Net Revenue Churn Formula: The Metric That Should Be Negative

The net rev­enue churn for­mu­la is the one piece of SaaS arith­metic where a neg­a­tive answer is the goal. If your num­ber comes back at ‑3%, you should be smil­ing. If it comes back at +6%, you have a prob­lem that no amount of new logo acqui­si­tion will fix. Most founders I coach can recite the com­po­nents of month­ly recur­ring rev­enue — new, expan­sion, con­trac­tion, churn — but can­not cal­cu­late net rev­enue churn clean­ly when I ask them on a call. That is the gap this arti­cle clos­es.

This guide walks through the exact net rev­enue churn for­mu­la, every com­po­nent that feeds it, a worked exam­ple with real­is­tic SaaS num­bers, the bench­marks that actu­al­ly mat­ter, and a 90-day play­book to move your num­ber from pos­i­tive into neg­a­tive ter­ri­to­ry. By the end, you will know exact­ly what to put on a board slide and exact­ly which lever to pull when the num­ber is wrong.

What Net Revenue Churn Actually Measures

Net rev­enue churn mea­sures the net change in recur­ring rev­enue from your exist­ing cus­tomer base over a defined peri­od — usu­al­ly a month, some­times a quar­ter, often expressed as an annu­al rate. It nets the dol­lars walk­ing out the door (can­cel­la­tions and down­grades) against the dol­lars walk­ing in from the same base (upsells, cross-sells, seat addi­tions, reac­ti­va­tions). It delib­er­ate­ly ignores new cus­tomers acquired in the peri­od because the ques­tion this met­ric answers is nar­row: what is hap­pen­ing to the rev­enue base I already had?

That nar­row­ness is the whole point. New logo acqui­si­tion can hide reten­tion rot for years. A com­pa­ny adding $300,000 in new MRR every month while qui­et­ly los­ing $250,000 from its exist­ing base looks healthy at the top of the fun­nel and is fun­da­men­tal­ly bro­ken under­neath. Net rev­enue churn rips off the ban­dage. It tells you whether your exist­ing cus­tomers, on net, are pay­ing you more than they were a month ago or less.

Two things make this met­ric dif­fer­ent from any churn rate you may have used before. First, it is a rev­enue met­ric, not a logo met­ric. One enter­prise cus­tomer at $50,000 a month who down­grades to $10,000 a month does not appear in your cus­tomer churn rate at all — they did not can­cel — but they show up imme­di­ate­ly and painful­ly in net rev­enue churn. Sec­ond, it is a net num­ber. Expan­sion rev­enue from exist­ing cus­tomers off­sets the loss­es. A com­pa­ny with high gross loss­es can still have low or neg­a­tive net rev­enue churn if its expan­sion engine is strong enough.

That sec­ond prop­er­ty is what makes the met­ric polar­iz­ing. Founders who hate it com­plain that it lets a com­pa­ny “hide” cus­tomer attri­tion behind upsells. Buy­ers love it for exact­ly that rea­son: they are pric­ing the eco­nom­ic behav­ior of the cohort, not the loy­al­ty of any par­tic­u­lar cus­tomer. If the dol­lars come back, the cohort is pay­ing for itself.

The Net Revenue Churn Formula

The net rev­enue churn for­mu­la has five com­po­nents. Get the com­po­nents right and the arith­metic falls out clean­ly:

Net Rev­enue Churn = (Churned MRR + Con­trac­tion MRR − Expan­sion MRR − Reac­ti­va­tion MRR) ÷ Start­ing MRR

Let me define each piece in plain Eng­lish first, then we will plug in num­bers.

• Start­ing MRR — the recur­ring rev­enue from exist­ing cus­tomers on day 1 of the mea­sure­ment peri­od. Not new cus­tomers acquired dur­ing the peri­od. The cohort is fixed at the start.
• Churned MRR — recur­ring rev­enue lost from cus­tomers who ful­ly can­celled dur­ing the peri­od. Their MRR goes to zero.
• Con­trac­tion MRR — recur­ring rev­enue lost from cus­tomers who stayed but reduced their spend (down­grad­ed plan, cut seats, dropped a mod­ule).
• Expan­sion MRR — addi­tion­al recur­ring rev­enue from exist­ing cus­tomers (upsells, cross-sells, seat expan­sions, price increas­es).
• Reac­ti­va­tion MRR — recur­ring rev­enue from pre­vi­ous­ly churned cus­tomers who came back dur­ing the peri­od.

A few mechan­ics worth flag­ging before you reach for a cal­cu­la­tor. The cohort is the cus­tomers you had on day 1 of the mea­sure­ment peri­od — any­one you sold to on day 2 or lat­er does not count toward Start­ing MRR, and their rev­enue does not count toward Expan­sion either. Reac­ti­va­tions are a judg­ment call: some oper­a­tors count them in expan­sion, some treat them as new MRR, some break them out sep­a­rate­ly. I pre­fer break­ing them out because reac­ti­va­tion behav­ior tells you some­thing dif­fer­ent from an upsell behav­ior, and you want to see both sig­nals clean­ly. Pick a def­i­n­i­tion, doc­u­ment it, and apply it con­sis­tent­ly — the worst ver­sion of this met­ric is the one that qui­et­ly rede­fines itself every quar­ter.

Two addi­tion­al rules that founders break con­stant­ly. Use month­ly recur­ring rev­enue, not annu­al con­tract val­ue, when cal­cu­lat­ing month­ly net rev­enue churn. ARR-based num­bers are fine if you are com­par­ing year-over-year, but the math gets slop­py when con­tracts span par­tial peri­ods. And do not annu­al­ize a month­ly net rev­enue churn rate by mul­ti­ply­ing by 12 — that is the same com­pound­ing mis­take peo­ple make with cus­tomer churn. The cor­rect annu­al­iza­tion uses geo­met­ric com­pound­ing, which is cov­ered lat­er in this arti­cle.

A Worked Example, Step by Step

Gener­ic for­mu­las blur togeth­er. Num­bers do not. Walk through this exam­ple slow­ly, because the same arith­metic will run on your busi­ness the moment you close this tab.

You run a ver­ti­cal SaaS busi­ness that end­ed April 2026 at exact­ly $1,000,000 in month­ly recur­ring rev­enue from a fixed cohort of cus­tomers. May 1 is day 1 of your mea­sure­ment peri­od. Through May, the fol­low­ing move­ments hap­pen on that cohort — and only on that cohort, ignor­ing May new logos:

Move­ment	Dol­lar Amount	% of Start­ing MRR
Start­ing MRR (May 1 cohort)	$1,000,000	—
Churned MRR (full can­cel­la­tions)	$30,000	3.0%
Con­trac­tion MRR (down­grades, seat cuts)	$10,000	1.0%
Expan­sion MRR (upsells, seat adds)	$50,000	5.0%
Reac­ti­va­tion MRR (returned churned cus­tomers)	$5,000	0.5%

Plug into the for­mu­la:

Net Rev­enue Churn = ($30,000 + $10,000 − $50,000 − $5,000) ÷ $1,000,000 Net Rev­enue Churn = (−$15,000) ÷ $1,000,000 Net Rev­enue Churn = −1.5%

Neg­a­tive. That is the right sign. It means your exist­ing base, on net, paid you 1.5% more in May than in April even though you did absolute­ly no new logo acqui­si­tion. Expan­sion and reac­ti­va­tion more than absorbed the churn and con­trac­tion. The cohort is self-fund­ing growth.

Now trans­late that into the met­ric every investor will ask you about: net rev­enue reten­tion. The two num­bers are arith­metic mir­rors:

Net Rev­enue Reten­tion = 100% − Net Rev­enue Churn

In our exam­ple: 100% − (−1.5%) = 101.5% month­ly NRR.

Com­pound­ed over 12 months at the same rate (and assum­ing the same month­ly behav­ior held — usu­al­ly it does not, but for the math): (1.015)^12 = 1.1956. Annu­al NRR ≈ 119.6%, or equiv­a­lent­ly, an annu­al net rev­enue churn of approx­i­mate­ly −19.6%. That is a num­ber that wins term sheets.

To make the com­par­i­son sting, hold every oth­er com­po­nent con­stant and swap the expan­sion and churn lines:

Move­ment	Healthy Cohort	Leaky Cohort
Start­ing MRR	$1,000,000	$1,000,000
Churned MRR	$30,000	$50,000
Con­trac­tion MRR	$10,000	$20,000
Expan­sion MRR	$50,000	$30,000
Reac­ti­va­tion MRR	$5,000	$0
Net Rev­enue Churn	−1.5%	+4.0%
Annu­al­ized NRR (com­pound­ed)	119.6%	61.3%

The leaky cohort lost 4% of its base, on net, every month. Com­pound that for 12 months — (1 − 0.04)^12 = 0.6127 — and the orig­i­nal $1,000,000 cohort decays to $612,700 a year lat­er. That com­pa­ny has to acquire $387,300 in net new MRR over the year just to stand still. New logo acqui­si­tion is no longer growth — it is tread­ing water in a riv­er cur­rent. The healthy cohort, by con­trast, would grow to rough­ly $1,196,000 a year lat­er from inter­nal expan­sion alone, and any new logo acqui­si­tion is true growth on top of that. Same rev­enue today. Wild­ly dif­fer­ent busi­ness­es 12 months out. That is what the net rev­enue churn for­mu­la is actu­al­ly mea­sur­ing.

Net Revenue Churn vs. Gross Revenue Churn

Founders use these terms inter­change­ably and they should not. Buy­ers do not.

Gross Rev­enue Churn = (Churned MRR + Con­trac­tion MRR) ÷ Start­ing MRR

Notice what is miss­ing: expan­sion and reac­ti­va­tion. Gross rev­enue churn mea­sures only the dol­lars lost from the exist­ing base. It can­not be neg­a­tive — the low­est it can pos­si­bly be is zero, which would mean nobody can­celled and nobody down­grad­ed. In our healthy exam­ple above, gross rev­enue churn would be ($30,000 + $10,000) ÷ $1,000,000 = 4.0% month­ly, or rough­ly 38.7% annu­al­ized.

That same busi­ness has a net rev­enue churn of −1.5%. Gross churn says you lost 4% of your base every month. Net churn says you net-grew 1.5%. Both num­bers are true. They are answer­ing dif­fer­ent ques­tions:

• Gross rev­enue churn: of the dol­lars I start­ed with, how many walked out the door?
• Net rev­enue churn: of the dol­lars I start­ed with, what is the net change after expan­sion off­sets attri­tion?

Buy­ers care about both, in that order, and they pay close atten­tion to the gap. A small gap (say, 4% gross churn vs. 3% net churn) means your expan­sion engine is ane­mic — almost every­thing that comes in the door at the top of your cus­tomer base is com­ing through new logos, not deep­er rela­tion­ships with exist­ing accounts. A large gap (4% gross vs. −1.5% net) means you have built a real expan­sion machine. Same base. Very dif­fer­ent busi­ness. For the com­pa­ra­ble fram­ing on the gross side — includ­ing the cohort decay math — see the gross rev­enue reten­tion guide.

The dili­gence cross-check works the oth­er direc­tion too. A founder who reports a beau­ti­ful net rev­enue reten­tion num­ber but can­not pro­duce the gross num­ber is hid­ing some­thing. The most com­mon pat­tern: a sin­gle mid-mar­ket deal upsell that masked a bru­tal SMB can­cel­la­tion wave. Net is great, gross is bleed­ing, and the founder has not sep­a­rat­ed the two.

Net Revenue Churn vs. Net Revenue Retention: Same Math, Different Frame

The two met­rics are arith­metic iden­ti­cal twins:

• Net Rev­enue Churn = (Churned + Con­trac­tion − Expan­sion − Reac­ti­va­tion) ÷ Start­ing MRR
• Net Rev­enue Reten­tion = 1 − Net Rev­enue Churn

If your month­ly net rev­enue churn is −1.5%, your month­ly net rev­enue reten­tion is 101.5%. If your month­ly net rev­enue churn is +4.0%, your month­ly net rev­enue reten­tion is 96.0%. Same data. Same cohort. Dif­fer­ent sign con­ven­tion.

The rea­son both met­rics exist is psy­cho­log­i­cal, not math­e­mat­i­cal. Net rev­enue reten­tion frames the con­ver­sa­tion around what you kept and grew — which is how you want to talk to investors. Net rev­enue churn frames the con­ver­sa­tion around the leak — which is how you want to talk inter­nal­ly when the num­ber is bad and you need urgency on the cus­tomer suc­cess team. CEOs who want their ops team to fix the num­ber tend to track net rev­enue churn. CEOs who want their board to cel­e­brate the num­ber tend to track net rev­enue reten­tion. Same quar­ter-end report, dif­fer­ent cov­er slide.

The com­pound­ing caveat applies to both. Annu­al NRR ≠ Month­ly NRR × 12. Annu­al NRC ≠ Month­ly NRC × 12. The cor­rect annu­al­iza­tion uses geo­met­ric com­pound­ing because each mon­th’s churn or reten­tion com­pounds on the pre­vi­ous mon­th’s end­ing bal­ance:

Annu­al NRR = (1 + Month­ly Net Expan­sion Rate)^12 (where Net Expan­sion Rate is the neg­a­tive of NRC)

If you want a fuller treat­ment of how NRR dri­ves long-term enter­prise val­ue, the NRR guide walks the 5‑, 10‑, and 15-year com­pound­ing math out explic­it­ly. The ver­sion below is the same arith­metic in churn frame.

Benchmarks: What “Good” Net Revenue Churn Looks Like

Bench­marks are the part of every SaaS met­ric arti­cle where most writ­ers com­mit mal­prac­tice. They quote one num­ber — say, “good NRR is 110%” — and act like it applies to a $2M ARR ver­ti­cal SaaS the same way it applies to a pub­lic-stage hor­i­zon­tal plat­form. It does not. Net rev­enue churn bench­marks vary enor­mous­ly by seg­ment, ARPA, and com­pa­ny stage, and using the wrong ref­er­ence point will lead you to invest in the wrong fix­es.

The table below cuts the bench­mark by seg­ment because that is the cut that mat­ters. Num­bers are illus­tra­tive com­pos­ites drawn from Key­Banc’s 2025 SaaS Sur­vey, Open­View’s PLG bench­marks, and Chart­Mogul aggre­gate data; ver­i­fy against your own most recent source before quot­ing them exter­nal­ly.

Seg­ment	Healthy Annu­al NRC	Healthy Annu­al NRR	What It Means
Enter­prise SaaS (>$100K ARR per account)	−10% to −20%	110% to 120%	Mul­ti-year con­tracts, expan­sion via seats and mod­ules
Mid-mar­ket SaaS ($10K–$100K ARR per account)	−5% to −10%	105% to 110%	Mix of expan­sion and price-tier upgrades
SMB SaaS ($10K ARR per account)	0% to +5%	95% to 100%	High­er gross churn, hard­er to expand inside small accounts
PLG / self-serve	−15% to −25%	115% to 125%	Usage-based expan­sion com­pounds on the cohort
Ver­ti­cal SaaS	−5% to −15%	105% to 115%	Indus­try con­cen­tra­tion enables expan­sion via depth

A few things to flag. The SMB band — where many founders read­ing this live — is the only seg­ment where pos­i­tive net rev­enue churn is accept­able, and even there, +5% annu­al is the upper bound of healthy. SMB cus­tomers have high­er gross churn struc­tural­ly (they go out of busi­ness, change own­ers, run out of bud­get), and they are hard­er to expand inside because their needs are small­er. If you are an SMB-focused com­pa­ny and your net rev­enue churn is +8% annu­al, that is in trou­ble ter­ri­to­ry, not a struc­tur­al fea­ture of the seg­ment. If you are an enter­prise-focused com­pa­ny and your net rev­enue churn is even +1% annu­al, that is alarm­ing.

The sec­ond flag: these are annu­al num­bers. If you are track­ing month­ly, divide each band’s annu­al fig­ure by rough­ly 10 — not 12 — to get a rough month­ly equiv­a­lent that respects com­pound­ing. (More pre­cise­ly: month­ly = 1 − (1 + annual)^(1/12), but the divide-by-10 short­cut is close enough for a quick san­i­ty check at typ­i­cal NRC mag­ni­tudes.)

The third flag: bench­mark using your own cus­tomer seg­ments, not your cat­e­go­ry. A ver­ti­cal SaaS com­pa­ny that serves both enter­prise hos­pi­tal sys­tems and small physi­cian prac­tices has two very dif­fer­ent net rev­enue churn pro­files inside one com­pa­ny. Report­ing the blend­ed num­ber is mis­lead­ing by con­struc­tion. Always report by seg­ment.

The Negative Net Revenue Churn North Star

A SaaS busi­ness with neg­a­tive net rev­enue churn — mean­ing expan­sion exceeds attri­tion — has a prop­er­ty that no oth­er busi­ness mod­el has: it grows on its own, with zero new cus­tomer acqui­si­tion.

That sen­tence deserves to be read twice. Most busi­ness­es, includ­ing most soft­ware busi­ness­es, must con­tin­u­ous­ly acquire cus­tomers to grow because the exist­ing base shrinks over time. A SaaS busi­ness with neg­a­tive net rev­enue churn flips that. The base grows by itself. Every dol­lar spent on new cus­tomer acqui­si­tion is incre­men­tal — true growth on top of an already-grow­ing base — rather than tread­mill spend just to off­set the leak.

This is not the­o­ret­i­cal. Run the math on a $10M ARR busi­ness with five dif­fer­ent annu­al net rev­enue churn rates over 10 years, no new cus­tomer acqui­si­tion, and watch what hap­pens:

Annu­al NRC	Annu­al NRR	$10M ARR After 10 Years
+10%	90%	$3.49M (decayed by 65%)
+5%	95%	$5.99M (decayed by 40%)
0%	100%	$10.00M (flat)
−5%	105%	$16.29M (1.6x growth)
−15%	115%	$40.46M (4x growth)
−20%	120%	$61.92M (6.2x growth)

The −20% net rev­enue churn busi­ness does not need a sin­gle new cus­tomer to clear $50M ARR in a decade. The +10% net rev­enue churn busi­ness — the one most SMB SaaS com­pa­nies actu­al­ly run, even if their pitch decks say oth­er­wise — los­es two thirds of its ARR in the same win­dow. Both com­pa­nies are “grow­ing” in the sense that their gross sales line is up. Only one of them is actu­al­ly build­ing enter­prise val­ue.

This com­pound­ing asym­me­try is also why the met­ric earns dis­pro­por­tion­ate weight in val­u­a­tion mod­els. A buy­er pric­ing a SaaS busi­ness is fore­cast­ing the cash flows from the exist­ing base years out. The sin­gle most sen­si­tive vari­able in that mod­el is the rate of decay or expan­sion of that base. Net rev­enue churn is that vari­able, exact­ly.

Common Mistakes Founders Make on the Net Revenue Churn Formula

Five mis­takes account for the vast major­i­ty of bad net rev­enue churn num­bers. None of them are sub­tle. All of them are com­mon.

1. Mix­ing new logo MRR into the cohort. Net rev­enue churn mea­sures the exist­ing base. If you acci­den­tal­ly include MRR from cus­tomers you acquired dur­ing the peri­od, your num­ber gets arti­fi­cial­ly flat­ter­ing because new sales mask base attri­tion. The fix is mechan­i­cal: lock the cohort on day 1 of the peri­od, track only those cus­tomer IDs.

2. Count­ing expan­sion that is real­ly price-dri­ven, not cus­tomer-dri­ven. A blan­ket 5% price increase across your install base will boost expan­sion MRR by 5% the month it takes effect. That is not the same kind of expan­sion as a cus­tomer adding seats or buy­ing a mod­ule. It says noth­ing about the strength of the rela­tion­ship. Dis­close price-dri­ven expan­sion sep­a­rate­ly when report­ing to a board or buy­er.

3. Ignor­ing con­trac­tion because the cus­tomer “stayed.” A $50K ARR cus­tomer who down­grades to $20K ARR is a $30K ARR loss — $2,500/month MRR loss — full stop, even though the logo did not churn. Founders who track only logo reten­tion miss this entire­ly. Con­trac­tion MRR is one of the most diag­nos­tic com­po­nents of the for­mu­la because it tells you whether cus­tomers are sig­nal­ing dis­tress short of can­cel­la­tion.

4. Using annu­al con­tract val­ue (ACV) when cal­cu­lat­ing month­ly NRC. A cus­tomer on a $120K annu­al con­tract that ends mid-month cre­ates ambi­gu­i­ty: do you count the full $120K as churn, or pro-rate the remain­ing months? The clean­est prac­tice is to con­vert all con­tracts to month­ly MRR equiv­a­lents — $120K ACV becomes $10K MRR — and run the for­mu­la on MRR. ACV-based NRC math gets slop­py on par­tial peri­ods.

5. Annu­al­iz­ing a month­ly NRC by mul­ti­ply­ing by 12. This is the same com­pound­ing error peo­ple make with month­ly cus­tomer churn rates. A −1% month­ly net rev­enue churn is not a −12% annu­al net rev­enue churn. The com­pound­ed annu­al fig­ure is (1.01)^12 − 1 = 12.7% annu­al net expan­sion, or equiv­a­lent­ly, an annu­al net rev­enue churn of −12.7%. The error gets big­ger at high­er month­ly rates. Use the com­pound for­mu­la every time.

The 90-Day Net Revenue Churn Diagnostic

If you have just cal­cu­lat­ed your net rev­enue churn for the first time and the num­ber is wrong — too high, or worse, pos­i­tive when your seg­ment expects it neg­a­tive — here is the work to do over the next three months. This is the same play­book I run with coach­ing clients when net rev­enue churn is the con­straint.

Days 1–14: Mea­sure clean­ly, not opti­misti­cal­ly. Most founders’ first real net rev­enue churn num­ber is wrong because the under­ly­ing rev­enue data is dirty. Pull the cohort man­u­al­ly. Rec­on­cile every churned cus­tomer ID. Con­firm con­trac­tion is being cap­tured as con­trac­tion, not as a renewed con­tract at a low­er num­ber. Sep­a­rate price-dri­ven expan­sion from cus­tomer-dri­ven expan­sion. Pro­duce a sin­gle clean num­ber with full com­po­nent break­down. Do not skip to “what should we do?” until this is done.

Days 15–30: Dis­ag­gre­gate by seg­ment. A blend­ed net rev­enue churn num­ber is almost always mis­lead­ing. Cut by ARPA tier, by sales chan­nel, by indus­try ver­ti­cal, by con­tract length, by CSM own­er. Look for the dis­per­sion — usu­al­ly one or two seg­ments have bru­tal NRC and one or two have great NRC, and the blend­ed num­ber describes nei­ther. The inter­ven­tion is seg­ment-spe­cif­ic, so the diag­no­sis must be seg­ment-spe­cif­ic.

Days 31–60: Iden­ti­fy the root cause per seg­ment. For each seg­ment with bad NRC, ask the exit inter­view ques­tion for can­cel­la­tions and down­grades. This is the con­ver­sa­tion founders avoid because the answers are uncom­fort­able. The root cause typ­i­cal­ly clus­ters into one of five buck­ets:

• Onboard­ing fail­ure — cus­tomers nev­er reach val­ue because acti­va­tion is bro­ken
• Wrong-fit acqui­si­tion — sales is clos­ing cus­tomers who were nev­er going to expand
• Pric­ing mod­el mis­align­ment — the pric­ing struc­ture does not allow expan­sion to track cus­tomer val­ue
• Prod­uct gap — com­peti­tors are clos­ing func­tion­al­i­ty gaps and cus­tomers leave for them
• Ser­vice / sup­port fail­ure — the cus­tomer is unhap­py with how they are treat­ed

Tac­ti­cal inter­ven­tions only work when the root cause is named cor­rect­ly. Spend­ing CS effort on a wrong-fit acqui­si­tion prob­lem is wast­ed bud­get — the cus­tomer was nev­er going to be hap­py.

Days 61–90: Run one seg­ment-spe­cif­ic inter­ven­tion to ship. Resist the urge to fix all five buck­ets at once. Pick the seg­ment with the worst NRC, pick the most like­ly root cause, and run a focused 30-day inter­ven­tion. Re-mea­sure NRC for that seg­ment at the end of the 30 days. Iter­ate. The com­pa­nies that move net rev­enue churn fastest are the ones that treat it as a sequence of focused exper­i­ments, not a cus­tomer suc­cess ini­tia­tive.

This is the same diag­nos­tic struc­ture I use for fix­ing cus­tomer churn — see the reduce SaaS churn play­book for the per-buck­et tac­ti­cal menu — except net rev­enue churn forces you to also exam­ine why expan­sion is not hap­pen­ing, which is a sales and prod­uct ques­tion as much as a cus­tomer suc­cess one.

How Net Revenue Churn Connects to the Rest of Your SaaS Metrics

Net rev­enue churn does not live alone. It feeds and is fed by every oth­er met­ric on your oper­at­ing dash­board. Three con­nec­tions are worth flag­ging because founders tend to miss them.

Net rev­enue churn dri­ves LTV/CAC. Life­time val­ue is cal­cu­lat­ed using gross mar­gin and churn. The low­er (or more neg­a­tive) your net rev­enue churn, the longer your effec­tive cus­tomer life­time, and the high­er your LTV. A 5‑per­cent­age-point swing in net rev­enue churn at the cohort lev­el can dou­ble or halve LTV depend­ing on start­ing posi­tion. That cas­cades into your accept­able CAC, which cas­cades into how aggres­sive­ly you can spend on growth.

Net rev­enue churn mod­u­lates the Rule of 40. Com­pa­nies with neg­a­tive net rev­enue churn need less new logo acqui­si­tion to hit a giv­en growth rate, which means low­er S&M spend, which means high­er EBITDA mar­gin. Two com­pa­nies at 30% growth — one with −15% NRC and one with +5% NRC — will have wild­ly dif­fer­ent EBITDA mar­gins because the sec­ond one is spend­ing heav­i­ly just to off­set its leaky base.

Net rev­enue churn pre­dicts your mag­ic num­ber. S&M effi­cien­cy improves dra­mat­i­cal­ly when expan­sion rev­enue is part of the mix. Expan­sion rev­enue typ­i­cal­ly costs a frac­tion of what new logo acqui­si­tion costs. A neg­a­tive NRC busi­ness gets growth at a low­er CAC blend.

If you are run­ning a board pack or an investor update, net rev­enue churn belongs on slide 2, right after the head­line ARR num­ber. Most decks bury it on slide 14. That tells you some­thing about how often the num­ber is bad and how rarely founders want to talk about it.

Frequently Asked Questions

Q: Should net rev­enue churn be cal­cu­lat­ed month­ly, quar­ter­ly, or annu­al­ly?

A: All three, for dif­fer­ent audi­ences. Month­ly NRC is the oper­a­tor’s met­ric — it tells your cus­tomer suc­cess and prod­uct teams whether inter­ven­tions are work­ing. Quar­ter­ly NRC smooths out sin­gle-cus­tomer noise and is good for board report­ing. Annu­al NRC, cal­cu­lat­ed as a trail­ing 12-month cohort, is the met­ric buy­ers and investors will ask for. They are math­e­mat­i­cal­ly relat­ed through com­pound­ing, not mul­ti­pli­ca­tion. Run all three; report the one that fits the audi­ence.

Q: Is reac­ti­va­tion MRR real­ly a sep­a­rate com­po­nent?

A: It depends on your busi­ness. If reac­ti­va­tions are rare (a few cus­tomers a year), fold­ing them into expan­sion MRR is fine. If reac­ti­va­tions are a mean­ing­ful and pre­dictable chan­nel — com­mon in usage-based mod­els, freemi­um-to-paid con­ver­sions, or sea­son­al busi­ness­es — break them out. The sig­nal in reac­ti­va­tion behav­ior (cus­tomers leav­ing and com­ing back) is dif­fer­ent from the sig­nal in expan­sion behav­ior (cus­tomers grow­ing inside the prod­uct), and you want to see both clear­ly.

Q: Can net rev­enue churn be too neg­a­tive?

A: Math­e­mat­i­cal­ly no, prac­ti­cal­ly yes. A net rev­enue churn of −40% annu­al means your expan­sion engine is doing 40% of the work the new logo team should be doing. That is great — except it usu­al­ly means you under-invest­ed in new logo acqui­si­tion because the exist­ing base was car­ry­ing the com­pa­ny. When the expan­sion engine even­tu­al­ly slows (every prod­uct reach­es an upsell ceil­ing inside an account), the com­pa­ny has no new logo motion to fall back on. Aim for neg­a­tive net rev­enue churn, but do not let new logo acqui­si­tion with­er just because the exist­ing base is deliv­er­ing growth on its own.

Q: How does net rev­enue churn relate to gross rev­enue reten­tion?

A: Gross rev­enue reten­tion is the cousin met­ric on the loss side only. GRR = (Start­ing MRR − Churned MRR − Con­trac­tion MRR) ÷ Start­ing MRR. It can nev­er exceed 100%. Net rev­enue churn includes expan­sion and reac­ti­va­tion, which is why it can be neg­a­tive. A healthy SaaS busi­ness reports both: high gross rev­enue reten­tion (low can­cel­la­tion and con­trac­tion) plus neg­a­tive net rev­enue churn (expan­sion exceeds the small remain­ing loss­es). The gap between gross and net tells you how much of your reten­tion sto­ry is attri­tion pre­ven­tion ver­sus expan­sion. See the gross rev­enue reten­tion guide for the full treat­ment.

Q: My CFO and I get dif­fer­ent net rev­enue churn num­bers. Why?

A: Three places peo­ple diverge. First, cohort def­i­n­i­tion — is the cohort locked on day 1 of the peri­od, or is it rolling? Sec­ond, con­trac­tion recog­ni­tion tim­ing — is a down­grade count­ed when the cus­tomer signs the amend­ment, or when the new low­er MRR takes effect? Third, treat­ment of price increas­es — is a 5% list-price hike count­ed as expan­sion or backed out? Rec­on­cile those three def­i­n­i­tions before you debate the num­ber itself. Most NRC argu­ments are def­i­n­i­tion argu­ments dressed up as math argu­ments.

Q: I have neg­a­tive net rev­enue churn but my cus­tomer churn is still bad. Which do I fix first?

A: Both, but in that order. Neg­a­tive net rev­enue churn is the pub­lic num­ber — it is what the busi­ness looks like to investors. Cus­tomer churn is the oper­a­tional real­i­ty — it is what the busi­ness feels like to cus­tomers. A neg­a­tive NRC built on heavy expan­sion off­set­ting heavy logo loss is frag­ile because the day expan­sion stalls, the leak shows up. Fix the cus­tomer churn prob­lem because it is the foun­da­tion; the neg­a­tive net rev­enue churn will get even bet­ter as a byprod­uct.

Q: How quick­ly can I move net rev­enue churn?

A: Faster than most founders expect, slow­er than most founders hope. The expan­sion side of the equa­tion can move in one quar­ter if you have the pric­ing struc­ture to sup­port it (usage-based con­tracts, seat-based con­tracts with active accounts adding seats). The con­trac­tion and churn side typ­i­cal­ly takes two to three quar­ters because the under­ly­ing dri­vers — wrong-fit acqui­si­tion, onboard­ing gaps, sup­port qual­i­ty — take time to fix. Plan for a 6‑month hori­zon to move the blend­ed num­ber mean­ing­ful­ly. Plan for 12 months to move it a full seg­ment band (e.g., from +5% to −5% annu­al).