Retention intelligence for an education subscription business.
Turning a customer base into actionable cohorts.
Weekly data-driven outreach replacing monthly gut-feel reviews.
Sector
Education, SaaS
Tech
pandas, segmentation models, Claude API, HTML reports
The challenge
A consumer edtech had a growing subscriber base but was losing customers before they hit the activation moment that drove long-term retention. The retention team was working from gut feel. There was no clean view of which cohorts churned, why, or when the drop-off actually started.
Our approach
- —Segmented the subscriber base by cohort, product, and usage pattern.
- —Surfaced the behavioural signals that precede churn. The quiet disengagement moments that happen before a customer cancels.
- —Built the dashboard the retention team uses weekly to prioritise outreach.
- —Paired the work with a strategic proposal on rebuilding their broader operations stack to be AI-first.
The outcome
A clear picture of where churn happens and why. A weekly-updated dashboard that drives the retention team's outreach calendar. The retention conversation inside the business shifted from 'we feel like we are losing users' to 'this cohort, at this point, for this reason.'
Where the real work was
Defining what activation actually meant for this product. Without a clear activation event, churn analysis has nothing to anchor to. We spent the first week just getting that definition right.
Tags