Believe it or not, we’re still in the first phase of Behavioral Targeting in online advertising. While there were some early attempts at targeting in 1999 and 2000 by Engage and DoubleClick, Behavioral Targeting as we know it didn’t really take off until 2002 and 2003, when vertical publishers like USAToday.com and Tribune Interactive and the portal Yahoo! deployed it and started creating some real success stories for advertisers. Since that time, we’ve seen hundreds of web sites deploy Behavioral Targeting capabilities; we’ve seen ad networks like Advertising.com start to use it; and we’ve seen thousands of campaigns run with Behavioral Targeting elements.
What have we learned in the process? A lot.
We’ve learned that Behavioral Targeting has a lot of promise. But, while the first phase of its use has demonstrated its extraordinary potential, it has also exposed many of its limitations. On the “potential” side, almost everyone that has used Behavioral Targeting has found the same thing: Used correctly, it can deliver more relevant advertising to consumers with less waste and more impact than any other form of advertising. It can help advertisers better “engage” consumers. It can help publishers better monetize their web pages, particularly those pages without a clear commercial context — the majority of web pages and searches for most publishers. It can help reduce the ad clutter on web pages, a real problem for our industry.
But for all its strengths, Behavioral Targeting services today also have a number of easily identifiable weaknesses: lack of scale, lack of consistency and lack of standards; complexity and consumer privacy concerns. If Behavioral Targeting is going to help drive the next wave of online advertising, this will have to change.
In my opinion, the second phase of Behavioral Targeting, which is just coming onto the scene with new behaviorally-targeted ad networks and second generation targeting services, will feature some rather important characteristics that counter the weaknesses of its current iteration. Its future will depend on a strong industry commitment to satisfying each of the following criteria: