Sign-ups don't
pay the bills.
Subscription and membership businesses live and die on retention curves. We build acquisition, onboarding and retention as one system — because they share the same P&L and the same customer.
Growth is retention plus new — in that order.
A subscription business that acquires 10,000 members and loses 8,000 in the first cycle is running an expensive marketing programme, not a growing business. LTV, cohort churn curves, payback and reactivation matter more than any acquisition metric — and each is a distinct operational programme, not a dashboard tile.
Acquisition → Trial/Join → Activation → Engagement → Renewal → Reactivation.
- Cohort economicsRetained recurring revenue
- 01Acquisition · cohortBy channel, plan, offer and creative
- 02Trial / joinThe moment expectations are set
- 03Activation · first valueFirst successful use, not first checkout
- 04Engagement · retentionRecurring value, not just billing
- 05Renewal · churn riskVoluntary and involuntary — modelled separately
- 06Reactivation · win-backExpected-value scored, continuously
Retention is the curve, not a KPI.
Illustrative cohort retention. Attributed measures each acquisition cohort by channel, plan and offer against the same curve — so honest payback is visible before the marketing budget is committed.
Where subscription growth stalls.
- Paid optimised to trial sign-ups, ignoring trial-to-paid conversion
- Onboarding designed for the checkout, not the first successful use of the product
- Low activation — new members never reach the moment that predicts retention
- Early churn concentrated in the first billing cycle, invisible in blended metrics
- Failed payments (involuntary churn) treated as billing admin, not a retention programme
- Weak dunning and card-update flows lose members who wanted to stay
- Pause, downgrade and cancel flows treated as legal admin, not retention
- Dormant members left alone until they cancel, when reactivation was cheaper
- Discount-led acquisition inflates sign-ups and destroys second-cycle retention
- Reactivation programmes running once a year, not continuously by expected value
- LTV assumed rather than measured against contribution margin
What Attributed builds for subscription and membership.
- Paid targets rebuilt around trial-to-paid and retained MRR, not trial sign-up count
- Onboarding flow tied to activation events that measurably predict retention
- Landing pages, pricing and comparison patterns tested honestly
- Referral and gifting mechanics designed to retain, not just acquire
- Discount discipline — offers priced against second-cycle retention impact
- First-value time reduced through onboarding, not just email nudges
- Cohort LTV and churn curves by acquisition channel, plan and offer
- Voluntary vs involuntary churn tracked separately, each with its own programme
- Dunning, card-update and payment-retry flows rebuilt as retention
- Pause, downgrade and cancel flows redesigned as retention moments
- Reactivation and win-back programmes with expected-value scoring
- Weekly view of retained MRR / ARR against contribution margin
What we sequence first.
- 1Define the activation event
Identify the specific in-product moment that predicts second-cycle retention — and instrument it end-to-end.
- 2Rebuild onboarding around it
Trim onboarding to what drives activation. Everything after the checkout is a retention programme, not a marketing one.
- 3Split voluntary and involuntary churn
Failed payments, expired cards and dunning are a separate programme from cancel intent — treat them that way.
- 4Instrument cancel and pause
Redesign cancel flows as retention moments with pause, downgrade, offer and reason capture — measured by saved MRR, not deflection.
- 5Reactivate by expected value
Score dormant and churned members continuously; run reactivation against expected LTV, not blanket discounts.
- 6Reconcile weekly
Retained MRR, CAC payback and LTV:CAC by cohort — reviewed with senior operator sign-off, not automated.
What the intelligence layer surfaces.
- Findings — cohorts where activation is collapsing, or channels whose sign-ups never reach paid
- Recommendations — onboarding, dunning and cancel-flow changes tied to retained MRR impact
- Alerts — payment-failure spikes, cancel-reason clusters, or activation drops in a live cohort
- Implementation Packs — activation event instrumentation, dunning rebuild, cancel-flow redesign, reactivation scoring
What we report against.
Sign-ups don't compound. Retained revenue does.
Paid targets rebuilt around trial-to-paid and retained MRR, not trial sign-up count.
Onboarding tied to the activation event that measurably predicts retention in your cohorts.
Cancel, pause, downgrade and dunning flows redesigned as retention moments with expected-value modelling.
Cohort LTV, churn curves, payback and reactivation scoring modelled honestly against contribution margin.
Related routes into the same system.
High-volume enquiry economics: speed-to-lead, qualification and revenue reconciliation.
Reconciled contribution margin across paid, organic and CRM.
Cohort retention, dunning, reactivation and LTV as first-class programmes.
The proprietary layer that reconciles cohorts against retained recurring revenue.
Book a Growth Review.
Bring your subscription platform and billing data. We'll walk through your retention curves in the session and show where retained MRR is actually being lost.