• by Gavin O'Malley  @mp_gavin, March 15, 2005

In news that unsettled many in the online advertising world, a new study by Jupiter Research revealed that four out of 10 Internet users delete cookies from their primary computers at least once a month.

The report found that about 12 percent of Internet users delete cookies on a monthly basis, 17 percent do so weekly, and 10 percent purge cookies every day. What's more, more than half--52 percent--said they had rid their computers of cookies at least once in the last year. For the study, announced yesterday, Jupiter Research surveyed 2,337 U.S. online consumers in March.

This study marks the first time Jupiter has examined how Internet users react to the cookies that wind up on their personal computers, said Eric Peterson, Jupiter Research analyst and author of the report. "It was commonly assumed, before this study, that users didn't have the sense or the inclination to fool with cookies," Peterson said, "so advertisers and marketers didn't factor the possibility into their tracking and targeting measurements."

Peterson attributed the findings to consumers' fears that cookies violate privacy. "It's a complex issue, which consumers think is easily resolved with the push of a button," he said.

The news that so many people were deleting cookies created a stir in the online advertising world, where marketers rely heavily on cookies to target consumers and analyze ad campaigns.

"The growing number of consumers who block cookies poses a serious challenge for many Web analytics vendors and the online businesses relying on their data, and not all vendors are equally capable of reducing the accuracy issues this can cause," Greg Drew, general manager of WebTrends, said in a statement.

But some advertisers said they were heartened by the findings that the vast majority of consumers retain their cookies for at least one week. "Seven days is ideal," said Dave Yovanno, general manager of online advertising network ValueClick Media. "If the majority of users were deleting cookies more frequently than that, the industry could be in real trouble."

Dave Morgan, CEO of behavioral targeting firm Tacoda, said he is not alarmed by Jupiter's findings, explaining that Tacoda had little need for data more than a month old. "The fresher the better, I find, and while cookies have been the cheapest and most efficient way to track behavior until now, we're developing other techniques to capture behavior without cookies."

Peterson did not disagree with Morgan's reasoning: "Advertisers using lifetime value metrics have to reexamine how accurate their findings are," Petersen said, "but advertisers measuring sites where users come back again and again have less to worry about."

Peterson predicted that Web analytics vendors will increasingly adapt their tools and measurement practices to more accurately track the basics: return visitors, unique visitors, multi-session campaign conversions, lifetime value, and users' broader product and content preferences.

Jupiter's report makes it clear that site operators should distance themselves from third-party cookies, and toward first-party cookies, citing the 28 percent of Internet users who are selectively rejecting third-party cookies. Yet, "with 15 percent of consumers reporting active blocking of all cookies," the report cautioned that the "exclusive use" of first-party cookies is no cure-all.

"It's expensive, but vendors have no choice but to switch from third-to first-party cookies," echoed Michael Leo, CEO of ad serving provider Trafficmac, who raised doubts about the accuracy of Jupiter's report. "To begin with," said Leo, "I have three computers, so when I go to a site's home page at work, it probably thinks that I've deleted the cookie it gave my home computer."

Peterson's report recommended that site owners consider a registration/log-in model, which would allow publishers to reset deleted cookies. For high-traffic sites, Peterson suggested that owners consider using Macromedia Flash's local shared objects--an opt-in system less likely to be spotted and removed by anti-spyware programs.

"You could say it's just cookies by another name, and anti-spyware services could end up targeting local shared objects the same way they target first- and third-party cookies," Peterson said. "This unproductive cycle will continue, I'm sure, until consumers learn how benign cookies can be."

In fact, industry observers said that spyware-removal programs will likely start deleting programs that are used as cookie substitutes. "You're just going to end up escalating this again," said Ari Schwartz, associate director of the Center for Democracy and Technology in Washington, D.C. He added that companies might come up with new ways of collecting information, but other companies will then find a way to stop such practices.

Spyware consultant Ben Edelman added that programs that bypass cookie deletion "might well cause consumer backlash," and might also prompt anti-spyware vendors to start deleting those programs.