• by Scott Howe , April 1, 2005

Is the online apocalypse upon us? Recently, Jupiter published survey results suggesting many online surfers purportedly delete their cookies on a regular basis. Since cookies have emerged as the standard of online measurement and the platform for many targeting variables, it stands to reason that an increase in cookie death would undermine the efficacy and interest in online segmentation and targeting. Is it time to panic? Hardly.

Unless users are deleting their cookies several times per day, online measurement should not be impacted. According to Young-Bean Song, director of analytics for Atlas and the Atlas Institute, "The vast majority of conversions - typically ranging from 70 to 90 percent - occur within a 24-hour window of the corresponding click or impression." As result, conversion data will not be significantly impacted.

Marketers employing online targeting schemes should have even less concern, as recent activity is often overlooked as an important influencer of segmentation models. A Web surfer who has visited a travel site within the past few days but failed to make a purchase is an ideal candidate for a cookie-based re-messaging scheme. In contrast, a surfer who visited the same travel site two years ago is probably a very poor candidate for a re-messaging scheme. Smart marketers have long realized that recent data is better data, and employ short windows for behavioral-based targeting schemes. The only impact of increased cookie death will be to limit the volume of impressions served to stale segments that wouldn't have converted anyway.

Even if cookie death continues to increase, marketers could still effectively employ all of the most valuable targeting schemes. Simply by capping daily frequency across all sites on the media plan, advertisers could suppress wasted advertising to the "impression eaters" -- the small portion of cookies that consume a disproportionate amount of daily impressions.

Likewise, advertisers could continue to employ geo-targeting (e.g. a regional airline which wishes to market only to users living in select city pairs), time of day/day of week schemes (e.g. a charity which generates the majority of its conversions on weekend evenings), connection speed targeting (e.g. the broadband provider that wants to convert dial-up users), domain name targeting (e.g. the concert promoter that seeks to target college campuses) and contextual targeting (e.g. the lender who wants to target users who are currently reading an article about mortgage rates). Many campaigns will employ combinations of these assorted variables - opening up a nearly infinite number of potential segmentation schemes.

If history is any guide, there will likely be many in the industry who point to cookie death as a harbinger of catastrophic change. Many of them will do this in order to get their name in print or sell you something you don't really need (anyone remember the slate of Y2K consulting projects?). Ignore the hysteria.

In short, cookie death will simply force marketers to acknowledge data "freshness" and design targeting schemes that fully utilize the vast array of targeting data that already exists. Smart marketers won't grieve cookie death - rather they will incorporate it into their campaign design.