Privacy and Behavioral Targeting: How Much Data Is Too Much?

I grew up in the direct response industry, where everything can be tracked and measured. Behavioral targeting is an old concept: list segmentation and databases, and predictive modeling had that down decades ago. With the onset of behavioral targeting online, the concept takes on new meaning. Consumers feel that Big Brother is looming as the technology become more sophisticated.

ISPs are among those jumping into the fray, truly testing whether tracking a consumer's behavior will be accepted. They've begun using or testing technologies that track their subscribers' online activities and serve ads based on those behaviors.

A quick definition: behavioral targeting tracks a Web visitor's browser clickstreams, often the last four to six visits, to predict what the user may want in the future and targets ads, content, or products based on those past behaviors. The amount of money spent on online behavioral targeting is growing. EMarketer theorizes the $1 billion it predicts for behavioral targeted ad spending in 2008 represents only 11 percent of the U.S. display, rich media, and video market. With the greater attention paid to overall ad targeting, however, and the rising focus on brand messages online, this market will nearly quadruple by the end of 2011.

If this growth is to continue, we as an industry must get some definitions in place on what's acceptable data collection and what violates a consumer's privacy

When it comes to targeting, I'm a firm believer that more is better than less. It keeps unwanted catalogs out of my mailbox, unwanted e-mail out of my inbox, and pop-up ads that I don't want to see out of my eyesight. It's sort of a green approach to marketing. But consumers have to give up something (data) to get something (being marketed to intelligently). And that's the perceived problem.

The line between an acceptable and intrusive amount of data collection is still evolving. It's a fine line. Soon, consumers will probably demand a say in what and how much is to be known about them. A couple of areas are being evaluated:

• Eventually, there will be opt-in or -out systems that allow ad recipients to indicate whether they want their online information tracked.
  
  
• Instead of using the clickstream, publishers may fall back on modeling their readers on an ongoing basis and in as close to real time as possible. They'll use aggregate data instead of personal data to predict behavior. In the short term, this may be unlikely. But it would probably meet with the highest acceptance.

Privacy advocates have pushed the federal government to adopt more stringent measures, such as requiring online business to enable consumers to opt out of data tracking and establishing a "do not track" list for consumers. Industry representatives have contended that such measures would be difficult to implement, among other issues.

In December, the FTC proposed voluntary guidelines, including a provision calling for an opt-out mechanism on all sites engaged in behavioral targeting. And the Interactive Advertising Bureau (IAB) is said to be gathering information on how companies track consumers and use that methodology to establish standards. If the IAB doesn't take measures that exceed the FTC's proposals, privacy advocates won't be pleased.

So the debate persists. We'll continue to explore how this issue evolves in coming months.

This is my first column with ClickZ, and I'm delighted to be in such illustrious company. Much has been written about behavioral targeting, but I'll explore what's current, offering up additional perspective and opinion, talking to suppliers, marketers, and just plain folks who'll share their own ideas and results. I'd love to hear your thoughts.