1997.
A company representative would not comment on whether it had hoped to pick a CEO by last month.
Industry watchers also point to numerous difficulties facing the US$414 million deal that created the company last June. At the time, Jupiter announced the combination with a grand vision to use raw data on Web site usage to provide detailed analysis and insight into the workings of the Net. Once expected to create a US$1 billion conglomerate of data and analysis expertise, the deal has instead resulted in a company with a current market capitalization of US$53 million, based on a recent share price of US$1.49. The shares have traded as high as US$29.18 in the past year.
Johnson painted a positive picture of the merger, saying the integration has moved ahead according to schedule. "You don't just slam things together and hope," he said.
Some analysts take a different view, however.
"The stock price speaks for itself," said Eric Apfel, an equity research analyst at Dresdner Kleinwort Wasserstein. "The vision that they had of blending Jupiter analysts and Internet data and cross marketing those services--that hasn't worked out."
Although the importance of Web rankings has waned with the collapse of the Internet bubble, online research services continue to wield influence, helping allocate advertising dollars and providing bragging rights for corporate executives. Regardless of who comes out on top, consumers and Web companies may eventually see benefits from competition in the form of more reliable statistics in a young industry long wracked by doubts about the accuracy of its data.
All of the major Web traffic measure firms have faced criticism over their methods, but Jupiter has been singled out in two recent incidents. Last month, for example, it included traffic counts for a form of online advertising that automatically launches a browser window, called a pop-up ad. That decision pushed miniature-camera maker X10 onto its list of the most-visited sites on the Web, along with portals such as Yahoo.
Nielsen/NetRatings filters pop-up ads from its results.
Also, Jupiter last year was forced to revise its rankings for the month of August after it incorrectly listed the eFront network of Web sites among the 20 most-visited destinations on the Internet. Jupiter later learned the traffic count for eFront was erroneous and issued a correction, but the company did not alter its methods for verifying the accuracy of its information.
Jupiter stands by its research, saying the August revision is the only time it has changed its monthly top 50 traffic count.
Nevertheless, online advertising experts say data-gathering practices of Web measurement firms are under more scrutiny than ever, raising new concerns over the value of the research and creating a potential competitive opportunity.
"With the dot-com collapse, now many of these problems with the methodology have been exposed," said Donna Hoffman, professor at Owen Graduate School of Management at Vanderbilt and a specialist in Internet marketing and e-commerce.
One of the biggest winners should Jupiter stumble is audience tracking giant Nielsen Media Research. Fearing the impact of the Internet on television viewing habits, the company two years ago set its sights on the online statistics market, joining forces with Web measurement company NetRatings to track Web traffic. They later added ACNielsen to the fold to produce international ratings.
Jupiter remains by far the largest Web traffic firm, with 2,200 clients according to a recent filing with the Securities and Exchange Commission. That compares with 750 for Nielsen/NetRatings, which has doubled its business since last year. In the same period, Jupiter increased its customer list by about 500, although it has lost more than 100 clients in the past few months, according to the regulatory filings.
Many of those lost accounts were dot-coms that went out of business. But a number of them, including News.com publisher CNET Networks, decided to abandon Jupiter in favor of Nielsen/NetRatings' measurement service.
"The tide is turning," said Rebecca Young, vice president of marketing for Nielsen. "You look at the financials and that's another big indicator."
Some customers that have switched said differences in the way the companies compile their data played a role in the decision. Three companies, including a major Web portal that asked not to be named, said they decided to drop Jupiter because of statistical discrepancies that the company could not reconcile. Some sources also said that Nielsen/NetRatings' metrics more closely matched their own internal site metrics.
Douglas McFarland, president of measurement services for Jupiter, denied that its data falls short compared to competitors. "We measure the full digital media landscape, including (America Online), ICQ and Word. We have the largest client base and have been at it a lot longer and have a larger research expertise," he said.
Both Nielsen/NetRatings and Jupiter use panels of Web surfers, who agree to allow their online activity to be tracked. The companies then extrapolate from the sample to estimate Internet-wide usage patterns.
The size of the underlying panels is one of the keys to the accuracy of results, particularly among less trafficked sites, where small variations can be greatly magnified.
Nielsen/NetRatings has about 225,000 panelists in 26 countries, including 68,000 in the United States. Jupiter's panelists number about 120,000, with 60,000 in the United States and the remainder split between 11 other countries.
Wing, for one, said she has doubts about the exactness of all Internet traffic data, but said she felt Nielsen/NetRatings' panelists represented the population on the whole better than competitors.
For example, Jupiter combines at-home and at-work usage for its results, a method that pushed American Express into the top 50. Nielsen/NetRatings offers combined and separate lists, providing more detail for customers, she said.
"People question both types of methodology and whether they're correct. If we matched up our actual usage, it's a good approximation, but it's not going to always match up 100 percent to our actual numbers. I'm just more comfortable with Nielsen's margin of error over" Jupiter's, said Wing.
Steve Coffey, chief development officer of Jupiter, said that the company has not lost its competitive edge. For example, he said, the company is now selling traffic reports that offer greater detail on subcategories such as Yahoo Travel.
"We're listening carefully to our clients to meet the changing needs of the industry," he said. "People are still buying the data, but for sure there's a change in the overall industry."
Beyond questions about the relevance of its research, Jupiter faces significant business and financial troubles, analysts say.
Like other Net companies, Jupiter has shed a sizable part of its work force, losing nearly 25 percent of its staff in two rounds of layoffs. In addition, several of its research analysts have left the company or are preparing to leave--a sign that some say it may be losing its intellectual firepower. Mike May, a longtime e-commerce analyst and research director, and Melissa Shore and Claudine Singer, two other research directors, left recently.
Among those slated to leave is interim CEO Johnson, who announced his resignation in March pending an executive search--a process that Johnson said is close to completion. Company president Gene DeRose also stepped back from his internal duties to focus more on external business development.
"They're hemorrhaging people--good people either leaving by choice or being laid off," said Tim Clark, a former analyst at Jupiter who says he quit the company because of what he considers its poor prospects.
The departures come as Jupiter faces tough competition.
Industry watchers say that Jupiter's chief rival Forrester Research has a better customer base and a more stable business model, with profitability for the last five years.
Financial analysts add that for a variety of reasons Jupiter has had a tough time committing customers to renewing its service, with roughly 50 percent to 60 percent of clients renewed for research compared to 80 percent historically.
Johnson brushed off those concerns, saying that the drop "is not that big...We're all adjusting to the change in the client mix due to the business evolving to the traditional companies and away from the dot-coms."
Meanwhile, Jupiter's conference business has tanked.
According to a recent filing, revenues from events fell some 81 percent in the first quarter, from US$5.2 million during the first three months of 2000 to US$1 million in 2001. The company has also canceled 12 of 30 scheduled conferences this year.
Meanwhile, attendance at some of its conferences has dropped dramatically. Clark said a company-sponsored business-to-business conference drew about 300 attendees in Boston this year. That performance pales to events thrown near the height of the dot-com bubble a year ago, when the same conference drew about 1,700 people.
Lastly, financial analysts add they are concerned with Jupiter's "burn rate," or the amount of money it spends each quarter. Analysts say at the rate the company spends money, about US$12 million a quarter, it should have about US$40 million at the end of 2001.
These figures "cut it too close in this environment," said one source, who requested anonymity.
In the first quarter of 2000, Jupiter reported losses, excluding charges, of US$10.8 million, or 30 cents a share. Revenues for the quarter were nearly US$30 million, up 8 percent from US$27 million in the year-ago quarter. Financial analysts estimate that the company will reach profitability in late 2002--assuming it plays its cards right.
"The market for (Jupiter) measurement and ratings has shrunk dramatically," said Vanderbilt's Hoffman. "There's a great sucking sound coming from the Internet right now in that sector."