Behavioral Targeting and A/B Testing

Imagine delivering a truly customized, dynamic consumer experience — generated in real time for each unique visitor to your site.

For many marketers, this is the promise of behavioral targeting — the possibility to deliver unique content and offers based on a customer’s profile and behavior. Now, imagine the knowledge needed to make informed “projections” about what your visitors want to see when they get to your site.

The potential payoff from behavioral targeting generates high enthusiasm among executives — and triggers a corresponding rush to action, without consideration as to the value and impact such a program can deliver. Effective behavioral targeting requires both extensive knowledge about your consumers — and a rich array of content to cater to their unique desires. Without a solid foundation, content may be off-target — or worse, off-putting.

Another option for marketers looking to optimize their messaging and content is to leverage A/B testing. A/B testing enables organizations to test multiple messages and/or creative directions across a broader audience. A/B can be a precursor to behavioral targeting — providing a foundation of knowledge to help build personas — and it can be leveraged to test content for companies already running a behavioral program. Assessing the complexity and maturing of your marketing model will help identify which solution is most effective for your campaign.

Do You Know Enough About Your Visitors?

How much actionable information do you have about your visitors? Can you predict their goals and needs? Are you working with ad networks, collecting data across multiple external sites and clickstreams? If so, the potential to deliver targeted content and offers based on their previous behavior is great. You can construct detailed matrixes that serve up content based on the family of sites they have visited — and the predicted traits and interests visitors to that family of sites demonstrate. However, if the first known point of contact with your visitor is their arrival at your site, predicting their targeted area of interest is a tricky proposition. A more effective approach may be performing an A/B test of two broader offers to your universe of site visitors and see which performs better. This would ensure that no audience is excluded or misdirected — and requires less historical data to drive the offer.

How Different Are Your Consumers?

Segmenting your visitors into unique profiles — or personas — is one of the first steps in most behavioral programs. High-performing behavioral targeting programs are based on having a rich profile of your consumers, and comprehensive, differentiated profiles illustrating how each segment is expected to behave. When you are marketing to a broad range of personalities (with different emotional needs and behavioral triggers), creating distinct personas is not a challenge. However, if your core market is limited to a relatively small set of profiles (e.g., information technology executives within financial instructions), there may not be enough variation between the visitors for immediate behavioral targeting to pay off. In cases where the persona set is small, A/B testing offers based on different value propositions may provide greater insight into what is driving consumer behavior and response.

Do You Have a Wide Array of Products and Services?

Are your products diverse enough to support uniquely rendered content based on consumer behavior? For companies with large, diversified product sets, behavioral targeting offers opportunities to suggest products and services within a broad category. Consumers shopping for CDs will have very different behaviors, search terms, and past site views than consumers shopping for sporting goods — which makes the probability of success for delivering targeted content to each group high. But for businesses that have relatively few large and complex solutions, serving up unique, relevant content may not be feasible. Is it worth developing comprehensive data sets and behavioral models if, regardless of the profile, the product and offer are fundamentally the same? If diversity of products is limited, A/B testing messaging that supports the buying decision can deliver category level insights applicable to a broader audience.

Whether you determine behavioral targeting, A/B testing, or some combination of both is the right approach for your business, keeping your focus on delivering relevancy and measuring performance will ensure your site’s optimization is headed in the right direction.

Related reading