It depends on your price point, where in the lifecycle it happens, and how many customers you have. Most benchmark articles ignore all three.
You already know your rate. This checks it against bands that apply at solo scale rather than enterprise benchmarks.
Bands are directional, drawn from published 2026 self-serve data. See sources below.
Whether your churn is too high depends on three things most benchmark articles ignore: your price point, where in the customer lifecycle it happens, and how many customers you have.
Retention correlates strongly with price. Published data indicates products under roughly $25 average revenue per customer churn several times faster than those above $1,000, which means a single benchmark applied across price points is close to meaningless.
| Your ARPU | Reasonable band | Concerning above |
|---|---|---|
| Under $25/mo | 5–9% monthly | 10% |
| $25–$100/mo | 3–7% monthly | 8% |
| $100–$500/mo | 2–5% monthly | 6% |
| Over $500/mo | 1–3% monthly | 4% |
If your rate looks alarming against a widely quoted 1% or 2% figure, check what that figure was measuring. Enterprise annual contracts and self-serve monthly subscriptions are different businesses.
Two products at 7% monthly can be in completely different positions.
Clustered in the first 60 days means people are leaving before they got value. That is an onboarding problem, it is fixable, and it is the most common pattern in early SaaS — the onboarding framework covers closing the gap to first value.
Spread evenly across cohorts means people get value and then stop needing it, or the value fades. That is harder and usually points at product or segment.
Concentrated after six months often means the underlying job ended — which can be structural rather than fixable, depending on whether there is a recurring use case adjacent to the one-off one.
Below roughly forty customers, your monthly rate moves several points on a single cancellation. That is not a small margin of error — it is larger than the difference between a healthy and an alarming rate.
This is the same sample-size problem that makes most published metrics unreadable at solo scale — the four metrics that matter under $10k MRR covers which ones survive. Use a three-month rolling figure, and treat direction as the signal rather than level. If you need to compute the rate itself from raw numbers, the churn rate calculator handles customer and revenue churn together.
Check involuntary churn before concluding anything. Failed and expired cards are commonly a substantial share of total cancellations, and they are not a retention problem — they are a billing problem with a much cheaper fix.
Bands here are directional, drawn from published 2026 data on self-serve and small SaaS retention. Where sources disagree the range is given rather than a single figure.
The wider solo SaaS benchmarks page covers trial conversion and involuntary churn alongside these figures, and is explicit about which numbers have no reliable benchmark at this scale.
It depends on your price point. Under $25 average revenue per customer, 5–9% monthly is a reasonable band. Between $25 and $100, 3–7%. Above $500, 1–3%. A single benchmark applied across price points is close to meaningless.
For self-serve products between $25 and $100 a month, roughly 3–7% monthly is normal territory and under 3% is strong. Widely quoted figures of 1–2% usually come from enterprise annual contracts, which is a different business.
More than the level does. Churn clustered in the first 60 days is an onboarding problem and is fixable. Churn spread evenly across cohorts points at product or segment. Churn after six months often means the underlying job simply ended.
Around forty as a minimum. Below that, a single cancellation moves your monthly rate by several points — a larger swing than the gap between a healthy and an alarming rate. Use a three-month rolling figure instead.
Track them separately. Failed and expired cards are commonly a substantial share of cancellations and they are a billing problem rather than a retention one, with a much cheaper fix. Blending them hides both.
Bring your rate and where cancellations cluster. Marcus names the type and the single fix worth making this quarter.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime