Raising the Post-Click Conversion Bar: BT And Web Site Optimization
Behavioral Insider: How did your approach to relevant content serving evolve? What limitations in current thinking about targeting and in particular the role of behavioral data is it designed to address?
Brian Deegan: Like most good ideas, relevant content serving came from customers. One of them said, 'you know we've spent over $800k on clicks and basically when all is said and done we've got about $400k for what we paid. Can you do anything else?'
Sellers have gotten used to thinking in terms of the constraint -- that the best they can hope for is that 4% conversion. All they allow themselves to hope for is that those sales will ultimately pay for all the clicks they've 'wasted.' For the most part, marketers think that once you've done a click-through campaign and calculated your conversion performance, that's about all you can do. But clearly something is missing there. They're not done. What's needed is a way of extending the opportunity to convert. The question is, what about those other 96%. For the most part, it's like they'd gone through a black hole.
BI: So how does an advertiser get more yield out of those wasted clicks?
Deegan: A key way to do that is to further segment customers after they click through. If someone clicked through with a certain Google keyword, say, and didn't buy, that needn't be the end. Not if you've learned from subsequent search query behavior that they're interested in particular kinds of luxury travel. The next time they visit, you're able to offer them a much richer, more personalized experience.
What we are attempting to do is make it more simple for an e-commerce site to tag the types of consumer segments they are targeting to particular kinds of messaging content they want to serve to that customer. Relevant content serving -- or RCS, as we call it -- is designed to allow Web sites to serve relevant content to specific consumers based on their past online behavior, not just the keywords they searched on before clicking but how they behaved afterward between visits. Let's say a consumer visits Google and searches for a new MP3. After they search, they click through and go to a landing page but they're one of the 96% that don't buy. Upon repeat site visits, they automatically will receive personalized content based on their specific tastes and interests.
BI: How is RCS deployed by a publisher?
Deegan: What we do is provide code snippets for marketers to create what we call 'Live Zones' on their Web site. Any content served up on the Live Zone can be associated with these content tags. We use data from any click-through history available from the site, whether it's related to search or email or affiliate marketing campaigns. The philosophy is that once a prospective customer clicks through and then leaves, you don't go back to square one with them but make their next visit one you're more ready for, because you've taken what they demonstrably are interested in and actually personalized your Web site to reflect those interests.
BI: Can you compare what you're doing to re-targeting?
Deegan: Re-targeting as advertisers or ad networks -- which we are not -- do it, is a related concept, certainly. But what we're focused on is less trying to repeat the same to get another click-through than how to actually make every return visit more optimized for conversion, based on a progressive ability to target personalized content on the site itself. So many advertisers spend all their resources trying to maximize click-through. They amass incredible analytic data about how to make advertisers build site traffic, but then they completely lose their focus once it comes to what people actually do on their site. So what we're doing is extending that concept.
BI: How do you expect to scale RCS?
Deegan: Where we can see this evolving is toward partnerships where we track consumers who've clicked through to multiple Web sites within an e-commerce category. So that you not only can personalize content based on what customers have searched for and what they've done when they clicked on your site, but if, say, a group of e-tailers banded together to share certain data points, what they've done at other shopping sites as well.
BI: I understand the platform is just being rolled out. What are your goals ? Looking ahead, how would you define success in the short to mid-term?
Deegan: The premise is to break the ceiling that currently exists regarding what an acceptable or good conversion rate is industry wide. We hope to say after six months or so of this program that if a customer is currently seeing a conversion rate of about 4.2% that by focusing on repeat visits by previous click-through customers they can improve that rate to 5% or 5.2%. That's the short-term goal here.