What a point of churn actually costs you

Churn is the quietest number on the management dashboard. Nobody celebrates it, few interrogate it, and almost nobody prices it. Yet a single percentage point of monthly churn, compounded across a year, can be worth more than most of the growth initiatives competing for the same budget. Here is how to put a defensible number on it.

Know your baseline before you benchmark

Subscription businesses average around 3.27% monthly churn — 2.41 points voluntary, 0.86 involuntary (failed payments and the like) — with 1–5% described as the normal band (Recurly). That involuntary slice deserves attention on its own: roughly a quarter of average churn is customers who didn't decide to leave, which is usually the cheapest churn to fix.

But averages hide the split that matters. In the same data, B2B businesses churn around 3.8% monthly while D2C runs around 6.5% (Recurly). B2B SaaS specifically sits near 3.5% monthly (Recurly via Vitally), US premium streaming averages 4.6% monthly (Antenna, 2025), and at the sticky extreme, T-Mobile reported postpaid phone churn of just 0.89% a month (T-Mobile, Q3 2025).

Two warnings for anyone quoting benchmarks in a business case. First, never mix monthly and annual figures: B2B annual churn benchmarks run from around 11% in energy and utilities through 19% in financial services to 31% in telecom (CustomerGauge) — an annual 19% is roughly 1.7% monthly, nothing like a monthly 19%. Second, in B2B the customer count is not the whole story: median net revenue retention for mid-ACV SaaS firms is about 102% (SaaS Capital, 2025), meaning expansion from surviving accounts can offset logo churn entirely. A D2C subscription business has no such cushion — every lost subscriber is lost revenue.

The revenue maths of one point

Here is an illustrative example — invented numbers, real mechanics.

Take a business with 10,000 customers paying £40 a month, churning at 4% monthly. That is 400 customers lost every month. Cut churn by one point, to 3%, and you keep 100 additional customers each month.

The naive version of the value is 100 × £40 = £4,000 a month. The real version is bigger, because retention compounds: the 100 customers saved in January are still around (mostly) in February, when another 100 are saved, and so on. Allowing for the saved customers themselves churning at the new rate, by month twelve the retained cohorts stack up to roughly 1,000 extra active customers, worth about £40,000 in that month alone — and the cumulative revenue difference across the year comes to roughly £280,000. One percentage point. No new marketing spend, no new sales headcount.

This compounding is why the academic finding on retention is so striking. The peer-reviewed customer lifetime value literature estimates a retention elasticity of 3–7 — improving customer retention by 1% is likely to improve customer and firm value by 3–7%, against an elasticity of about 1 for margin improvements and only 0.02–0.3 for acquisition spend (Gupta, Lehmann & Stuart, 2004). On those estimates, a point of retention is worth an order of magnitude more than a point of acquisition.

Handle the famous claims with care

Two retention statistics appear in almost every deck, and both need handling carefully if your case is going to survive review.

The first is "a 5% increase in retention increases profits by 25–95%". That range is actually two separate claims welded together: "more than 25%" was about financial services (Bain, 2001), while the 95% figure came from a study of dot-com-era e-commerce companies (Reichheld & Schefter, HBR, 2000). Cite whichever applies to your sector, with its year — or better, lead with the Gupta elasticity above, which is peer-reviewed.

The second is "acquiring a customer costs 5 to 25 times more than retaining one". This appears in a Harvard Business Review article which itself provides no citation and hedges with "depending on which study you believe" (HBR, 2014). Use it as an HBR-published rule of thumb if you must, and label it as exactly that. A reviewer who catches you presenting it as research will discount everything else in your document.

Building the case

The structure that survives scrutiny looks like this:

• Baseline honestly. Your current churn rate, split voluntary versus involuntary, benchmarked against your sector — B2B and D2C are different worlds.
• Model the improvement as a range. Nobody can promise a one-point reduction. Model half a point, one point, and a stretch case, and show the compounding month by month rather than a flat annual multiplier.
• Show the arithmetic. Cohort-by-cohort retention maths is exactly the kind of calculation that should be visible line by line, not buried in a spreadsheet tab.
• Commit to measuring it. Churn moves for many reasons. A projection that names its baseline and promises a post-implementation comparison is far harder to dismiss than one that doesn't.

---

The churn reduction calculator runs exactly this cohort maths from your own numbers — conservative, moderate and optimistic scenarios, a full assumptions audit trail, and multi-year NPV so the compounding is priced rather than asserted. The customer retention calculator approaches the same economics from the lifetime-value side. Your first calculator is free, with unlimited re-runs.