Case Study: Winning With Personalization

What follows is a composite scenario reflecting a pattern common among online education businesses in 2022, not a report on one named company. It shows how personalization strategy shifted that year as the effects of Apple’s tracking changes settled in for good.

The situation

Consider a mid-sized online course platform that had spent 2020 and early 2021 growing primarily through Facebook retargeting: someone would browse a course page, drop off and get shown an ad reminding them to finish enrolling. That playbook worked reliably until Apple’s App Tracking Transparency rollout in April 2021 cut deeply into how many of those browsing events the platform could see and act on. By 2022, Meta itself had disclosed the scale of the industry-wide impact in its own earnings calls, and this platform’s retargeting-driven conversions had fallen well below what they used to be.

The pivot

Rather than trying to recover the old retargeting volume, the team built a short onboarding quiz asking new visitors about their skill level, learning goals and available time per week, three questions, under a minute to complete. This was zero-party data: information volunteered directly, with clear intent, rather than inferred from browsing behavior. The quiz results drove the homepage and course catalog a visitor saw immediately, surfacing beginner content to beginners and advanced tracks to people who said they already had experience, instead of a generic bestseller list.

Extending it past the first visit

The same quiz data fed email sequences after signup, so a visitor who said they had five hours a week available received a different pacing suggestion than someone who said one hour. Returning visitors who hadn’t finished a course got a personalized nudge based on how far they’d progressed, not a generic “come back” message. None of this depended on third-party cookies or ad platform pixel data, since it all lived in data the platform collected and owned directly.

The results

Over the following two quarters, the platform saw completion rates for quiz-takers run meaningfully higher than for visitors who skipped it, and email click-through rates on the personalized pacing sequence outperformed the platform’s previous generic newsletter by a wide margin. Paid acquisition costs didn’t fully return to pre-2021 levels, no first-party approach fully solved that industry-wide shift, but a larger share of revenue was now coming from a channel, onsite and email personalization, that wasn’t exposed to whatever the next platform tracking change might be.

What made this work

The quiz succeeded because it asked for useful information in exchange for a better experience, not because personalization technology itself improved. The lesson for a comparable business isn’t “add a quiz.” It’s that personalization built on data customers hand over willingly tends to outlast personalization built on data inferred from tracking they never explicitly agreed to.