Turning Casual Viewers into Subscribers
Overview
Our client was acquiring subscribers efficiently and losing them just as efficiently, which meant the growth was rented rather than owned.
Mohamed El Alali
Sr. Marketing Consultant / fairsystems
Subscription businesses are often measured on acquisition because acquisition is easy to attribute. Retention decides whether the business works. When churn runs ahead of the payback period, every marketing pound buys a subscriber who leaves before they have paid for themselves. fairsystems was engaged by a regional streaming platform whose growth numbers looked healthy and whose economics did not.
The Client
Our client operates a subscription video platform serving several markets with a mix of licensed and original content. Subscriber acquisition was strong and marketing was sophisticated, but monthly churn had crept upward for six consecutive quarters. The team had a great deal of engagement data and no clear model of which behaviours in the first weeks predicted whether a subscriber would still be there in six months, so retention effort was spread evenly rather than aimed.
Roadmap
1
Assess
fairsystems analysed the full subscriber lifecycle, linking acquisition channel, onboarding behaviour, content consumption and cancellation into a single view the client had never assembled. Our consultants built a churn propensity model and, more usefully, identified the specific early behaviours that separated retained subscribers from lost ones. The finding that reframed the engagement was that subscribers who did not complete a second title within their first fortnight churned at several times the base rate, and that a large share of acquisition spend was going to channels that reliably delivered exactly those subscribers.
2
Deliver
Our team rebuilt onboarding around that fortnight, replacing a generic welcome sequence with personalised recommendations calibrated to get a new subscriber to a completed second title. fairsystems reallocated acquisition spend away from channels whose subscribers systematically failed that test, accepting a lower headline acquisition number for materially better cohort economics. We introduced targeted intervention for accounts the model flagged as at risk, and worked with the content team so that commissioning decisions were informed by retention contribution rather than by launch-week viewing alone.
3
Continue
fairsystems supports a quarterly cohort review in which acquisition, product and content teams look at the same retention picture, which had previously been three separate conversations. We retrain the propensity model as the content slate changes and extend it as new markets launch. Our consultants also help the client run structured retention experiments with proper controls, so that changes to onboarding and intervention are evaluated on evidence rather than adopted because the following month happened to look better.
Solution Details
Acquisition is a cost. Retention is the business.
Growth measured on acquisition rather than cohort economics
Deliverable: fairsystems rebuilt reporting around cohort payback, making it immediately visible which channels delivered subscribers who never paid for themselves.
No model of which early behaviours predicted retention
Deliverable: Our consultants identified the specific first-fortnight behaviours separating retained from lost subscribers, giving retention effort a target.
Generic onboarding for every new subscriber
Deliverable: We rebuilt onboarding around reaching a completed second title within two weeks, the single strongest predictor of long-term retention.
Acquisition spend flowing to reliably churning channels
Deliverable: fairsystems reallocated budget on cohort quality rather than cost per acquisition, trading headline volume for subscribers who stay.
Retention effort spread evenly across the base
Deliverable: Our team introduced propensity-based intervention so effort concentrates on accounts that are both at risk and worth saving.
Content commissioned on launch-week viewing alone
Deliverable: We gave the content team a retention contribution measure, so commissioning weighs a title's effect on subscriber life rather than its opening weekend.
Acquisition, product and content working from different numbers
Deliverable: fairsystems established a quarterly cohort review on shared metrics, replacing three separate conversations with one.
Retention changes adopted without controlled evaluation
Deliverable: Our consultants introduced properly controlled retention experiments, so what gets kept is what demonstrably worked.
Result:
Monthly churn reversed its six-quarter trend and cohort payback shortened materially, even though headline acquisition volume fell as spend moved away from channels that had been delivering subscribers who never stayed. The client trades acquisition vanity for subscriber lifetime deliberately now, with the numbers to justify it. Onboarding is aimed at the behaviour that actually predicts retention, and content, product and marketing finally review the same cohort picture rather than three versions of it.
0
Monthly churn reduction
0
Cohort payback period reduction
0
Second-title completion in first fortnight
0
Subscriber lifetime value increase
0
Acquisition cost per retained subscriber


