Behavioral Targeting
After 10 years of experience with the Internet, advertisers are finally figuring out how to reach audiences based on where they go and what they do online. Until recently, the only way to target a banner ad effectively was by the content on the page visited, or by the keyword being searched. A visitor to a page about football could be served an ad for event tickets. A person searching about the stock market could be given an ad for personal finance products.
But new technology is enabling website publishers to track visitors’ page views over time (anonymously, using a cookie), and serve up advertising based on where the visitor went, and how often, in the past—not just where they have landed today. The concept is called “behavioral targeting,” and is enabled by companies like TACODA Systems, Revenue Science, and Blue Lithium.
Behavioral targeting allows advertisers to reach more targeted prospects, and, one hopes, enjoy a higher response. For consumers, who feel bombarded by ad messages, behavioral targeting presumably increases ad relevancy and interest, and improves their experience online. From the publisher perspective, the benefits include the ability to broaden the salability of their banner ad inventory, not to mention premium ad prices—publishers using TACODA are getting between $4 and $25 CPM on inventory that would otherwise have sold for $2 or less per thousand. Sounds like a win/win/win situation.
Consider the example of NTT DoCoMo’s U.S. division, which was looking to reach IT decision makers, and set up a test on ft.com, the website of Financial Times. NTT DoCoMo’s objective was to increase brand awareness and improve brand image among this influential group. An ad was served up to a control group in a standard run-of-site media buy across the ft.com’s site. The same ad was delivered to a test group identified by Revenue Science’s targeting capabilities. The results? The test group performed better on all counts, including cost. On average, 61% better, with one brand attribute (NTT DoCoMo is a “pioneer in the field of mobile telecom”) achieving an 83% lift in awareness.
Results like these are driving a huge growth in the use of behavioral targeting among U.S. advertisers. According to a June 2006 JupiterResearch study, usage will soon double, moving from 13% of advertisers now, to 25% who said they expect to use this kind of targeting next year.
Is behavioral targeting working for all kinds of marketers? Products involving deep consideration and a distinct, relatively long purchase cycle are most likely to see improvements with behavioral targeting versus standard banner ad media buys. According to TACODA Systems, the top categories include automotive, mortgage loans, insurance, consumer electronics and travel.
In order find a large enough audience, most behavioral targeting service providers have built networks of publishers who make their visitors available to advertisers. TACODA has pulled together over 3,000 publishers, including such well-known brands as The Food Network, NBC and cars.com. Blue Lithium offers over 1,000 sites in its network.
With this kind of reach, the numbers become interesting to general advertisers. TACODA has segmented the visitors on its network into 31 pre-canned affinity groups of ad targets based on a variety of interests, for example:
- 3.5 million avid sports fans
- 5 million entrepreneurs
- 6.8 million gourmet cooks
- 2.6 million late-stage automobile shoppers
- 1.2 million camera bugs
With such growth, it’s no surprise that new techniques and innovations are emerging almost daily. For instance, Blue Lithium offers a service called AdPath that lets advertisers follow prospects as they move along the sales cycle, whether at the advertiser’s site or in the network as a whole. Marketers can thus serve even more relevant messages as the customer relationship evolves. And TACODA recently announced that its advertisers can serve up not only standard banner units, but also video ads, via behavioral targeting. What will come next? Even more effective ways to market, we hope.