• by Omar Tawakol , April 13, 2006

At the recent OMMA Hollywood conference, the buzz about behavioral targeting was bigger and better than ever. Growth in share of online ad dollars and a recent glowing Forrester report had advertisers and media companies excited. However, while confidence was high, there was still significant confusion in one area: what constitutes "a behavior," and how do you match the right behaviors with strategy?

When you're making an online behavioral media buy, do you know which behaviors you're really buying? When you plan a buy, do you know whether the behavior is signaling the consumer's intent to purchase, or an interest in the topic?

This distinction has always been critical in advertising: advertisers display brand ads to interested prospects and direct response ads to people who are in the market to buy. It is no less critical in the online world of behavioral targeting.

Interest Versus Intent

If users visit a news site within a travel section, we can assume that they are interested in travel. But we can't assume that their behavior means that they are about to purchase a trip--not unless they click to a travel reservation site, that is. But when you purchase a behavioral media buy, are you getting that level of distinction? Probably not.

By not being transparent about the precise behaviors being sold, the media seller or network is doing a disservice to you. Your ads will not perform if audiences aren't highly targeted, and an ad without a clear call to action may lose the sale with in-market buyers.

By not being fully transparent about behaviors, we're also doing a disservice to the industry. Behavioral targeting will never completely fulfill its promise for advertisers until we remove complexity and make it crystal-clear which behaviors we are buying and selling.

Bundling the Good, the Bad, and the Ugly

For years online media buyers and sellers have played a little game. The rules are unwritten, but everyone knows them. Sellers bundle different types of inventory to achieve the right price. Buyers agree to buy the bundle in the hope that the good part will be enough to make it worthwhile.

With the industry still in its infancy--as behavioral targeting is--confusion shouldn't reign. Greater transparency would make for much more effective behaviorally targeted ads--and the resulting boon would be good for both buyers and sellers.

Toward More Effective Behavioral Media Buys

To have an effective behavioral media buy, two conditions are needed:

1. Behaviors must be fully transparent. The buyer must know whether the audience has a general level of interest in the product, or intent to purchase.
2. The purchased audience must match the campaign goals of the buyer.

If these two conditions are met, you have all the necessary elements for effective behavioral targeting. Some examples of how this would look in action:

Direct Response Example: A financial services company wants to target in-market buyers with its IRA accounts. The buyer purchases an audience whose behaviors indicate intent to purchase. When John types "open IRA" into his search engine, the company's direct response ad is delivered to John, even when he later visits a blog.

Brand Example: An athletic shoe company running a brand campaign purchases an audience whose behaviors indicate interest, such as reading articles on the running section of a sports site or visiting a running blog and submitting an entry that mentions a passion for running. So when Carla surfs a running site and then leaves to visit a weather site, she sees brand ads for running shoes.

Breaking the Black Box

Similar behaviors are clearly not of the same value to advertisers. The person reading an article about cars is not in the same mindset as the person configuring car options for purchase. The same ad could have a big effect on one reader and none on the other.

But for too long now, media sellers have lumped similar behaviors into the same bucket and tried to use a "black box" to hide their value from buyers. It's time to abandon this pose and shine a spotlight on the distinct behaviors that are being bought and sold.

With more transparency, ad buyers and sellers will have the information they need to accurately value specific behaviors. Advertisers can then make better purchase decisions and realize more effective campaigns. Sellers can get maximum value for their ad inventory.

We don't need a black box to confuse the issue. Today, some companies are trying to make money off that confusion. In the search world, that confusion has been used to lump contextual ads into the same bucket as search ads. Clearly, search is about imminent action and intent, and contextual advertising is about interest in a topic. Both can be valuable--but for very different reasons.

Advertisers have pressured the networks to unbundle the two approaches for quite some time. The only reason that search networks could get away with this bundling in the first place was because they first achieved critical mass in search (intent) and they then added volume through contextual (interest). The key point here is that bundling works once you get to critical mass, but before that it simply gets in the way.

So let's do something that will help us all: admit that all behaviors aren't created equal, and provide more transparency for all.