Churn Meaning: The Straight Answer for Solo SaaS

Churn means the share of customers or recurring revenue lost during a defined period. For a solo SaaS, calculate it from your own billing data; do not borrow an enterprise benchmark. The ProvenStartups cohort does not disclose enough loss data to support a real solo-SaaS churn range, so publishing one would mean inventing it.

Contents

This page gives the definition, calculation rules, available founder evidence, the dataset’s contradiction, and a practical measurement process. Use the links below to jump directly to the part you need; the core distinction is simple: revenue evidence can establish that a business earns money, but it cannot establish retention without loss data.

• ·What churn means
• ·How to calculate churn correctly
• ·What the ProvenStartups evidence shows
• ·Where our data contradicts the popular claim
• ·What a solo founder should do
• ·FAQ

What churn means

In plain English, churn rate means how much of a recurring customer or revenue base disappeared during one measurement window. Customer churn counts lost accounts; revenue churn counts lost recurring dollars. The denominator must be the base present at the start, not every customer who appeared at any point during the period.

The basic customer formula is:

customer churn = customers lost during period / customers at start of period

Gross revenue churn measures recurring revenue lost to cancellations and downgrades against starting recurring revenue. Net revenue churn also credits expansion from retained customers, so strong upgrades can offset losses. State which version you use whenever someone asks, “What is the churn rate?”

Both Stripe’s guide to churn measurement and Investopedia’s definition of churn rate treat churn as loss over a period. The definition is easy. The hard part is keeping the numerator, denominator, billing interval, pauses, reactivations, and downgrades consistent.

How to calculate churn correctly

Calculate churn from a frozen opening cohort and an explicit event log. Count which opening customers cancel before the period closes, then divide those losses by the opening count. For revenue churn, run the same operation on recurring revenue. New customers belong in growth reporting, not in the churn denominator for that window.

Use these inputs:

• ·opening_customer_ids
• ·cancelled_opening_customer_ids
• ·opening_recurring_revenue
• ·lost_or_downgraded_recurring_revenue
• ·expansion_revenue, only when reporting net revenue churn

Data Fetcher reports $23K/mo [F], 600 paying customers [F], and an 85% margin [F]. Those figures still cannot produce churn. We would need the opening customer set and the customers lost during the same window. Dividing cancellations by the current 600 paying customers [F] would mix stocks and flows.

Define cancellation behavior before querying. Decide how failed payments, refunds, paused plans, annual contracts, and reactivations map to states. Then preserve that rule so movement in the metric reflects customer behavior rather than a changed query.

What the ProvenStartups evidence shows

ProvenStartups can show a wide revenue distribution among relevant products, but not a defensible churn distribution. The full matching cohort contains 229 projects, including 138 solo-run projects. Among 86 projects with a clean monthly figure, the median is $30K/mo and the range is $6/mo to $2.2M/mo. These are cohort aggregates, not one case claim.

The cohort spans SaaS, consumer apps, AI services, simple tools, plugins, directories, and ecosystem products. Its recorded evidence split is 34 [V], 0 [F], 0 [C], and 0 [U]. That split does not turn revenue into churn; the missing variables are still opening base, losses, period, and cohort definition.

That table is the useful boundary. It gives checkable commercial context while refusing to label success as retention. Browse the full startup index for the cases, and read the grading method before treating [V], [F], [C], and [U] as interchangeable.

Where our data contradicts the popular claim

The popular claim that solo SaaS businesses fit one “healthy” churn range is not supported by this dataset. The matching cohort contains substantial revenue evidence but no customer-loss series sufficient to calculate that range. ProvenStartups therefore refuses to print an enterprise benchmark as if it described products run by one person.

This matters because revenue screenshots create false confidence. Letterly’s $250K/mo [C] does not reveal how many subscribers left. Revid’s $600K+/mo [F] does not reveal whether growth masks heavy cancellation. The AEO Service’s $2,000/mo retainer [F] comes from a single client [F], where losing an account has a completely different shape from churn across hundreds of low-priced subscriptions.

Our data also contradicts the shortcut that scale proves retention quality. Revenue, downloads, profit, and margin answer commercial questions; churn answers a cohort-loss question. Unless a case exposes compatible opening and loss figures, the honest churn value is “not disclosed.”

What a solo founder should do

Track the churn rate your own product can calculate consistently, segmented by plan and acquisition source. Start with customer churn and gross revenue churn; add net revenue churn when expansion matters. Refuse comparisons that mix monthly self-serve users with annual contracts, services, or consumer downloads, because their cancellation opportunities and denominators are structurally different.

1. 1.Freeze the opening customer IDs and opening recurring revenue.
2. 2.Record cancellations, downgrades, reactivations, pauses, and failed-payment recovery as separate events.
3. 3.Calculate customer churn and gross revenue churn from the same time boundary.
4. 4.Segment by plan, billing interval, and acquisition source before chasing a blended headline.
5. 5.Save the query and definition beside each result so future measurements remain comparable.

Do not “improve” churn by changing the denominator or silently removing failed payments. If the product is young and the sample is sparse, report the raw inputs next to the rate. A transparent small cohort is more actionable than a polished benchmark imported from a different business model.

FAQ

Churn questions usually become ambiguous because “churn” can refer to customers, gross recurring revenue, or net recurring revenue. The answers below use the strict loss-over-opening-base definition. When the necessary inputs are absent, the correct answer is not an estimate; it is that churn was not disclosed or cannot be calculated.

What is churn rate in business?

Churn rate is the portion of an opening customer or recurring-revenue base lost during a specified period. For subscriptions, customer churn tracks cancelled accounts, while revenue churn tracks cancelled and downgraded recurring revenue. Always name the period and metric. Without both, “churn rate” is too vague to compare or debug.

What is customer churn rate?

Customer churn rate is lost opening customers divided by customers present at the start of the period. It should not use the closing customer count or include newly acquired customers in the denominator. Keep account cancellation separate from contraction, because a retained customer who downgrades affects revenue churn but not customer churn.

What is a churn rate worth comparing?

A churn rate is worth comparing only when the products use compatible definitions, periods, billing intervals, and customer types. A self-serve monthly app is not directly comparable with an annual SaaS contract or a service retainer. Internal cohort trends are usually more diagnostic than a broad external benchmark with hidden methodology.

Can revenue prove that churn is low?

No. StoryShort.ai reports $35K/mo across 3 apps [F], but that claim contains no opening base or customer losses. Revenue can rise while churn remains high if acquisition is faster than cancellation. To prove low churn, publish the period, opening cohort, losses, and treatment of downgrades and reactivations.

What churn range should a solo SaaS target?

No defensible universal solo-SaaS range is available from this cohort. The supplied cases disclose commercial outcomes, not the consistent loss series needed to calculate a distribution. Target improvement against your own stable definition, then compare like-for-like cohorts. ProvenStartups would refuse to fabricate a percentage simply because searchers expect one.