The Personalization Spectrum – Bridging the Gap Between Testing and Personalization

It’s not uncommon in the digital marketing world to hear the mandate: “Personalization is our top priority this year.” But what does that really mean? Personalization is not a single thing to be had; it’s actually a spectrum of carefully planned strategies. And if you’re on the testing side of the chasm, personalization can seem pretty darn far away.

Fully automated personalization - machine-controlled tailored, unique experiences for every individual who visits a site, gets an email, or sees an ad - is of course the ultimate goal, but typically requires a pretty decent investment in both time and capital. Additional tools and data are usually required. Many times additional resources are needed to bring new expertise. But there are steps you can take to continue to move your optimization program forward as you build out your personalization toolset and team.

By using your existing testing tools and focusing on “Rules Based” or “Audience Based” targeting, you can bridge the gap between business-as-usual testing and automated personalization. It might also be a great way to show the true value in more personalized experiences while you’re making the case for those initial investment dollars.

1.    Start simply - Single Focused Audience Based Targeting

Using your testing platform, start with a simple, single focus by analyzing the data available to you. Target audiences based on a single factor like location, channel, device type, or a single onsite behavior like whether the visitor is new to the site or making a return visit.

Your audience size will be fairly large and as long as you have adequate traffic to your site, statistical significance of your test should still be relatively attainable.

Implementation of your personalized experiences will take place in the form of testing on your key pages (think biggest ROI, most trafficked, etc.) and modal windows. Variations will focus on imagery and light content changes.

In this solution, you’re providing a very simplified form of personalized experiences but still learning valuable information nonetheless.

2.     Get more granular - Multi Focus Audience Based Targeting

Once you’ve got those single focus experiences under your belt, you’re ready to graduate to slightly more complex, multi-factor audiences but still utilizing available tools and data.

Again using your testing platform as your tool du jour and with the help of your DMP, expand your targeting factors to incorporate multiple data sources (i.e. 1st & 3rd party) as well as multiple behaviors like a returning visitor who also viewed two product pages but did not make a purchase previously.

Implementation of these personalized experience will take place in the form of testing in all the similar key areas as your single focus tests, followed by continued iteration, and can also broaden from single page personalization to multi page experiences (for instance, updating each page of the conversion funnel to help highlight benefits most important to the specific audience).

This solution is bringing you closer to more, deeply personalized experiences but still requires a lot of manual data analysis and configuration to execute.

3.     Automate & Complexify Automated Personalization

And finally, the crème de la crème, fully automated testing. By the time you reach this step, you will need a personalization tool at your fingertips. Targeting factors are robust for these experiences; automated based on algorithms incorporating all available data to reach all visitors (known & unknown) at all moments across the customer journey.

Execution will expand across the customer journey both onsite and offsite and your target audiences will be small so testing may not be as feasible. Personalization tools will also auto adjust audiences to maximize performance which goes beyond the abilities of most testing tools.

Note that “automated personalization” is only touched on here as an end to a means and deserves its own separate blog (or several) to really talk about the full potential and offerings (ie. Prescriptive, predictive, fully automated, etc).

So, as with most things, personalization requires different strategies depending on what tools, data, and dollars you have available to you. By adopting some of the tactics above, you can connect testing and personalization in a meaningful way, as you strive to achieve a fully automated solution simultaneously offering your customers a better, more personalized experience. Happy optimizing!