The results

The streaming media client leveraged Subsets to build a sophisticated AI-powered lifecycle automation engine, running dozens of experiments and multiple always-on programs that reach 10 million monetizable users with targeted emails and in-app messages across the full subscription journey.

Impact (18 experiments)

• +296% engagement of high-risk audience
• +109% new trials started and +158% trial converted
• +10.1% increase in retention of trial subscribers

The challenge

Despite its scale, the company faced three subscriber lifecycle challenges.

• Low-tenure subscribers within 125 days of sign-up were approaching renewal with very low engagement. These users represented elevated churn risk, but traditional CRM segmentation lacked the behavioral context required to intervene effectively.
• Free users experiencing significant ad load, particularly the top 25% most exposed,  were showing declining engagement. Many of these users consumed monetized content but were not converting to Premium, indicating monetization friction.
• High-risk trial subscribers were not experiencing meaningful value before trial end. Standard onboarding sequences were not sufficient to trigger conversion behavior in this cohort.

The team needed a way to surface high-impact audiences, run structured experiments without engineering dependency, and measure retention and conversion lift.

The solution

Using Subsets, the company was able to identify audiences using AI to identify behavioral risk patterns across the subscription lifecycle.

For the low-tenure subscribers approaching renewal, Subsets created cohorts with declining engagement. The team launched personalized journeys serving genre-specific content aligned with each subscriber’s listening preferences. This directly addressed usage decline before renewal and resulted in +296% engagement.

By identifying high-ad-load free audiences with declining activity and monetized listening behavior, the team deployed targeted upsell journeys promoting ad-free Premium plans, aligning monetization messaging with observed friction signals. This ramped up the new trials by +109% and increased trial conversions by +158%.

Finally, the high-risk trial subscribers were enrolled in onboarding flows with more assertive, personalized journeys toward the end of the trial period. These interventions were tested against control groups and measured across renewal outcomes.

Throughout each experiment, Subsets automatically tracked engagement, trial progression, and retention metrics in real time, validating statistical significance and enabling proven interventions to be turned into always-on lifecycle flows.

‍