Behavioral Insider: In your report, you distinguish between two definitions of BT. What are they?
Riley: 'Extended content targeting' specifically means an advertiser will be targeting someone who has looked at a particular type of content as defined by both the advertisers and the publisher. And then they target an ad to that person when [he's] on a different type of content.
The second type, 'purchase intent targeting,' is slightly more truly behaviorial--and this arises when an advertiser has a goal to find somebody within a purchase cycle. They are looking for someone who is actually researching a product or talking about buying. It is a way to take advantage of a much wider reach, because sites like automotive sell out quickly. Advertisers in the space understand the value of the site itself and are looking for extended reach, since they know there is such a short window of time when people will be researching.
BI: Is there some confusion among advertisers about these definitions and how each works?
Riley: I think there is. And there is 'retargeting.' All of these things being called behavioral targeting can be confusing. If you look at our report, not a ton of people have adopted behavioral targeting yet, and I think one reason is they really aren't sure how it fits into their business model. With advertisers who can define a purchase cycle and find people researching a purchase, then behavioral targeting is an obvious tool to test. With brand advertisers, it becomes a little hazier.
BI: Does industry education remain a problem?
Riley: It [varies] by vertical. In travel and automotive and even the finance vertical, behavioral targeting has caught on. With a large swathe of brand advertisers and several direct response type advertisers it's less familiar. Two different groups are trying to educate advertisers. The first are Web properties, and Yahoo is a perfect example. Their Yahoo Fusion is essentially extended content targeting, but they train their salespeople to bring this up as an example of extended reach for somebody who feels they have gotten to the outer limits of their buy with Yahoo. The ones with proprietary technology or within a large network... who have seen success with it, are able to bring it up in the sales process. To a large degree, though, it really is coming from agencies. They are pretty well educated when it comes to direct response and how it leverages behavioral targeting--and they are starting to pick it up on the brand side, and that is where the larger advertisers will start dipping their toes.
BI: Your surveys showed how advertisers that had tried BT had different view of the challenges to using the technique. They tended to see things like finding the right sites, negotiating with sites, and measuring ROI across media more challenging. Does that raise any red or yellow flags for the BT industry?
Riley: It's hard to understand how to determine your target. There is no standard yet, especially with individual advertisers that aren't in one of these easy verticals like automotive where you can find people researching a product. The reason is that just because somebody reads two or three articles on the subject doesn't necessarily mean they are better audience for you. You need to figure out, if that's not the right definition, then what is? As Yahoo and [others] collect data over large swatches of advertisers, they are going to be more effective about consulting with the advertiser, but it's still very early in the process.
BI: Is BT going to be appropriate for shorter-term-consideration purchases?
Riley: Even short-purchase-cycle advertiser could benefit from BT. The issue is, can you identify the behavior quickly enough to show them the right ads. I don't think that people have gotten used to coming up with the right definition of what a useful behavior is. The technology needs to advance or the network get large enough or tapped into e-commerce sites effectively enough to figure out that people are researching a pretty quick buy. Their window is very small. For example, I think eBay partnering with Yahoo would give Yahoo potentially a great window into something like a short purchase cycle with millions of purchases going on each month.
BI: Do you see a place for BT in branding?
Riley: It's still up to the advertiser to figure out what the right behavior is to target. I am always a proponent of experimenting and it takes a long time to figure out the right way to do any type of advertising. But in many cases it's very hard to figure out what is a behavior online that indicates that this person is more likely to listen to my message.
BI: What are the key weaknesses in the marketplace of BT offerings from both the vendors and the publishers?
Riley: There are two different issues right now. With the more popular verticals like automotive, it is really selling out the extended buys, so someone like ValueClick has recently partnered with Jumpstart in an attempt to increase their available inventory for auto advertisers. There still is a limited reach, especially in industries where not everyone is going to be buying your product all the times.
On the content side, a lot of it is just figuring out what content is worthwhile to target and what's not. Just because someone reads an article about golfing, is that an indicator of what they then want to buy? And how do you categorize different types of content across different Web sites? Do the people who read golf articles on ESPN.com have the same interest as someone who reads the golf article by chance on the Wall Street Journal [site]? That comes into play a lot. And then there is overlaying BT with other types of targeting. Your audiences shrink dramatically. If you have a demographic or geo-target you want to stick to, suddenly you are left with 50 impressions a day.