Net Revenue Retention (NRR): Formula and Benchmarks

Net revenue retention (NRR) answers one question. "If we stopped acquiring new subscribers tomorrow, would revenue from the people we already have grow or shrink?" MRR can't tell you that. Churn rate can't either. NRR can.

At Codecademy, while we grew from $10M to $55M ARR (2017-2021), very little of the growth came from more traffic. Most of it came from making existing subscribers worth more: better pricing, smarter packaging and plan mix, lower churn, and expansion. All of that shows up in one number.

Most consumer subscription teams I talk to don't track NRR. They treat it as a B2B SaaS investor metric. It isn't. This guide covers the formula, how NRR relates to GRR and NDR, what "good" looks like (and why consumer and membership businesses shouldn't copy SaaS headline numbers), how to pull it from your own data, how to present it to leadership, and the levers that move it. There's a free NRR calculator too.

---

NRR at a Glance

---

How to Calculate NRR (and NDR)

The Formula

NRR = (Starting MRR − Churned MRR − Contraction MRR + Expansion MRR) ÷ Starting MRR × 100

All four terms come from the same cohort: the subscribers who were paying at the start of the period. New subscribers acquired during the period are excluded entirely.

• Starting MRR: recurring revenue from that cohort at the start of the period
• Churned MRR: revenue lost from subscribers in the cohort who cancelled, voluntarily or because a payment failed
• Contraction MRR: revenue lost from subscribers who downgraded or moved to a cheaper plan
• Expansion MRR: revenue gained from upgrades, add-ons, extra seats or usage, and price increases on existing subscribers

Use the same currency and the same definition of recurring revenue in every term. Exclude one-off charges, and decide once whether annual plans are normalized to monthly.

Worked Example (Hypothetical)

A consumer subscription business starts the month with $500,000 MRR from existing subscribers. During the month:

• $30,000 is lost to cancellations (churn)
• $5,000 is lost to downgrades from Premium to Standard (contraction)
• $12,000 is gained from Standard → Premium upgrades (expansion)
• $15,000 is gained from a price increase rolling through a legacy cohort (expansion)

NRR = ($500,000 − $30,000 − $5,000 + $12,000 + $15,000) ÷ $500,000 × 100 = 98.4%
GRR = ($500,000 − $30,000 − $5,000) ÷ $500,000 × 100 = 93.0%

That's a typical consumer picture: a meaningful leak (GRR 93%) that expansion almost, but not quite, fills. The roadmap question is obvious. Is it cheaper to cut that $35,000 of monthly losses, or to grow the $27,000 of expansion? To get this business to 100% NRR, it needs $35,000 of expansion a month, which is $8,000 more than it has, or $8,000 less churn and contraction. The NRR calculator works out that "expansion needed to hit target" line for your own numbers.

NRR Formulas You'll Need

Monthly NRR      = (Starting MRR − Churn − Contraction + Expansion) ÷ Starting MRR × 100
GRR              = (Starting MRR − Churn − Contraction) ÷ Starting MRR × 100
Expansion rate   = Expansion MRR ÷ Starting MRR × 100
Annualized NRR  ≈ Monthly NRR ^ 12        e.g. 1.02^12 ≈ 1.268 → ~126.8%

The annualized figure is an approximation. It assumes this month's rate repeats on a compounding base. For leadership reporting, prefer a true 12-month cohort NRR: the cohort's MRR today ÷ the same cohort's MRR 12 months ago.

How to Pull Cohort NRR From Your Own Data

If your billing data lands in a warehouse, this query gives you true 12-month cohort NRR and GRR, plus the decomposition. It assumes a table with one row per customer per month and their normalized recurring revenue. Adjust the table name, the dates and the date syntax for your warehouse.

-- mrr_monthly: customer_id, month (first day of month), mrr (normalized, excl. one-off charges)
WITH start_cohort AS (
  SELECT customer_id, mrr AS start_mrr
  FROM mrr_monthly
  WHERE month = DATE '2025-10-01' AND mrr > 0
),
cohort_later AS (
  SELECT s.customer_id, s.start_mrr, COALESCE(m.mrr, 0) AS end_mrr
  FROM start_cohort s
  LEFT JOIN mrr_monthly m
    ON m.customer_id = s.customer_id
   AND m.month = DATE '2026-10-01'      -- 12 months later
)
SELECT
  SUM(start_mrr)                                                       AS starting_mrr,
  SUM(CASE WHEN end_mrr = 0 THEN start_mrr ELSE 0 END)                 AS churned_mrr,
  SUM(CASE WHEN end_mrr > 0 AND end_mrr < start_mrr
           THEN start_mrr - end_mrr ELSE 0 END)                        AS contraction_mrr,
  SUM(CASE WHEN end_mrr > start_mrr THEN end_mrr - start_mrr ELSE 0 END) AS expansion_mrr,
  ROUND(100.0 * SUM(end_mrr) / SUM(start_mrr), 1)                      AS nrr_pct,
  ROUND(100.0 * SUM(LEAST(end_mrr, start_mrr)) / SUM(start_mrr), 1)   AS grr_pct
FROM cohort_later;

Two choices to make once and write down. First, this compares endpoints, so a subscriber who cancelled and came back within the year counts as retained. If you'd rather count them as churn plus new, filter on continuous payment. Second, new subscribers who joined after the start month never enter the cohort. That's the point.

To see how retention and churn shape your long-run revenue, use the Growth Ceiling Calculator. To see how plan mix changes revenue per subscriber, try the Annual Plan Mix Calculator.

NRR vs GRR

Gross revenue retention (GRR) counts losses only: churn and contraction, no expansion. It can't exceed 100%. NRR adds expansion back, so it can go above 100%.

You can have a GRR of 92% and an NRR of 108%. That means you lose 8% of starting revenue to churn and downgrades, and expansion more than makes it up. Both numbers matter. GRR shows the health of the core product (do people stay?). NRR shows the health of the model (does each subscriber become more valuable?). ChartMogul's definitions match this split.

NRR vs NDR

They're the same metric. "Net dollar retention" and "net revenue retention" differ in wording, not formula. Public companies use several labels: "net revenue retention rate," "dollar-based net retention," "dollar-based net expansion." The underlying idea is the same. Cohort and time-window definitions vary, so read the footnote before you compare two companies.

---

What "Good" NRR Looks Like, and Why Consumer and Membership Differ

There's no single "good NRR" that applies to every subscription business. Most NRR figures you'll see quoted come from public B2B software companies, where expansion is built in: seats grow with headcount, usage grows with data, and products cross-sell into the same account. Consumer benchmarks are scarcer, and they look very different.

A Consumer Benchmark

One of the few large consumer datasets comes from billing data rather than surveys. ChartMogul's retention report (Kyle Poyar, Dec 10, 2025) looked at annualized 2025 retention for software companies with at least $250K ARR:

The same report found median NRR 10 to 20 percentage points higher on annual plans than on monthly ones. Two caveats: this is consumer software, not media or membership, and ChartMogul's sample includes many smaller self-serve products. Use it to calibrate expectations. If your consumer app is near 49%, you're at the median, not an outlier. Then manage your own trend.

Public Software Companies, for Contrast

Here are recent reported figures from public companies, so you can see what built-in expansion looks like.

Definitions vary by company. Compare with that caveat.

The takeaway isn't "aim for 120%." It's that NRR is a structural property of how you charge. Consumption and multi-seat products expand automatically. A consumer subscription with one or two plans mostly doesn't.

Why Consumer and Membership NRR Sits Differently

A meditation app, a news subscription or a professional membership has no seats to add and no usage meter. For these businesses:

• GRR does most of the work. Churn, including involuntary churn from failed payments, is usually the biggest term in the formula. Even within software, ChartMogul reports a gap by price. Top-quartile companies with ARPA under $50/month reach only about 60–70% annual gross retention. Companies with ARPA above $500/month reach about 88%. Low price points leak more.
• Expansion comes from a different set of levers:
  • Plan mix: moving subscribers to higher tiers (Standard → Premium, individual → family or household)
  • Price increases on existing subscribers, including ending legacy or promo pricing
  • Add-ons and bundles: extra products, print plus digital, a partner bundle
  • Billing changes that raise realized revenue, such as converting monthly to annual. This affects retention too, so model it rather than assuming.
• Contraction shows up as downgrades to cheaper tiers, retention discounts, and annual-to-monthly switches.

For consumer, membership, and media businesses, the useful question isn't "are we at 120%?" Ask these instead:

1. Is NRR trending up or down, on a consistent definition?
2. Which term moved, churn, contraction or expansion?
3. How does it vary by segment (plan, acquisition channel, tenure cohort, billing interval)?

Can a Consumer Subscription Have NRR Above 100%?

Yes, if expansion (upgrades, plan mix, add-ons, price increases) is bigger than churn plus downgrades. It's less common than in seat-based software, but it's the right thing to manage toward. A consumer business at 95% that knows exactly why is in better shape than one at 102% that can't explain the number.

---

How to Show NRR to Leadership

One number that moved, with nobody able to say why, is how NRR loses leadership's trust. Here's the one-slide version I recommend, filled in with the worked example from above. The prior-period columns are hypothetical, there to show how the comparison reads.

Under the table, add three bullets: what changed, why, and what you're doing about it. In this example, NRR jumped mostly because of a one-time price increase, GRR was flat, and involuntary churn is improving year on year. That's a very different story from "NRR is up 3 points." The NRR calculator has this slide as a tab that fills from your inputs.

A few rules that save painful leadership conversations:

• State the window and stick with it. Monthly, trailing 12 months or annual cohort. Don't switch mid-discussion.
• Split involuntary churn out. It's usually the cheapest line to improve. See failed payment recovery.
• Show price increases as their own expansion line. A price-led NRR jump and a product-led one tell very different stories about durability.
• Segment once a quarter. Plan, channel and tenure cohort. A blended number can hide a great segment and a terrible one.

What NRR doesn't tell leadership: new-subscriber growth, acquisition efficiency (LTV:CAC and CAC payback) or cash. Put those on the next slide, not in the NRR footnotes.

---

The Three Levers, and the Playbook

NRR is the net result of three forces. Improving it means pulling one or more of them.

Lever 1: Reduce Churn

Every cancelled subscriber comes straight off NRR. Start with the cheapest fix: involuntary churn. These subscribers wanted to keep paying. Recurly estimates involuntary churn at 20–40% of total churn across subscription businesses.

• Turn on card account updater and smart retries (failed payment recovery)
• Run a real dunning process
• Set grace periods on App Store and Google Play subscriptions

For voluntary churn, the move that pays off most is a cancellation flow. It asks why people are leaving and offers a matching alternative: pause, downgrade, or a time-limited discount. For the full picture, see the churn rate guide and customer retention strategies.

Lever 2: Reduce Contraction

Some contraction is healthy. A downgrade beats a cancellation. Heavy downgrade volume, though, usually points to packaging:

• Tiers don't match how people use the product, so subscribers pay for features they don't use and eventually notice.
• The gap between tiers is too big. People who can't justify the top tier drop all the way to the bottom.
• Retention discounts never expire. A save offer that quietly becomes permanent shows up as contraction forever.

If most downgrades skip your middle tier, your middle tier isn't doing its job.

Lever 3: Increase Expansion

This is where NRR goes from "protecting revenue" to "growing revenue." For consumer and membership businesses, the main paths are:

• Upgrades and plan mix. Give subscribers a natural reason to move up (more content, family or household plans, premium features). Prompt at the moment of need, like hitting a limit or trying a gated feature, not in a random newsletter. More in the upselling guide.
• Add-ons and bundles. Things that feel like natural extensions, not nickel-and-diming.
• Price increases. For many consumer subscriptions this is the biggest expansion lever, and the riskiest for goodwill. Test it, grandfather it thoughtfully, communicate the value clearly, and watch churn in the affected cohort. The way I think about it: price is the conversion rate on the value you've delivered. A price increase lands much more easily right after you've made the product meaningfully better, because customers can see what they're paying more for. That said, almost every company does too little price testing.
• Annual plans. They mostly help retention rather than expansion, but they change the revenue mix. Model them with the Annual Plan Mix Calculator.

Seat and usage expansion matter for B2B products. If any part of your pricing scales with usage, expansion happens as customers succeed.

The Playbook

1. Know your number. Pick a cohort (everyone paying on January 1). Record its MRR, then the same cohort's MRR a month later, including cancellations, downgrades and expansion. Divide. Do it for the last 12 months, because you want the trend.

2. Decompose it into GRR and expansion rate. If GRR is 88%, you have a leak to fix before expansion matters much. If GRR is 96% and expansion is 1%, your problem is packaging and upgrade paths.

3. Fix the biggest lever first:

   • GRR is weak for your segment: churn first (involuntary, then voluntary). What counts as weak depends on who you sell to. For self-serve or low-priced B2B, about 90% annual GRR is the peer bar. SaaS Capital's 2025 benchmarks put median GRR at 90% for companies with contracts under $12K a year. For consumer subscriptions, 90% is the wrong bar. In ChartMogul's 2023 retention report, best-in-class consumer gross retention tops out around 75–80%. Only 5.3% of businesses charging under $10 a month clear 85%. Compare yourself to your own segment, and to your own trend.
   • GRR is solid but expansion is low: add a tier, an add-on or upgrade prompts.
   • Both are solid: revisit price.

4. Build expansion into the product, so the upgrade shows up at the moment of value.

5. Revisit pricing and packaging if NRR is structurally capped. A single flat plan with no expansion path won't get above 100% through retention tactics alone.

---

When NRR Is the Wrong Metric to Chase

NRR isn't the right scoreboard for everyone. Weight it lower if:

You sell a single flat-rate plan with no expansion path. Getting above 100% would mean adding tiers, add-ons or price increases that may not fit the product. Make GRR your main scoreboard and track NRR as context.

Your cohort data is short or noisy. With less than a year of history, or a small subscriber base, one large account or one promo cohort can swing the number. Track retention curves until churn stabilizes.

Your revenue is mostly services or transactions. NRR measures recurring revenue. If a big share of revenue doesn't recur, track recurring and non-recurring separately, or use GMV and take rate.

---

FAQ

What is net revenue retention?

Net revenue retention (NRR) measures whether existing subscribers' recurring revenue grew or shrank over a period, after churn, downgrades and expansion. It excludes new subscribers. Above 100% means the existing base is a net growth engine. Below 100% means it shrinks without new acquisition.

What is the NRR formula?

NRR = (Starting MRR − Churned MRR − Contraction MRR + Expansion MRR) ÷ Starting MRR × 100. All terms come from the same group of subscribers who were paying at the start of the period. Use the same currency and the same definition of recurring revenue throughout.

Is NRR the same as net dollar retention (NDR)?

Yes. NRR and NDR describe the same metric. "Dollar" versus "revenue" is wording, not a different formula. Companies define cohorts and time windows slightly differently, so check the definition before comparing two reported numbers.

What's the difference between NRR and GRR?

Gross revenue retention (GRR) counts churn and contraction but excludes expansion, so it can't go above 100%. NRR adds expansion back and can exceed 100%. Report both: GRR shows how leaky the base is, and NRR shows whether expansion more than fills the leak.

What is a good NRR?

There's no single good number for every subscription business. Public B2B software companies with seat or usage pricing report figures above 100%. In ChartMogul's 2025 data, median NRR was 82% for B2B SaaS and 49% for consumer SaaS companies with at least $250K ARR. Consumer and membership businesses rely much less on seat or usage expansion, so copying a SaaS headline is misleading. A more useful test: is NRR trending up, and do you know whether the change came from churn, contraction or expansion?

Can a consumer subscription business have NRR above 100%?

Yes, if upgrades, plan mix, add-ons and price increases add more revenue than churn and downgrades remove. It's less common than in seat-based software, but it's the right thing to manage toward rather than a metric to ignore.

How do I calculate how much expansion I need to hit a target NRR?

Expansion needed = (Target NRR × Starting MRR) − (Starting MRR − Churned MRR − Contraction MRR). In the worked example, hitting 100% on $500,000 starting MRR with $35,000 of churn and contraction needs $35,000 of expansion. The business has $27,000, so the gap is $8,000 a month. The NRR calculator does this for you.

How often should we measure NRR?

Calculate it monthly for operating decisions. Report a trailing 12-month or true 12-month cohort view to leadership, because single months are noisy. Say which window you're using every time.

Can NRR be above 100% if churn is high?

Yes, if expansion from the subscribers who stay outweighs what you lost. A business losing 5% of revenue to churn each month but gaining 8% from expansion has 103% NRR. That's a fragile position, so always show GRR next to NRR to keep the leak visible.

How does NRR relate to LTV?

Higher NRR means higher customer lifetime value. If subscribers spend more over time, simple LTV formulas based only on cancellation rate undercount what a cohort is worth. See the LTV:CAC guide.

---

What to Do Next

NRR tells you whether your pricing works, whether your packaging leaves room to grow, and whether you're making more from the subscribers you already have, or stuck on the acquisition treadmill.

1. Calculate monthly NRR and GRR for the last 12 months on one consistent definition.
2. Decompose the most recent quarter into churn (voluntary and involuntary), contraction and expansion.
3. Fix the biggest lever first. For most consumer businesses, that's involuntary churn, then the cancellation flow, then plan mix and price.
4. Put the one-slide decomposition in front of your leadership team every quarter.

More on the surrounding metrics: subscription metrics, churn rate, LTV:CAC and CAC payback.

Get new guides and teardowns by email. Browse the archive or subscribe below.

If your NRR is below where it needs to be, or you need a leadership-ready decomposition and a prioritized roadmap, I do a free 30-minute call. No pitch.

Book a free intro call →