Behavioral Insider: Could you give a little background on how Active Athlete approaches behavioral targeting as a niche vertical network?
Robert Tas: When Active Athlete started, our goal was to be the premier venue for aggregating sports sites in a more customized way. What we wanted to do was fill a gap and allow large brands like Nike, Ford, Chevy and others to reach passionate sports enthusiasts in a maximum high-engagement setting, and to do that with scale.
At the beginning we started with a couple of core sports verticals like endurance (running, cycling and triathlon) -- areas I personally was passionate about -- and used those as building blocks to evolve a network, which today encompasses a wide array of sports, both participatory and also including passionate fans.
We specifically decided the network was not going to try to be everything to everybody but instead would focus on finding sports sites with the most passionate kinds of attachment. By focusing on real enthusiast sites we're able to develop a deeper kind of domain knowledge and intelligence about sports-related behavior than would be possible in a more generic behavioral network. This more-focused effort benefits the brand since we can bring them even closer to the consumer. And I'd like to qualify 'niche' as 'premium,' not necessarily 'small.' We have 18 million monthly uniques.
BI: What do you mean by domain knowledge, and how is it applied?
Tas: For instance, PowerBar was looking to reach newbie athletes who were new to endurance sports. The challenge, of course, is how do you define exactly who the new athlete is? It's a much more granular segment than 'sports enthusiast' or even 'active sports participant,' and the behavioral cues or triggers that are going to be most relevant to the brand aren't as susceptible to an automated formula.
What we did was actually learn in some detail from the brand what they're trying to do. In this case it was to present PowerBar as an integral component of a comprehensive training plan and regimen to new competitors. PowerBar sought to present itself as an educational resource, providing information about diet and nutrition as it relates to performance. Knowing this, we were able to customize our approach to behavioral segmentation. The project became identifying what sorts of behavioral markers differentiate athletes new to training. Once you frame it that way, you begin to look for people who are signing up for their first 5K race, people who are looking for background information about the rules and regulations of particular events, or for information about how to start a training regimen or how to purchase the basic gear, how new they are to the community, and athlete goals -- information that specifically identifies an enthusiast as being new to training. To do that sort of targeting you need a depth of vertical-specific domain knowledge.
BI: Do you see what you're doing as supplanting more automated approaches?
Tas: There's a business to applying generic models of behavioral segmentation in cases where that kind of more general target market is all the brand is looking for. But if you're an auto company and have a campaign in mind that addresses a segment you're thinking of as 'adrenaline junkies,' a conventional behavioral network is at a loss. Because you have to figure out first what exactly the brand wants. Is that 17- to 25-year-olds who've demonstrated an avid interest in extreme sports competition? Or are there different or better behavioral cues to be focused on?
This sort of thing is hard to automate. What you need to be able to do is take all the data you've aggregated and all the knowledge you have as a vertical network and translate it into success. To take the PowerBar example again, they could have run a campaign across a network based on a behavioral segment made up of all sports-site visitors. But in segmenting by 'new athletes,' they were able to fit their message of nutrition and training uniquely to suit the specific needs and interests of a target market.
It's not for everybody. We can't deliver a mass group of in-market car buyers like Edmonds. But we can offer an auto brand, for instance, the ability to differentiate their message to highly focused segments. From talking with auto brands, this is on-target with their current marketing objectives.
BI: What are some other things vertical-specific behavioral segmentation can do that transcend more conventional approaches?
Tas: Having highly developed vertical domain knowledge allows you also to target markets in terms of idiosyncrasies in their behavior that would be counterintuitive to conventional behavioral marketing. For instance, we focus heavily on seasonality data. We know that most athletes follow up to seven sports, and they do so 12 months a year. Even though the conventional wisdom would say you do a running-related campaign in the summer and fall and do skiing in the winter, we know that runners still look heavily to running sites and avidly consume information about running in January, even at the peak of skiing season. And people interested in skiing are just as likely to visit a ski site in August.
BI: What are your current innovation focuses?
Tas: The goal is to constantly improve both the data and the understanding of how sports enthusiasts engage content. We're pushing very hard this year to use content management and social networking tools to better understand user-generated data, consumer engagement behavior and metrics. The further we get, the farther we can take our brands.