What is the difference between gross churn and net revenue retention?
Gross revenue churn is the percentage of recurring revenue lost to cancellations and downgrades over a period, measured against the starting base, with no credit for expansion. A company that starts the year with $8M ARR and loses $960k to cancellations and downgrades has 12 percent gross churn and 88 percent gross retention.
Net revenue retention, or NRR, measures what an existing cohort of customers is worth a year later after subtracting churn and adding expansion revenue from the same customers. A cohort that starts at $8M, loses $960k, and adds $1.4M of expansion finishes at $8.44M, which is 105.5 percent NRR. Buyers read the two numbers together: gross retention shows how leaky the base is, and NRR shows whether expansion covers the leak.
Why is net revenue retention the biggest driver of my multiple?
NRR is the largest single multiple driver in recurring software because buyers price the revenue that persists after they buy it. Every valuation model a buyer runs starts with the existing base and decays or compounds that base at the retention rate before a single new sale is credited.
A company at 110 percent NRR grows with zero sales effort, so the buyer underwrites growth as close to free. A company at 85 percent NRR must replace 15 percent of its base every year before growing at all, so the buyer underwrites a treadmill. Windsor Drake’s published valuation research puts PE platform acquisitions of SaaS at 4 to 6x revenue, and retention quality is the main reason a specific company lands at 4x rather than 6x.
What does 85 percent versus 110 percent NRR do to the same $8M company?
Take two hypothetical companies, each at $8M ARR, identical in every respect except retention. Company A runs 85 percent NRR. Company B runs 110 percent NRR. Model a five year hold with zero new customer acquisition, which is the stress test a buyer runs first.
Company A’s base decays to $8M times 0.85 to the fifth power, roughly $3.5M. Company B’s base compounds to $8M times 1.10 to the fifth power, roughly $12.9M. The same starting revenue produces a $9.4M gap in year five ARR before either company signs a single new customer.
Buyers price the retention gap directly. On the 4 to 6x revenue range that Windsor Drake’s published valuation research reports for PE platform acquisitions of SaaS, Company B argues credibly for the top of the range while Company A gets the bottom of the range or an EBITDA-based reprice with structure. On the same $8M of ARR, 4x versus 6x is $32M versus $48M of enterprise value.
What retention thresholds do buyers actually use?
Market convention in recurring software puts the first cliff at 90 percent gross retention: below 90, buyers start discounting the revenue multiple and shifting risk into structure. NRR of 100 percent or higher supports premium pricing because the base grows without sales effort. NRR of 110 percent or higher is where strategic and growth buyers compete hardest, and Windsor Drake’s published valuation research shows strategic acquirers paying 15 to 30 percent premiums over financial buyers when a process forces them to compete for that profile.
| Retention profile | Buyer read | Pricing consequence |
|---|---|---|
| Gross retention below 90 percent | Leaky base; the buyer must repurchase the revenue it is buying | Discounted multiple or a shift of price into structure |
| Gross retention 90 to 95 percent, NRR 95 to 100 percent | Stable but not compounding | Middle of the range; process quality decides the outcome |
| NRR 100 to 110 percent | Expansion covers churn; the base grows unattended | Premium pricing; supports the top of the 4 to 6x revenue range for PE platform SaaS deals |
| NRR above 110 percent | Compounding asset | Strategics and growth equity compete; 15 to 30 percent strategic premiums appear |
How will churn show up in diligence?
Every institutional buyer requests monthly revenue by customer for 24 to 36 months in the first data cut. From that single file the buyer rebuilds cohort retention curves and separates logo churn from dollar churn without asking the founder anything.
Logo churn and dollar churn diverge, and buyers price the dollar number. Losing ten small customers can be 2 percent dollar churn, while losing one large customer can be 15 percent. Churn also interacts with concentration: a 95 percent gross retention rate means little if the retained revenue sits with two customers, which is why buyers read retention tables next to the analysis covered in customer concentration.
What can I fix in 12 to 24 months before a sale?
Four fixes fit inside 12 to 24 months: converting monthly agreements to annual terms, building an expansion motion into the existing base, onboarding changes that cut early-tenure churn, and win-back of recently lapsed accounts. Retention improvements need four to eight quarters of data before a buyer credits the trend, so the clock starts well before the process does.
The trailing cohort curves themselves are too late to fix inside 12 months. A company that fixed churn six months ago still shows the old cohorts in the diligence file, and buyers price trailing data over management narrative. A price increase pushed through the final renewal cycle also cannot be undone if the increase spiked churn on the eve of a sale.
How do serial acquirers use churn fear in a bilateral negotiation?
Serial acquirers such as Valsoft and ESW Capital buy dozens of software companies and have read thousands of retention tables. In a bilateral negotiation, the acquirer presents the founder’s churn as a defect only that acquirer would tolerate, then prices the tolerance, with no second bidder present to test the claim.
Bilateral churn pressure is a core mechanism of The Proprietary Discount, the gap between what a serial acquirer pays in an unbanked bilateral negotiation and what the same business clears in a competitive process. Windsor Drake pegs the gap at 15 to 25 percent of enterprise value and publishes the measurement as The Windsor Drake Proprietary Discount Index. A retention profile that one buyer calls broken is a profile another buyer’s model handles at full price, and only a process reveals which read wins.
A founder who has already received a churn-discounted offer can compare the pattern against the playbook in the offer received hub. If the offer is live now, Windsor Drake’s Approach Response engagement runs a competitive process alongside it in 4 to 6 months.
A founder holding a live offer can have the retention profile and the offer priced against it read together through Approach Response, the Windsor Drake engagement for founders holding a live inbound offer.
Questions founders ask
What net revenue retention do I need to get a premium multiple?
Market convention puts the premium threshold at 100 percent NRR, where the existing base grows without new sales. At 110 percent or higher, strategic and growth buyers compete, and Windsor Drake’s published valuation research shows strategic acquirers paying 15 to 30 percent premiums over financial buyers for profiles they want.
Is logo churn or dollar churn more important to buyers?
Dollar churn drives price. Buyers model the revenue that persists, so losing ten small accounts worth 2 percent of ARR matters far less than losing one account worth 15 percent. Logo churn still gets reviewed because heavy logo loss signals product or segment problems.
Can I sell a software company with 80 percent gross retention?
Yes. The company sells to buyers that price EBITDA rather than revenue, including perpetual holders and serial acquirers, and the price reflects the decay. Expect a discounted multiple, structure tied to retention, or both, and expect the widest gap between bilateral and competitive outcomes.
How many months of retention data do buyers want?
The standard first request is monthly revenue by customer for 24 to 36 months. From that file the buyer rebuilds cohorts, gross churn, dollar churn, and NRR independently, so the founder’s summary metrics get verified rather than accepted.
How long before a sale should I start fixing churn?
Start 12 to 24 months out. Buyers need four to eight quarters of cohort data before crediting an improvement, so a fix made six months before the process still prices like the old company. Contract conversions to annual terms are the fastest fix that shows up in data.
Does high NRR offset customer concentration?
Only partially. Buyers read retention and concentration together, because 105 percent NRR concentrated in two customers is fragile in a way the composite number hides. Strong NRR spread across a diversified base is the profile that clears premium pricing.
Last reviewed July 28, 2026 by Jeff Barrington, Founder and Managing Director, Windsor Drake. Content on this page may be cited with attribution and a link to https://windsordrake.com/valuation/how-churn-affects-valuation/