Why churn is a reciprocal, not a percentage
Churn is usually reported as a percentage and thought about as a small loss. That framing hides its real behaviour, because the quantity that matters is its reciprocal. If each month a subscriber has a 5% chance of cancelling, they stay on average 1 ÷ 0.05 = 20 months. At 10% they stay 10 months. At 2.5% they stay 40.
So halving churn does not improve things by a few percent — it doubles the average lifetime, and therefore doubles lifetime value. Nothing else in a subscription business has that property. Doubling your price does not double LTV once you account for the subscribers a higher price drives away; halving churn does, exactly, with no offsetting loss.
The same reciprocal shows up in the second result on this page, and it is the one that surprises people. If you add a subscribers a month and lose c of your base each month, the base converges to a ÷ c and stops. Not slows — stops. At 100 signups a month and 5% churn the ceiling is 2,000 subscribers, and at 2,000 the 100 arriving each month are exactly the 100 leaving. You can be doing everything right, adding subscribers every single month, and be asymptotically approaching a wall.
The model, its assumption, and where it breaks
The lifetime formula comes from a geometric survival model. Each month, every subscriber independently has probability c of cancelling. The chance of surviving n months is (1 − c)ⁿ, and the mean of that distribution is 1 ÷ c months. Twelve-month retention is (1 − c)¹², which at 5% monthly churn is 0.95¹² = 54.0% — barely more than half of today's subscribers still paying a year from now, from a rate that sounds trivial.
The assumption baked into that model is memorylessness: a subscriber who has stayed two years is exactly as likely to cancel this month as one who joined last week. Real subscriptions do not behave that way. Cancellation risk is heavily concentrated in the first few months and falls sharply afterwards, so a single blended churn rate understates the loyalty of your long-tenured base and overstates the value of a brand-new signup. If your churn differs sharply by tenure, compute LTV separately for new and established cohorts and expect the true blended figure to sit above what a single rate predicts.
Net revenue per subscriber must be net, not the sticker price. Platform commission, card processing and any per-member fulfilment cost all come off before this number, which is exactly the figure the membership tier break-even calculator and the newsletter revenue calculator produce. Using the gross price inflates LTV by 15–30% at typical fee levels, which is enough to turn a bad acquisition decision into a good-looking one.
Acquisition cost is total acquisition spend divided by subscribers acquired. If growth is entirely organic, enter zero and read the LTV figure on its own; if you run paid promotion, sponsorships or referral incentives, include all of it, including your own time if you value it consistently elsewhere.
Worked example: 800 subscribers, 5% churn, 100 signups a month
Net revenue of $6.67 per subscriber per month and an acquisition cost of $12.
- Average lifetime. 1 ÷ 0.05 = 20 months.
- Twelve-month retention. 0.95¹² = 54.0%.
- Lifetime value. $6.67 × 20 = $133.40.
- LTV to CAC. $133.40 ÷ $12 = 11.12.
- Subscribers lost per month. 800 × 5% = 40.
- Net growth. 100 − 40 = 60 a month. So 40 ÷ 100 = 40% of every month's signups are replacements, not growth.
- Ceiling. 100 ÷ 0.05 = 2,000 subscribers.
- Months to double. The base follows
S(t) = 2,000 + (800 − 2,000) × 0.95ᵗ. Setting S(t) = 1,600 gives 0.95ᵗ = (1,600 − 2,000) ÷ (800 − 2,000) = −400 ÷ −1,200 = 1/3, so t = ln(1/3) ÷ ln(0.95) = (−1.0986) ÷ (−0.051293) = 21.42 months.
Now hold everything else and halve churn to 2.5%. Lifetime becomes 40 months, LTV becomes $266.80, monthly losses fall to 20, net growth rises to 80, and the ceiling jumps from 2,000 to 100 ÷ 0.025 = 4,000. One change, and both the value of a subscriber and the size of the business you can ever build have exactly doubled. Doubling signups to 200 a month instead also lifts the ceiling to 4,000 — but leaves LTV at $133.40 and costs 100 × $12 = $1,200 a month more in acquisition, where the churn fix costs nothing per subscriber.
What each output should make you do
Average lifetime is the number to quote internally, because months are intuitive in a way that percentages are not. Twenty months means a subscriber acquired today is, on average, gone before the end of next year. That reframing usually changes how much attention onboarding and the first ninety days get.
The LTV to CAC ratio is a spending decision, not a scorecard. The widely used rule of thumb is that a ratio of 3 or better means acquisition is worth funding; below 1, every additional acquired subscriber loses money and spending more makes it worse. Between those, you are buying subscribers who repay their acquisition cost but leave little toward producing the thing they subscribed to. Note that a very high ratio is not automatically good news either — it often means you are under-investing in growth and could profitably spend more.
The replacement share is the figure that explains why growth feels harder than it is. At 800 subscribers and 5% churn you must find 40 people a month before the count moves at all, and that requirement grows with the base: at 2,000 subscribers it is 100 a month, which is your entire current signup rate. This is the mechanism behind the ceiling.
The ceiling is the strategic output. If it sits below where you want to be, no amount of patience gets you there — the curve flattens toward that number and stops. Only two inputs move it: signups and churn. And because the ceiling is a ÷ c, a given percentage improvement in churn moves it by exactly as much as the same percentage improvement in signups, while usually costing far less.
Churn, lifetime, retention and value
| Monthly churn | Average lifetime | Still paying after 12 months | LTV at $10/month net |
|---|---|---|---|
| 1% | 100.0 months | 88.6% | $1,000.00 |
| 2% | 50.0 months | 78.5% | $500.00 |
| 3% | 33.3 months | 69.4% | $333.33 |
| 5% | 20.0 months | 54.0% | $200.00 |
| 7% | 14.3 months | 41.9% | $142.86 |
| 10% | 10.0 months | 28.2% | $100.00 |
| 15% | 6.7 months | 14.2% | $66.67 |
| 20% | 5.0 months | 6.9% | $50.00 |
Read the first and last columns together: the relationship is a reciprocal, so the step from 1% to 2% churn costs $500 of lifetime value while the step from 15% to 20% costs $17. Improving already-low churn is worth far more per point than improving high churn — but high churn is usually far easier to improve.
Assumptions and honest limitations
- Churn is assumed constant with tenure. In reality it is front-loaded: most cancellations happen in the first few months. A single blended rate therefore understates the value of long-tenured subscribers and overstates that of new ones.
- Annual plans distort monthly churn. A subscriber on an annual plan cannot cancel in most months and then can cancel decisively in one. Compute annual and monthly cohorts separately, or you will read a misleadingly low rate eleven months of the year.
- Involuntary churn is included but not separated. A meaningful share of cancellations are failed cards rather than decisions. That portion is recoverable through payment retries in a way voluntary churn is not, so measure it separately if you can.
- No discounting is applied. LTV here is the undiscounted sum of future monthly revenue. Over a twenty-month horizon the difference is small; over a hundred-month lifetime at 1% churn it is not, and a formal LTV would discount the stream.
- Revenue per subscriber is assumed flat. Price rises, tier upgrades and downgrades all change it over a subscriber's life, and upgrade-heavy businesses genuinely earn more than this model reports.
- The ceiling assumes signups stay constant. They rarely do — they usually rise with audience size, which pushes the ceiling up over time. Treat the ceiling as the answer to "what if today repeated forever", not as a prophecy.
- LTV excludes the cost of serving the subscriber. Net revenue per subscriber should already have fulfilment in it; if it does not, the ratio to CAC is flattering you.
Reducing churn beats every other lever
Because lifetime is a reciprocal, retention work compounds in a way acquisition work does not. A subscriber you keep costs nothing to reacquire, raises the ceiling, raises LTV and improves the LTV-to-CAC ratio simultaneously. A subscriber you acquire improves only the count, and costs money each time.
The practical implications follow from where churn actually happens. It is concentrated early, so the first month matters disproportionately: whether the subscriber ever used what they paid for, whether the value was obvious immediately, whether they received anything at all in the first week. It is also concentrated in payment failure, which is a solved problem — retry logic and expiry reminders recover a meaningful share of cancellations that nobody intended.
Annual billing is the blunt instrument that works. It converts twelve monthly cancellation decisions into one, and it removes eleven chances for a card to fail. The trade is that you are paid before the work, and that a year's churn arrives all at once on the renewal date rather than spread out.
Finally, use LTV to decide what growth is worth buying. If a subscriber returns $133 and costs $12 to acquire, cross-promotion, referral rewards and paid growth are all obviously worth funding. If they return $30 and cost $25, they are not, and the fix is on the retention or pricing side first. The inputs to that decision come from the net figures produced by the newsletter revenue calculator and the membership tier break-even calculator, and the alternative income lines they compete with are modelled in the sponsorship rate calculator and the YouTube ad revenue calculator.
Key terms
- Monthly churn rate
- Subscribers who cancel in a month divided by subscribers at the start of that month. Measured on subscriber counts, not on revenue — revenue churn is a different metric that can be negative when upgrades exceed losses.
- Lifetime value (LTV)
- The total net revenue expected from one subscriber over their whole relationship with you. Here it is undiscounted and computed before acquisition cost.
- Customer acquisition cost (CAC)
- Total spend on acquiring subscribers divided by the number acquired in the same period. Include everything spent to win them, not only advertising.
- Involuntary churn
- Cancellation caused by a failed payment rather than a decision to leave. It is recoverable through retries and card-expiry reminders, which makes it worth tracking apart from voluntary churn.