Why a flat churn rate hides where you're actually losing money
If you quote a single monthly churn percentage, you're averaging two very different populations. A 1,000-person new-subscriber cohort from a discovery-box brand loses roughly 18% of its members in month 1 alone, 13% of the remainder in month 2, and 10% in month 3. By the start of month 4 you're down to about 650 of the original 1,000 — and that surviving 650 churns at 5-7% per month for the rest of the retention curve. Quoting the 12-month average (~8.5%) makes the product look healthier than it is and obscures the real problem: the first 90 days.
This analyzer models the cohort curve explicitly. It applies an early-stage multiplier (1.6x in M2-M3 for boxes, 1.4x in M1 for apparel, 1.25x in M1 for consumables) that matches what we see in actual operator data across ~40 DTC subscription brands. The result is a retention curve that decays fast through M6, then flattens — the shape that matches reality, not the exponential-decay curve you get from applying one churn rate.
Category-by-category churn benchmarks
| Category | Monthly churn | M1 retention | M12 retention | Typical LTV months |
|---|---|---|---|---|
| Discovery box | 8-12% | 78-84% | 22-32% | 5.8-8.2 |
| Apparel subscription | 10-14% | 80-86% | 18-28% | 5.0-7.1 |
| Consumable (coffee/supp) | 4-7% | 88-94% | 40-54% | 10-16 |
| Pet food / care | 3-5% | 91-95% | 52-68% | 14-22 |
| SaaS / digital | 3-6% | 92-97% | 48-65% | 14-24 |
Three cohort scenarios
Scenario 1: Discovery box · 1,000 cohort · $32 ARPU · $48 CAC
A curated snack box brand, 1,000 new subscribers signed up in one month, $32 monthly ARPU, $48 blended CAC. Plug in 9% monthly churn + "discovery box" category. The analyzer shows: M1 retention 82%, M6 retention 43%, M12 retention 21%. Total LTV lands around $220 — LTV/CAC ratio of 4.6x. CAC payback is ~2.3 months. Healthy unit economics. But the M1 cliff still loses 180 subscribers — worth $96k in lost lifetime revenue on this cohort alone. Fixing M1 to 87% retention (via onboarding improvements) adds $21k of LTV to this 1,000-person cohort and compounds across every future cohort.
Scenario 2: Apparel subscription · 500 cohort · $58 ARPU · $95 CAC
A women's styling subscription, 500 new subscribers, $58 ARPU, $95 CAC. Category = apparel, 11% monthly churn. Results: M1 retention 85%, M6 retention 38%, M12 retention 17%. LTV ~$304, LTV/CAC 3.2x — at the healthy floor but not strong. The problem is the M3-M4 window: subscribers have received two or three boxes, decided the fit/style doesn't work for them, and cancel before the fourth charge. Stitch Fix famously solved this with AI-driven personalization that improved M4 retention 12 points. For this brand, the lever is a human stylist follow-up at day 75 offering to rework the style profile — historically this converts 18-25% of would-be cancellers into retained subscribers.
Scenario 3: Pet food consumable · 1,200 cohort · $42 ARPU · $65 CAC
Small-batch dog food brand, 1,200 new subscribers, $42 ARPU (for a medium-size dog box), $65 CAC. Category = consumable, 5% monthly churn. Output: M1 retention 94%, M6 retention 74%, M12 retention 56%, M24 retention 31%. LTV $640+, LTV/CAC ratio ~9.8x, CAC payback 2.4 months. This is the business model venture funds write checks for — consumable subscription with strong retention compounds into enormous LTV once the cohort matures. The operator lever here is not retention (already excellent) but ARPU expansion: adding a $14/mo toy-and-treats add-on lifts ARPU 33% and flows straight through to LTV because the retention curve is unchanged.
The four churn-fix levers, ranked by impact
- M1 onboarding + first-box delight. Every percentage point of M1 retention gain compounds through the entire retention curve. The single highest-ROI retention project you can run.
- Day-20 to day-28 pause flow. Replace the "cancel" button with a "pause for 30/60/90 days" option on days 20-28 of the cycle (before the next charge). Pausers reactivate at 40-55% vs 8-12% for cancellers.
- Annual billing upsell at month 2. Subscribers who loved their first two boxes are primed for an annual offer. Converting 20% of a cohort to annual at a 15% discount lifts blended 12-month retention 8-14 points.
- Win-back sequence at day 60 post-cancel. 15-22% of cancellers reactivate with the right offer (free box + free shipping). Cheaper than net-new CAC by 60-75%.
What this analyzer doesn't model
Seasonal cohort effects — a BFCM-acquired cohort churns 20-35% worse than a March-acquired cohort because discount-driven signups have lower intent. If you know your Q4 cohorts churn worse, run the analyzer separately for Q4 vs non-Q4 acquisitions and blend the LTVs.
It also doesn't handle reactivations. A brand with a good win-back program will see 8-15% of churned subscribers return within 6 months, which flattens the 18-24 month tail of the retention curve. That's a real LTV lift — budget it separately and don't double-count it as retention.
Related tools
Pair this with CAC calculator to recompute LTV/CAC after a CAC change, profit margin calculator to confirm your contribution margin assumption, and pricing calculator if an ARPU lift is on the table.