GRR vs NRR: Why SaaS Companies Need Both

Two retention metrics, one common mistake

Gross Revenue Retention and Net Revenue Retention have become two of the most closely watched numbers in SaaS. They show up in every board deck, every investor update, every diligence process. And they generate a predictable set of questions in every finance team:

What's our NRR? What's our GRR? Why are they different? And the one that comes up most: which metric matters more?

That last question is the wrong one — and it's worth understanding why, because the framing itself causes confusion. GRR and NRR aren't competing measures of the same thing, where one is the better version. They answer different business questions. Asking which matters more is like asking whether a car's speedometer or its fuel gauge is more important. They tell you different things, and you need both to drive.

Understanding both — and, crucially, understanding what it means when they diverge — gives finance a far more complete picture of customer health than either number alone. This article is about what each metric actually measures, why they can tell opposite stories about the same business, and why mature finance organizations always report both together.

If you need short definitions first: see NRR and GRR in the glossary. The rest of this piece is about reading both numbers together — in the board pack and in diligence.

What is Gross Revenue Retention (GRR)?

Gross Revenue Retention measures how much recurring revenue you keep from your existing customer base — before counting any expansion.

That "before expansion" part is the whole point of GRR. It deliberately ignores upsells, cross-sells, and any growth within the base, and looks only at what you retained or lost. It reflects:

Customer retention — did customers stay?
Downgrades — did they move to cheaper tiers?
Contractions — did they reduce their spend?
Churn — did they leave entirely?

Because it excludes expansion, GRR can never exceed 100%. The best you can do is keep every dollar you started with — you can't "gain" in a metric that ignores growth. A GRR of 90% means you retained 90% of your starting recurring revenue and lost 10% to some combination of churn and contraction, regardless of how much you expanded elsewhere.

GRR answers one focused question: how well are we protecting the revenue we already have? It's a measure of the durability of your customer base — the leakage before any of the growth story is layered on top.

What is Net Revenue Retention (NRR)?

Net Revenue Retention measures how the revenue from your existing customers changes over time — after accounting for everything, growth included.

Where GRR stops at retention, NRR adds the expansion back in. It reflects:

Expansion revenue — customers spending more.
Upsells — moving to higher tiers.
Cross-sells — buying additional products.
Contractions — customers reducing spend.
Churn — customers leaving.

Because it includes expansion, NRR can exceed 100% — and for healthy SaaS businesses, it often does. An NRR above 100% means your existing customers, as a group, are worth more today than they were a year ago, even after accounting for the ones who left or shrank. That's the hallmark of a business where growth compounds within the base, not just from new logos.

NRR answers a different question than GRR: how much additional value are our existing customers creating over time? It measures the overall economic trajectory of the customer base — not just whether you kept it, but whether it grew.

GRR measures stability. NRR measures growth.

Here's the distinction that makes both metrics necessary, stated as simply as possible:

GRR answers: how well are we protecting the revenue we already have?

NRR answers: how much additional value are our existing customers creating?

GRR is a stability metric. It tells you how leaky the bucket is — how much revenue drains away before you pour anything new in. NRR is a growth metric. It tells you whether the water you already have is somehow multiplying — whether expansion is outrunning churn inside the existing base.

Both perspectives matter, and neither substitutes for the other, because each can hide what the other reveals:

High NRR can hide weak retention. A company can post an impressive NRR while quietly losing a meaningful share of customers — because strong expansion from the customers who stay masks the churn of the ones who leave. The growth story looks great; the retention problem is invisible in the NRR alone.
Strong GRR with weak NRR can signal limited growth. A company can retain almost every dollar (high GRR) but generate little expansion (NRR barely above 100%). The customer base is loyal and stable, but it isn't growing — which may point to limited upsell opportunity or an under-monetized product.
Strong performance requires understanding both. A genuinely healthy business retains well and expands well — high GRR and high NRR. You can only confirm that by looking at both. Either number alone can tell a flattering story that the other would complicate.

This is why the "which matters more?" question misleads. The valuable question is: what do GRR and NRR say together, and what does the gap between them reveal?

When GRR and NRR tell different stories

This is where interpretation gets genuinely useful — and where a finance leader adds value that a metric definition can't. Consider two companies:

Company A: GRR = 85%, NRR = 115%
Company B: GRR = 95%, NRR = 102%

Glance at NRR alone and Company A looks like the stronger business — 115% versus 102%. But that reading misses almost everything important, and the two numbers together tell a much richer story.

Company A is losing 15% of its recurring revenue to churn and contraction each year (GRR of 85%). That's significant leakage. What rescues the NRR is powerful expansion — the customers who stay are spending so much more that they more than offset the substantial losses, pushing NRR to 115%. This is a business with an excellent expansion motion papering over a real retention problem. The 115% NRR is impressive, but it's masking a leaky bucket. If expansion ever slows — a saturated base, a tougher economy, a pricing ceiling — that retention weakness will surface fast, because there'll be nothing offsetting the 15% leak. Company A should be worried about why customers are leaving, even while celebrating expansion.

Company B looks less exciting on NRR (102%) but is retaining almost everything (GRR of 95%). Only 5% of recurring revenue leaks away. The modest NRR tells you expansion is limited — existing customers aren't growing their spend much. This is an exceptionally loyal, stable customer base with less expansion opportunity. The question for Company B isn't retention — that's excellent — it's whether there's untapped upsell potential, a product-expansion path, or pricing power being left on the table. Company B should be asking how to grow its very sticky base, not how to keep it.

Same category of metrics, opposite underlying stories. Company A has a growth engine hiding a retention hole; Company B has a retention fortress with an underused growth lever. Neither is simply "better" — they have different strengths and different risks, and you can only see any of that by reading GRR and NRR together. That's the interpretation finance owes its executive team: not "NRR is 115%," but "NRR is 115% because expansion is masking 15% churn, and here's why that matters."

How the math actually works

Both metrics start from a cohort: the customers (and their ARR) you had at the beginning of the period. New logos acquired during the period do not belong in classic NRR/GRR. They belong in new ARR on the ARR waterfall.

From that starting ARR:

Subtract contraction (downgrades, seat loss, tier moves down).
Subtract churn (full logo or ARR exits).
For NRR only, add expansion (upsells, cross-sells, seat growth, price increases on renewals).

A clean way to say it:

GRR = (starting ARR − contraction − churn) ÷ starting ARR
NRR = (starting ARR − contraction − churn + expansion) ÷ starting ARR

If your waterfall cannot reproduce both from the same movements, the retention slide is not finished. GRR and NRR are not separate models. They are two readings of one bridge — see also waterfall and ARR.

A simple example

Start the quarter with $10.0M ARR in the existing base.

Expansion: +$1.8M
Contraction: −$0.7M
Churn: −$0.9M

Ending ARR from that cohort = $10.0M + $1.8M − $0.7M − $0.9M = $10.2M.

NRR = $10.2M ÷ $10.0M = 102%
GRR = ($10.0M − $0.7M − $0.9M) ÷ $10.0M = 84%

NRR says the installed base grew slightly. GRR says you kept 84 cents of every starting dollar before upsell. Both are true. Presenting only 102% invites the wrong celebration.

Why investors watch both metrics

Sophisticated investors never look at one of these in isolation, because each tells them something the other can't.

High GRR signals to investors:

Strong customer satisfaction — customers stay because the product delivers.
Stable recurring revenue — the base is durable and predictable.
A predictable business — low leakage means the revenue floor is solid.

High NRR signals something different:

Product expansion — the product has room to grow within accounts.
Pricing power — customers accept paying more over time.
Customer adoption — usage and reliance are deepening.
Long-term growth potential — the base compounds, so growth doesn't depend solely on new logos.

Notice these are different signals about different qualities. GRR speaks to the safety of the revenue; NRR speaks to its growth potential. An investor evaluating a SaaS business wants both — a durable base (GRR) that also compounds (NRR). A company strong on one and weak on the other raises exactly the questions the two-company example illustrates. This is why neither metric should be evaluated independently in diligence, and why a finance team that reports only its most flattering number invites skepticism rather than confidence.

Why finance needs consistent methodologies

Everything above assumes GRR and NRR are calculated consistently — and that's a bigger assumption than it sounds. Like ARR, neither metric has a single universal formula, so calculating them reliably requires documented methodology decisions:

Customer cohorts — how you define the starting base and the period being measured.
Expansion timing — when an upsell counts toward the period.
Contract amendments — how mid-term changes flow through.
Usage revenue — whether and how variable revenue counts toward retention.
Downgrades — how contractions are measured.
Reactivations — how a returning customer is treated.
Foreign exchange — how currency movement is handled, if applicable.
Partial-period customers — how customers who start or leave mid-period count.

A few definition traps that show up in the same meetings:

CRM ARR vs billing ARR. If Sales' expansion sits in the CRM and Finance's retention sits in billing, NRR and GRR will disagree with bookings commentary. The metrics are only as trustworthy as the source definitions behind ARR — and as a documented ARR methodology the whole company can reuse. Pick one governed definition for board retention and stick to it.
Including new logos in NRR. That inflates the rate and breaks comparability. New business is a growth line, not a retention credit.
Price increases counted as expansion without disclosure. Legitimate in many ARR policies — but boards should know when NRR is partly a pricing story.
Multi-year ramps and delayed starts. Ramp deals can look like expansion when they are really contracted schedule. Your ARR policy has to say which is which, consistently, every close.

Each of these is a judgment call, and the same underlying business can produce different GRR and NRR figures depending on how they're answered. Which leads to the same principle that governs ARR: methodology consistency matters more than matching another company's reported numbers. Comparing your NRR to a competitor's headline NRR is nearly meaningless if you've defined the metric differently — and you probably have. What matters is that you calculate GRR and NRR the same way every period, document how you do it, and can explain it. (This is the retention-metric case of the discipline we've written about in ARR governance and why there's no standard ARR calculation.)

Consistency is also what makes the trend trustworthy. A rising NRR only means something if this quarter's NRR was computed like last quarter's. Inconsistent methodology turns your retention trend into noise — which is why consistent calculation, traceable to source, underpins any useful retention reporting. (Inconsistent underlying data is a fast way to undermine this; see why poor financial data holds back finance teams.)

Executive reporting should include both

Given all this, board and executive reporting should never present one retention number in isolation. A complete retention picture includes:

GRR — the stability of the base.
NRR — the growth of the base.
Churn — the revenue lost to departures.
Expansion — the revenue gained within the base.
Contraction — the revenue lost to downgrades.

Presented together, these let executives see the full story rather than a curated slice. GRR and NRR frame it; churn, expansion, and contraction explain it. You can see not just that NRR is 112%, but that it's 112% because expansion of 22% offset churn and contraction of 10% — which is a completely different situation than 112% built on 4% churn and 8% expansion.

That's the real standard for retention reporting: executives should understand why NRR changed, not simply whether it went up or down. A number that moves without an explanation invites the follow-up questions that erode confidence. A number presented with its drivers — here's the expansion, here's the churn, here's what changed — is one the board can actually use to make decisions.

Board questions you should be ready for

These come up whether or not they appear on the slide.

"Is NRR above 100% because we retained customers — or because a few whales expanded?" Be ready to show concentration: top-10 expansion vs the rest of the base. Cohort NRR without a concentration note is easy to misread.
"What would GRR be if we excluded involuntary churn / bankruptcies / one large logo?" Have a policy before the meeting. Adjustments are fine when disclosed. Silent exclusions destroy trust.
"Why did GRR fall while NRR held?" Classic pattern: churn or contraction worsened, and expansion plugged the hole. That is an early warning, not a win.
"Are these logo-weighted or ARR-weighted?" Almost all board NRR/GRR should be ARR-weighted (dollar retention). Logo retention is a different chart. Mixing the two mid-answer is how meetings derail.
"Does this match the waterfall on the prior slide?" If ending ARR, expansion, contraction, and churn on the waterfall cannot regenerate the retention percentages, stop. Fix the pack before you defend a narrative. Related failure mode: when ARR, cash, and the P&L tell different stories because each came from a different export.
"Is starting ARR the same population RevOps uses for renewal forecasts?" Definition drift between finance and RevOps is one of the fastest ways to lose a room. Align the cohort rules in writing.

How to present GRR and NRR in the pack

A practical pattern that keeps the board oriented:

Waterfall first — beginning ARR → new / expansion / contraction / churn → ending ARR.
Retention second — GRR and NRR for the same period, same cohort, same movements.
One sentence on the gap — "NRR is 108%; GRR is 91%; the 17-point spread is expansion in the existing base, concentrated in enterprise."
Optional: segment — enterprise vs mid-market GRR/NRR often matters more than the blended number.

Keep commentary tied to the same figures as the table. If the narrative says expansion saved the quarter, the expansion dollar amount on the waterfall should be the one in the sentence. That is the same trust standard as the rest of SaaS board reporting: traceable numbers beat a polished story that cannot be checked.

When CFOs start hearing "which number is correct?" about retention, the root cause is usually conflicting definitions — not a missing chart. The fix is one cohort, one set of movements, two clearly labeled rates. See also why CFOs stop trusting their own numbers.

The goal isn't choosing one metric

Pull it all together and the conclusion is clear.

GRR and NRR are complementary, not competing. One measures the strength of your customer retention — how well you protect what you have. The other measures the growth potential of your existing customers — how much more they become worth over time. Together, they tell you both the stability and the scalability of your recurring revenue, which is exactly what finance, executives, and investors need to understand the health of a SaaS business.

Choosing between them means throwing away half the picture. The gap between them — as the two-company example showed — often contains the most important insight of all: whether expansion is masking churn, or whether loyalty is masking a growth ceiling.

So the principle worth leaving with:

Great SaaS finance organizations don't ask whether GRR or NRR is more important. They understand that each metric tells a different part of the company's growth story.

GRR vs NRR at a glance

The two metrics side by side, to keep the distinction clear:

Gross Revenue Retention (GRR) — Measures retained recurring revenue; excludes expansion; focuses on retention quality; highlights churn and contractions; capped at 100%; indicates revenue stability.
Net Revenue Retention (NRR) — Measures retained and expanded recurring revenue; includes expansion; focuses on customer growth; highlights total customer value; can exceed 100%; indicates long-term growth potential.

FAQ

What is Gross Revenue Retention?

GRR measures how much recurring revenue a company retains from its existing customers before any expansion — accounting only for churn, downgrades, and contractions. It's capped at 100% and reflects the stability and durability of the customer base.

What is Net Revenue Retention?

NRR measures how the recurring revenue from existing customers changes over time after including expansion, upsells, and cross-sells alongside contraction and churn. It can exceed 100%, and a figure above 100% means existing customers are worth more over time even after accounting for losses.

What is a good GRR?

It varies by segment, but higher is better and the ceiling is 100%. Enterprise-focused SaaS businesses often target GRR in the low-to-mid 90s, while lower figures can be normal for SMB-focused businesses with naturally higher churn. What matters most is measuring it consistently and understanding the trend.

What is a good NRR?

NRR above 100% is generally considered healthy, since it means expansion is outpacing churn and contraction. Best-in-class businesses often report meaningfully higher. But NRR should always be read alongside GRR, because a strong NRR can mask weak retention.

Why do investors care about NRR?

Because NRR indicates whether growth compounds within the existing customer base rather than depending entirely on new-logo acquisition. High NRR signals product expansion, pricing power, and deepening adoption — all markers of long-term growth potential. Investors read it alongside GRR to confirm the base is both durable and growing.

Should SaaS companies report both GRR and NRR?

Yes. They answer different questions — GRR measures retention stability, NRR measures growth within the base — and each can hide what the other reveals. Reporting both, alongside churn, expansion, and contraction, gives a complete and honest picture of customer health.

Should new customers be included in NRR?

No. Classic NRR and GRR measure the existing starting cohort. New logos belong in new ARR on the waterfall, not as a retention credit.

How do GRR and NRR tie to the ARR waterfall?

They should regenerate from the same movements: starting ARR, contraction, churn, and (for NRR) expansion. If the retention slide and the waterfall disagree, fix the pack before you defend a narrative.

Where SMPL.ai fits

SMPL.ai is built to compute retention metrics the way your business defines them — a finance operating system that applies your methodologies rather than imposing its own.

SMPL reads and reconciles data from your connected systems — billing, CRM, and the general ledger — and does not replace your systems of record. It computes GRR, NRR, and the churn, expansion, and contraction movements beneath them from one reconciled base, deterministically and repeatably, so the same inputs always produce the same outputs and this quarter's retention is computed like last quarter's. Every reported number can be traced back to its originating source, and validation and reconciliation occur before anything reaches executive reporting.

The AI explains financial performance — why NRR moved, what drove the gap between GRR and NRR — but it does not invent your financial methodologies. Your definitions govern the calculation; the AI describes the result. (Authentication today uses magic links.)

We do not claim a certification status here. Trust in board metrics comes from definitions, reconciliation, and the ability to drill from the percentage to the customers underneath.

If you'd like to see your own GRR and NRR computed consistently and traced to source, book a demo and we'll walk it on data that looks like yours.