Building Profiles With Cookies

The true currency of the Internet is relationships. The companies that are best at managing online relationships will eventually emerge the winners. Managing relationships, however, is no simple feat. That’s why I find it hard to swallow the fact that many online marketers are giving away a fundamental tool used to manage online relationships.

It is critical that online marketers have the ability to recognize individuals who have interacted with their online presence. Obviously, the human brain is incapable of tracking the millions of relationships that can result from a good online presence, so technology must step in and fulfill that need. Currently, cookies are the best tool for recognizing web users after users have interacted with your brand in some capacity.

Whenever a request is made to a web server, that server has the opportunity to set a cookie on the machine that made the request. (Of course, there are exceptions. Web users can refuse cookies, but I defy anyone to make it through an entire e-commerce site without accepting cookies. You’ll drive yourself nuts.) Thus, it follows that user recognition of a cookie is an easy and effective way for you to recognize someone who has been to your web site.

But cookies can be just as easily leveraged to understand a prospect’s level of interaction with an advertising campaign. Almost all third-party servers set a cookie on users whenever a banner ad request is made. This cookie can be used for banner sequencing (showing ads in a certain order), for tracking beyond the click, and for database profiling. The fact that this cookie belongs to the third-party server and not to the advertiser makes it difficult for that advertiser to tie advertising-related data to site-based behavioral data.

As a general rule, a cookie can be read and understood only by the server that sets it. If an advertiser is using a third-party ad server, that cookie belongs to that ad server. This makes it tough for the advertiser to build user profiles that contain any information about a user’s interaction with an ad campaign. If the advertiser were to set his or her own cookie, it would be possible to create database profiles that integrate information about a user’s behavior on a site and his or her interaction with the corresponding ad campaign.

Such a profiling system could provide valuable insight into consumer behavior that could steer future marketing efforts. Through collaborative filtering techniques, a database of behavioral information could reveal that users who responded to an ad for a specific product also had a tendency to purchase another product. This might lead an online advertiser to comarket those two products in the future, thus saving money and increasing effectiveness. From a “big picture” perspective, an end-to-end profiling system can provide numerous insights into how users become aware of your web presence and how they interact with it.

Before the “‘Big Brother’ Squad” trashes this idea completely, I’d like to state for the record that I think an end-to-end profiling system could be built that would provide valuable marketing insight without linking consumer behavior to any information that would personally identify a web site user. Any insight gleaned from such a system would be valuable only if data were examined on an aggregate basis.

If you advertise online and are trying to understand your consumer better, think about how tying together your ad campaign data and your site usage data can help you to achieve that goal.