In late May, more than five million Web users vanished
In late May, more than five million Web users vanished.
The disappearing act came when Nielsen/NetRatings, a leading company in measuring Internet traffic, sharply cut its previously reported statistics for the financial Web site Entrepreneur.com to 2 million unique visitors in April, from 7.6 million.
Why the change? For millions of Web surfers, Entrepreneur.com visited them — and not the other way around, the measurement company said.
As computer users visited other sites, new browser windows popped up containing articles from Entrepreneur.com, according to Scott Ross, senior product manager for Nielsen/NetRatings.
Pop-up windows appear all over the Internet, including the Web site of The New York Times. But they are typically used as advertising to pitch a product or a service.
Entrepreneur.com’s pop-ups were unusual because they contained news content, like articles on how to start a small business, making them hard to distinguish from an intentional visit to Entrepreneur.com’s site. This hailstorm of pop-ups more than tripled Entrepreneur’s reported traffic before it was detected and factored out a month later.
The technique of using pop-ups to gain readers underscores just how important sheer numbers have become in the online media business. Advertisers are shifting their marketing dollars to the Internet, but the rates they pay are low compared with traditional media.
Consequently, publishers who have struggled for years to find a way to make money online are taking aggressive steps to get their Web pages in front of as many eyes as possible.
Entrepreneur.com, owned by Entrepreneur Media in Irvine, Calif., did not return calls seeking comment. But it is not the only online publisher to use pop-ups, according to Benjamin G. Edelman, a Harvard doctoral student who has compiled a large database by installing on his computer many kinds of software, known as adware, that generates pop-ups.
(Mr. Edelman has also provided expert testimony on behalf of publishers, including The New York Times Company and other newspaper concerns, in a lawsuit involving adware. The publishers had sued to prevent pop-ups, by the Gator Corporation, from appearing on their Web sites. The suit was settled in 2003 under terms that were not disclosed.)
Other sites that appear to have used pop-ups for content in the last year include Concierge.com, the Web site of Condé Nast Traveler magazine; ForbesAutos.com, part of the Forbes financial publishing group; and Heavy.com, a popular humor site, Mr. Edelman said.
The concern over pop-up content goes beyond traffic numbers. Many advertisers pay premium prices to reach readers of certain Web sites. Through pop-ups, these advertisers may find their orders are being fulfilled with low-cost page views that users never requested and may never have seen.
This list of sites surprised Scott Symonds, vice president for media at Agency.com, who advises companies on where to spend their online advertising budgets. Pop-ups delivered by adware are usually seen as a “nuisance form of advertising,” and most mainstream publishers avoid them, he said.
“You would hope that publishers of high-quality content would use advertising techniques that were in keeping with that,” Mr. Symonds said. He added, though, that pop-ups could be a legitimate way to reach new viewers if the publisher took certain precautions, like not using pop-ups to inflate traffic or satisfy orders from advertisers.
There are legal issues as well. Many sellers of pop-up ads have been sued by regulators and consumers, who say the software to allow pop-ups is often installed without a users’ consent. The adware hitches a ride on another application, like a game or a screen saver, and the pop-up function can be buried in the fine print. Sometimes it is never disclosed.
“You can almost look at it like steroids,” Mr. Ross said of pop-up content, which his firm calls “non-user-requested” traffic. Others in the online media business call it “push traffic,” because Web pages are actively pushed to computer users who do not request them.
Used indiscriminately, push traffic is like printing extra copies of a magazine and tossing them onto doorsteps or, because pop-ups can be intentionally hidden behind other windows, simply dropping them in an alley.
Mr. Ross and executives at comScore, a rival measurement company, say they can usually detect such activity and remove it from their data. But both companies concede that they cannot catch everything.
“It is a cat-and-mouse game,” said Magid M. Abraham, comScore’s chief executive.
In the case of Concierge.com, which has articles on luxury resorts and expensive spas, the site recently bought pop-up services from Zango, one of the largest adware companies. In one case, Zango’s software caused a page featuring “Hot List Hotels 2006” and other travel articles (created by Concierge.com) to appear unexpectedly while Mr. Edelman was browsing other sites, he said.
Concierge.com declined to comment, but Zango has faced harsh criticism from the Federal Trade Commission. In early November, Zango agreed to pay a $3 million fine to settle the commission’s charges that it had used unfair and deceptive practices to install its software on personal computers and to make it difficult to remove.
“If consumers choose to receive pop-up ads, so be it,” Lydia B. Parnes, director of the commission’s Bureau of Consumer Protection, said in announcing the Nov. 3 settlement. “But it violates federal law to secretly install software that forces consumers to get pop-ups that disrupt their computer use.”
Zango, based in Bellevue, Wash., said that third-party affiliates were the source of the problems and that it had long since cut ties with them. Zango also said it had operated within the requirements of the F.T.C. settlement, including verifying computer users’ consent, since Jan. 1 and had hired an outside auditor to confirm its compliance.
Zango’s chief executive, Keith Smith, said he would not identify or discuss any of his company’s clients. Asked about Zango’s potential as a tool to inflate traffic numbers, he said that publishers and measurement companies need to work out the issue themselves. “The measurement piece is still evolving,” he said.
Zango is just one of the adware providers that work with online publishers. Until late last year, ForbesAutos.com, an auto-related offshoot of the financial site Forbes.com, was using the services of eXact Advertising, whose adware is sometimes bundled with free games and other applications.
Screen images from December 2005 show several cases in which eXact delivered unsolicited pages — in one instance, a review of a BMW Z4 coupe — from ForbesAutos.com. The pages were actually “pop-unders,” positioned so they were mostly obscured by the main browser window, to be revealed when that window was closed.
Forbes.com declined to say how long it used pop-ups or how many pages were generated that way, but Jim Spanfeller, Forbes.com’s chief executive, said they accounted for a “very small fraction” of its page views. He also said the site abandoned the practice last year.
“We decided in 2005 to stop using pop-ups of any sort, delivered by adware or otherwise, for site promotion after determining they were of less utility than other efforts,” Mr. Spanfeller said. A spokeswoman for eXact Advertising declined to comment.
Heavy.com, a site that tries to attract young men with its irreverent video clips and animation, is also a page popper, though the company says it has taken steps to avoid inflating traffic statistics or upsetting advertisers. Citing data from Hitwise, a traffic measurement company, Heavy.com said that in October it was the second-largest entertainment video site after YouTube.com. Rather than rely on videos alone, though, the site has also been using pop-ups to position its pages in front of users.
In a recent session, Mr. Edelman said he saw two Heavy.com pages appear on his screen within the space of a minute, each generated from a separate piece of adware. In one case, the Heavy.com home page appeared while he was browsing Netflix, the video rental service.
Some users seem to be bombarded by Heavy.com pop-ups. “HELP! tons of pop-ups from heavy.com and webcrawl.com!!” was the plea posted a few weeks ago in the forums of SpywareInfo.com, which offers advice on shedding unwanted software.
Heavy.com acknowledges using pop-ups, but Andy Morris, a spokesman for the company, said it did not use adware or condone its use. Instead, Heavy.com works with ad networks whose member sites initiate the pop-ups, a practice it calls an “effective marketing tool.”
Presented with Mr. Edelman’s pop-up examples, Mr. Morris said they were not generated at Heavy.com’s request. In some cases, he said, the site has been used by “unscrupulous third-party Web operators” that tried to use its videos as bait; in others, Heavy.com worked with ad networks that then violated the terms of their agreement by using adware.
“Quite simply, we’ve been ripped off,” Mr. Morris said.
Heavy.com said its advertisers were never charged for pop-up pages and that the pop-ups did not inflate its traffic because it flagged those pages so that comScore could exclude them from its statistics. ComScore confirmed this arrangement.
“They did the honorable thing,” Mr. Abraham said.
But Nielsen/NetRatings was unaware of Heavy.com’s pop-up campaign until recently, a NetRatings spokeswoman said. When it began excluding those pop-ups in October, Heavy.com’s traffic dropped 35 percent from the previous month, to 1.8 million, although NetRatings said it was unclear how much of the decline was related to removing the pop-ups. (ComScore reported 7.8 million visitors to Heavy.com in October, more than four times the NetRatings number and a large gap even in the inexact world of Web measurement.)
Heavy.com strongly disputes the accuracy of the NetRatings data. Among other things, it argues that the company does a poor job of tracking sites like Heavy.com that use Flash multimedia software throughout their pages. It also said that NetRatings had relatively few college students, who are a large part of Heavy.com’s audience, in its survey group.