Aiming for 1:1 Personalization? Here’s a Better Approach.

For most of the last decade, A/B testing has been the cornerstone of most site optimization programs.

We learned that measuring in controlled environments can give us tangible ROI while empowering marketers to prove their campaigns’ worth. At the beginning, big wins were plenty, and testing took off.

As those programs and experiences became more optimized, and programs imitated (flattered?) other programs’ successes to develop their own best practices, the big sitewide wins were a bit more difficult to find.

Segmentation became the way to look, post-hoc, at test results and find big wins on portions of the population. (The appropriateness of that “win” is another conversation.)

The new, shiny idea is personalization. The things we have learned in almost a decade of A/B testing should inform the way we embrace personalization—and we can do it profitably without waste.

A personalization continuum

In most cases, testing programs measure potential impact. You choose sections of the site to test, then test that experience. As Reid Bryant, our VP of data science and analytics, says, this is the “1:Most” delivery—one experience for most people, informed by data and psychology.

Some advanced programs are in the “1:Few” section of the testing and personalization continuum. That is, one experience feeds segmented portions of the overall population.

Sometimes that segment is purchased (third-party data), or dictated to the program by the highest paid person in the room or, best of all, uncovered through analysis of previous test results. Rarely, however, are those segments as precise and effective as we believe they could be.

The wish is for 1:1 personalization. 1:1 sounds great. The idea of delivering content uniquely targeted to one person is alluring. It seems like a guaranteed way to move the needle.

Indeed, at the Tealium Digital Velocity 2016 conference, I saw companies that have identified over 17,000 segments of customers. The problem is that those segments don’t have 17,000 different desired endpoints on the site. They don’t need 17,000 different experiences!

(The creative and development teams are thrilled about that, by the way.)

The best place to be right now is in the 1:Few group        

Ramping up your personalization program to 1:Few is an evolutionary process that is scalable and exciting for your culture of optimization.

Give the data team the power to find predictive, contextual behaviors to define segments. Data scientists and analysts thrive on researching, not just reporting results!

Answering questions that begin with, “How could I have predicted…?” is much more fun than answering, “How much did they buy?”

This kind of ownership gets everyone excited about a culture that will breed tangible wins.

Today’s analytics toolset provides the opportunity to know a lot about your customer, not just the computer they’re using. This data can come from offline, online, yesterday, today or three years ago.

Finding the mix of longer-term information and right-now context that best indicates conversion will be the path to personalization.

Inversely, finding the mix of behaviors that best predicts a never-convert customer can save large amounts in advertising and retargeting efforts. Identifying both positive and negative correlations will help you maximize ROI.

An A/B personalization program is already attainable—and it will only continue to grow with the omnichannel information you can stitch together about your customers.

This information will give your data teams the power to use predictive methods and your marketing teams the power to do great things to influential groups.

How is this A/B personalization?

The process of A/B testing is data-informed, statistically rigorous, and iterative. A/B personalization needs to be the same iterative, evolving process that is improved through testing. Personalization programs need to be measured as a whole for ROI and tested iteratively on the “few” groups that most valuably maximize that ROI.

Personalization is a program, not a project.

Dave Rose has worked with many retail and subscription clients as a senior optimization analyst at Brooks Bell and now leads the consulting group in analytics. 

—

Brooks Bell helps top brands profit from A/B testing, through end-to-end testing, personalization, and optimization services. We work with clients to effectively leverage data, creating a better understanding of customer segments and leading to more relevant digital customer experiences while maximizing ROI for optimization programs. Find out more about our services.