Digithoughts

The world of behavioral analytics does not have clearly drawn lines about what it is (and isn’t).  Where I work, we normally think of behavioral analytics as the way we recognize patterns of behavior based on site activity. Once we understand what visitors are doing on a website, we can create visitor segments based on those patterns to discover what is driving those behaviors.

For example, we can use onsite behavior to uncover which site assets are being downloaded during the first visit of a visitor who would later register on the site.  By understanding what content drives future registration, we can help our clients prioritize which assets to create (assuming registration is their goal).

Last week, I was invited to speak on a panel at OMMA Behavioral in San Francisco (recorded video here), regarding how behavioral analytics allows brands to connect with their target audience online.  Unlike the post-click analysis we typically do, OMMA Behavioral is primarily relevant for brand advertisers who use analytics to purchase media targeted at behavioral segments. Instead of post-click engagement and conversion, behavioral advertisers measure success by brand lift and purchase intent.

In a way, Digitaria’s first-party data analysis made us an outlier within the panel; however, it became an excellent chance to learn from the other panelists and share some ideas about why brand marketers are so often disconnected from site-side performance.

The perspectives shared at OMMA behavioral (during both our panel and vendor meetings) gave me a few ideas about analytics trends to watch for. These three trends will help everyone develop a more sophisticated behavioral approach:

1. Better behavioral pattern recognition

Any web analyst worth his salt understands that sometimes pulling insights from web analytics data can be akin to digging holes – lots of them. Insights don’t come easily and part of what makes someone a great web analyst is the intuition and perspective necessary to determine what data to look at and where to start digging. I could see a future where onsite behavior becomes more quickly recognized by pattern recognition services. This would help web analysts become increasingly efficient and more productive with their time. Analysts could spend more time understanding why behavioral patterns occur, instead of trying to find them.

2. Merging 3rd party audience definitions with on-site behavior

While we may be able to understand on-site behaviors through web analytics data, we still have a long way to go before we can understand who those customers are. Certain verticals -- retail, travel, finance -- are accustomed to understanding their customers over multiple touch points, because their customers hand over their data at the time of purchase; but what about consumer packaged goods (CPG), business to business (B2B), and entertainment? In the case of B2B, the purchase cycle can be so long and protracted that by the time your visitor tells you who they are, it may be too late to impact the buying process. Although CPG, B2B, and entertainment have a deep understanding of who their customers are, they have difficulty identifying them online.

For industries that focus on branding, the future may be in connecting on-site activities and sources of traffic to personas identified by 3rd parties on the web. These personas (e.g. Nielsen PRIZM) combined with behavioral analysis from 3rd party cookies, could be a powerful tool for learning more about who is coming to the site, what other sites they may have visited, and the factors that may be driving their behavior. In addition, we should see on-site segments powering behavioral ad buys more and more in the future.

3.  Working with brand marketers to validate behavioral ad buys

My last observation is a human problem, possibly the one that will take the longest to resolve. There is a fundamental disconnect between online ad buys that set out to accomplish branding objectives, and the web analysts who measure online success. Too often, these two sides of the house don’t communicate enough or effectively, and they’ve never “walked in the other’s shoes”  -- e.g., few web analysts have a background in brand advertising, while  few media planners have done serious web analysis.

This disconnect is confusing to marketing managers, who often hear quite different stories from each group: the media buyer says the campaign was a success on the basis of frequency and reach; while the web analyst complains about the above average bounce rates, and lower time on-site associated with visitors from that campaign.

The truth is that modern web analysis tools often fail to capture the value of branding campaigns, while brand marketers don’t give enough attention to understanding how on-site activity insights can improve the quality of ad buys. It is essential to initiate the conversation between ad buyers and web analysts before the campaign, in order to reach an agreement of how success will be measured.

My OMMA Behavioral experience was a chance to see and hear from the “other side” (pre-click analysis), and learn more about how ad effectiveness is being measured and talked about. Web analysts should start thinking about how they can cross the divide between brand measurement and on-site activity -- and a great place to start is by attending these conferences.

Matthew Parry Sr. Analyst & Web Analytics Evangelist

Read more from the Analytics category. If you would like to leave a comment, click here: Comment or stay up to date with this post via RSS, or you can Trackback from your site.