THE TERM “TOSSING YOUR COOKIES” HAS NEVER HAD AN exactly positive connotation, but for marketers it's becoming even more of an unsavory concept.
Recent studies suggest consumers are deleting cookies from their Web browsers quicker than ever, diminishing their usefulness as measurement tools.
A user who deletes a cookie from a Web site and then revisits that site will get another cookie and will be counted as a unique visitor — even though he's returning. And if someone clicks through a search ad to a landing page but doesn't purchase until a later online session, removing the cookies might mean the search ad won't get credit for promoting that later sale.
Reactions to the perceived threat have been varied. Some have stepped up efforts to work with trade groups on definitions that would distinguish cookies from spyware or adware, while others are trying to educate consumers.
Still, those are long-term solutions. Some search engine optimization and Web analytics firms are finding ways to remedy the deletion problem today.
BROWSER COOKIES ARE no longer the tracking tool they once were.
A study published last spring suggests consumers are deleting cookies more quickly than before, throwing into question their reliability as measurement tools. JupiterResearch's report found 39% of all respondents eliminated them once a month; a similar poll from Nielsen/NetRatings said 43.7% routinely delete them monthly. And a June survey by online ad network Burst! Media determined that 38.4% of Internet users remove cookies at least monthly. That figure goes up to some 42.1% among 21- to 54-year-olds.
These deletions cause tracking problems for online advertisers and marketers — and no wonder. Cookies are short pieces of code inserted into a Web visitor's browser. They allow Web advertisers and their vendors to track behavior on Web sites and monitor which search ads are converting well.
A user who deletes a cookie received on a Web site and then revisits that site will get another cookie and will be counted as a unique visitor even though he's returning. If someone clicks through a search ad to a landing page but doesn't purchase until a later online session, removing the cookies might mean the search ad won't get credit for promoting that later sale.
Reactions to the perceived threat have been various. Some forces in the online advertising industry have stepped up efforts to work with anti-spyware trade groups on definitions that would clearly distinguish cookies from either malicious spyware or adware (downloaded software that serves pop-up ads). Other groups, like the Interactive Advertising Bureau and SafeCount.org, say they're preparing plans to educate consumers on the matter. The hope here is that better understanding will prevent or at least slow deletion and preserve cookies' value as a measurement tool.
But those are long-term efforts. In the meantime, advertisers need to know now how well their ads and Web sites are working. So some search engine optimization and Web analytics firms are finding ways to remedy the deletion problem.
The most common solution is to switch from using third-party cookies — those placed on a user's browser by an outside firm such as a Web analytics company — to first-party cookies, which are downloaded directly by the Web site operator. Many users set their browser to accept only the latter. And consumers who run anti-spyware programs are much less likely to delete cookies downloaded by Web brands they recognize than those served up by outside firms.
“We began to see a rise in third-party cookie blocking rates about 18 months ago,” says Jeff Seacrist, director of product marketing for Web analytics firm WebTrends. “By the end of last year, we estimated that 17% of users were blocking third-party cookies.” While that proportion had settled at about 13% by July, it was still much higher than the less than 4% of users who blocked first-party cookies served by Web site operators.
Many site owners already serve their own cookies to let users personalize content, store shopping cart information and get product recommendations. WebTrends developed a technology that uses those first-party cookies to accomplish the tracking and monitoring functions that companies need to keep their sites efficient and optimized. Approximately 500 of WebTrends' 2,800 subscribers have made the transition from third- to first-party cookies since the company began encouraging them to switch this past May, Seacrist says.
Among the difficulties involved in swapping cookies is bridging to the historical data obtained with the third-party version. WebTrends was able to engineer a proprietary solution that eased this data-retention difficulty for its customers, according to Seacrist. “We started life as a software company before introducing a hosted service about five years ago,” he says. “So a lot of our customers had WebTrends software installed in their own data centers. The data they were using to analyze Web behavior was coming from their own environments, so it was first-party by nature.” That installed base in customer data centers is an advantage WebTrends enjoys over some of its Web analysis competitors, which only offer the hosted-services model, Seacrist adds.
Other Web metrics firms have also begun using first-party cookies to produce more accurate readings for clients. In July, Coremetrics published research from its LiveMark Index of 150 customers showing that 13.8% of their traffic registered as anonymous in May — meaning users had either deleted or blocked third-party cookies. The blockage rate was highest in the online retail sector, where it reached 16.4%. By contrast, Web sites using first-party cookies experienced an average anonymous visit rate of 0.6%. As a result, Coremetrics instituted a policy of encouraging migration to a first-party cookie tracking system.
Web site testing firms are also feeling the impact of cookie deletion. Many testing firms have relied on JavaScript cookies embedded in Web pages to send them a signal when a visitor is downloading a Web page; those signals prompted the firms to send out a test version of the dynamic HTML content. Users who block placement of those third-party cookies won't get the revamped content meant to be tested.
Some testing firms have made the switch to first-party cookies to get around this obstacle and provide more reliable testing results. Boston-based SiteSpect has developed a platform for doing A/B and multivariant Web site testing using only first-party cookies without JavaScript. SiteSpect began devising its platform to get around other limitations of using JavaScript for testing. “But a lot has happened in the industry in terms of people being increasingly concerned about cookies, security and third-party JavaScript,” says Eric Hansen, SpiteSpect's president. “Those trends have worked to differentiate our platform.”
Other Web testing and optimization firms have made the switch to first-party cookies. But Matt Roche, founder and CEO of Offermatica, which deploys first-party cookies, says the furor over cookie deletion has largely been the product of self-serving tactics by pop-up ad servers. These behavioral ad networks are making a false equation between cookies used to personalize Web sites or monitor Web traffic and those used to trigger pop-up ads based on the pages a consumer is seeing.
“There's a legitimate consumer concern about malicious software, and that explains the market for the spy blockers and pop-up ad blockers,” Roche says. “Most of the messaging associating cookies with spyware is coming out of that [pop-up ad] sphere. They're trying to associate cookies with spyware and saying, ‘Don't block us, because adware is benign and we perform a valuable consumer function.’ These companies have a vested interest in associating with cookies because they're so valuable.”
One ad server has taken the unique approach of allowing visitors an opt-out choice within the actual ad as they view it. New York-based AlmondNet, which debuted its service last January, offers a “post-search” ad network which gathers data on what users have been looking for on the search engines. It then builds recent search profiles for those users with third-party cookies. When those users visit Web pages that are part of AlmondNet's network, the company delivers ads that relate not to the content of the page they're looking at but to the area of their most recent searches.
Users who view the ads can click on a “Powered by AlmondNet” link at the bottom and go directly to an opt-out page that will stop the network from serving them any more ads. “We tell them what we do — that we show ads based on recent searches, but we have no idea who they are and don't collect personally identifiable information — and they can click here if they don't value the service,” says Roy Shkedi, AlmondNet's founder and CEO.
“But we also tell them, ‘By the way, don't erase our cookies, or we'll have no way of knowing who you are and might then serve you another ad next time.’”