Perlego maximized subscriber value by optimizing Google acquisition campaigns to predictive LTV signal
Making Learning Affordable: Perlego’s Origin Story
Founded in 2017 in London, Perlego is a digital library boasting unlimited access to over a million academic books across 900 disciplines and five languages. Perlego was born from the personal experience of its founders, Gauthier Van Malderen and Matthew Davis, who recognized the challenge of the rising textbook costs for students. Upon graduating from university, they set out to create a solution to make accessing essential educational material more affordable and convenient for students everywhere. The result is an ‘all you can read’ buffet for hungry learners.
“Implementing Churney’s predictive LTV model has been a game-changer for our user acquisition strategy. By identifying high-potential subscribers early, we’ve not only increased the lifetime value of our customers but also significantly improved retention rates and overall campaign performance. For any subscription business looking to maximize growth and profitability, adopting a pLTV approach is essential.”
Director of Organic & Paid
Optimizing Growth with a Lifetime Value Strategy
Perlego, a subscription-based platform, provides users with flexible access plans—monthly, semesterly, and yearly—granting full access to its extensive library and features. Recognizing the diverse behavior of its subscribers, with some opting for short-term use and others remaining committed for years, Perlego sought to explore whether a lifetime value-based strategy could enhance their user acquisition efforts. The goal was to scale their growth more aggressively while ensuring sustained profitability.
Identifying Lifelong Learners Early & Leveraging This Insight
Perlego faced two main challenges in enhancing their Google Ads targeting to attract lifelong learners—users who would maximize long-term subscriber value. First, they needed to find a reliable method for early identification of high-potential, long-term users who would benefit most from their offerings. Additionally, there was the challenge of effectively communicating to the ad network which newly acquired users were most likely to become lifelong learners.
Learning new insights with Churney’s predictive LTV model
Perlego collaborated with Churney by providing access to a PII-preserving version of their first-party data, which included detailed information on how users interacted with the platform throughout their journey as learners. Churney utilized this data to develop and train its causal deep-learning predictive LTV models. These models were designed to predict a 3 months future customer value based on the consumption patterns observed within the first 24 hours of a newly acquired learner’s engagement.
Once the predictive LTV (pLTV) model was established, Churney created an activation pipeline that fed the model’s output directly into the Google Ads platform. A dedicated campaign was then launched on Google Ads, using the pLTV signal provided by Churney as the primary optimization event for the ad network. To ensure ongoing effectiveness, Churney’s model continuously monitored the performance of newly acquired users, automatically adjusting for any model drift or unexpected feedback loop outcomes.
Leading the class
After just two months, Perlego realized they could significantly enhance the quality of newly acquired users. This improvement was evident in user engagement, as Churney’s users demonstrated higher activity levels—measured by the average number of books read by subscribers—over time: 25.6% more active after 1 month, 26.8% more active after 2 months, and an impressive 36.0% more active after 3 months.
From a performance standpoint, Churney’s optimization efforts yielded remarkable results for Perlego’s campaigns.
