Gross revenue retention (GRR) measures what you kept from existing customers after churn and downgrades, capped at 100%. Net revenue retention (NRR) adds expansion revenue on top, which is why it can climb past 120% while telling you almost nothing about how many customers you're actually losing. Investors ask for both because one shows the leak, the other shows the water level — and founders who only lead with NRR tend to get a very specific follow-up question in diligence.

If you've pitched NRR as your headline growth metric and watched a VC's eyes narrow slightly, this is probably why.

What GRR and NRR Actually Measure

Start with the formulas. The gap between them is the whole story.

Gross Revenue Retention = (Starting MRR − Churn − Contraction) / Starting MRR

GRR only ever moves in one direction from 100%. You can't expand your way to a higher number — upsells and cross-sells are deliberately excluded. It's the most conservative retention metric in SaaS, and that's the point. It isolates one question: of the revenue you started the period with, how much is still there?

Net Revenue Retention = (Starting MRR − Churn − Contraction + Expansion) / Starting MRR

Add expansion revenue back in — seat growth, upsells, cross-sells — and NRR can sail past 100%, 120%, even 150% in strong PLG businesses. Mathematically, NRR is always greater than or equal to GRR. There's no scenario where it isn't, because it's GRR plus a non-negative expansion term.

That relationship is exactly why a single NRR number gives you an incomplete picture. Two businesses can post identical NRR figures with wildly different underlying health, because the same end number can be built from two completely different combinations of churn and expansion.

The two metrics respond to different levers, too. NRR is a whole-business number — onboarding speed, time-to-value, customer success quality, even sales comp structure for expansion deals, all move it. GRR is narrower and more diagnostic: mostly product value and churn prevention. When GRR is soft, it's rarely a sales problem. It's usually a product-market fit or customer success problem wearing a sales metric's clothing.

The Masking Problem: Why NRR Alone Lies to You

Here's the uncomfortable bit. A company can lose a quarter of its customer base every year and still report NRR above 115%, as long as the customers who stick around spend enough more to cover the hole. On paper, that looks like a growth story. In reality, it's a shrinking customer base propped up by a shrinking group of accounts paying more each — a much riskier thing to underwrite.

Call it the "leaky bucket" problem. NRR tells you whether the water level in the bucket is rising. GRR tells you how big the hole in the bottom is. A rising water level feels great until you ask how fast it would drop if expansion slowed for two quarters — which, in a tighter budget environment, it often does.

Sophisticated investors have seen this movie before. A founder who leads every deck with NRR and never volunteers GRR raises a specific kind of suspicion: not that the business is necessarily bad, but that the founder either doesn't understand the distinction or is hoping the investor won't ask. Neither is a great look in a data room. The fix is simple and costs you nothing — show both, unprompted, before anyone has to ask.

The businesses that get the benefit of the doubt are the ones where GRR and NRR tell a consistent story — stable or improving GRR, with NRR climbing on top of it through genuine expansion. That combination says: we're not losing people, and the people who stay are growing with us. It's the difference between compounding and compensating.

The Valuation Story: Why GRR Is the Sanity Check Behind the Multiple

NRR gets the headline because it correlates hard with valuation multiples. According to Software Equity Group data cited by Digital Applied, public SaaS companies with NRR above 120% have traded at a median of roughly 9.3x EV/revenue, compared to about 3.1x for companies below 100% NRR. That's roughly a 3x multiple gap driven by one metric. No wonder founders optimise for it.

But that multiple is priced on an assumption: that the NRR is real growth, not churn wearing an expansion costume. This is exactly where GRR becomes the sanity check. An investor underwriting a 9x multiple on the back of 125% NRR needs to know whether that's sitting on top of 95% GRR (excellent, durable business) or 70% GRR (a business hemorrhaging logos that happens to have a few whales expanding fast enough to hide it).

The second scenario is a much worse bet at the same headline multiple, because whale concentration risk compounds. If two or three expansion accounts are carrying your NRR number and one of them churns, your growth story doesn't just slow — it can go negative overnight. Investors who've been burned by this before will specifically model what happens to NRR if your top 5 expansion accounts flatline. GRR gives them the floor; NRR gives them the ceiling. They need both to price the range in between.

This is also why growth-stage and buyout investors increasingly ask for GRR and NRR broken out by cohort rather than as a single trailing-twelve-month snapshot. A single NRR number can be flattered by timing. A cohort view — how the customers who signed in Q1 2023 have retained and expanded through today, compared to the Q1 2024 cohort — shows whether the retention profile is improving, stable, or quietly degrading as the company scales. One clean number is easy to produce. A consistent multi-cohort trend is much harder to fake, which is exactly why sophisticated diligence teams ask for it.

Same NRR, Wildly Different Business: A Worked Example

This is the part most explainers skip — and it's the part that actually changes term sheets. Take two SaaS companies, both reporting 115% NRR at year end.

Company A: 92% GRR + 115% NRR This company lost 8% of its starting revenue to churn and contraction, then grew what was left by 23 points through expansion. That's a strong, sticky customer base with real upsell motion. Few logos leaving, and the ones that stay are buying more. This is widely regarded as one of the healthiest retention profiles in SaaS — sometimes called "land and expand done right."

Company B: 75% GRR + 115% NRR This company lost a quarter of its starting revenue — a serious churn problem by any standard — and then needed a massive 40-point expansion lift from the survivors just to land at the same 115% NRR headline. Same number on the slide. Completely different business underneath it.

What would an investor actually do differently with Company B?

They'd dig into customer concentration immediately — how much of that 40-point expansion comes from the top 3-5 accounts? They'd ask for logo-level churn reasons, not just the aggregate rate, to separate "bad fit" churn from "we're losing to a competitor" churn. And they'd almost certainly haircut the valuation multiple relative to the headline NRR, because the growth is structurally fragile — it depends on a shrinking pool of accounts continuing to expand at an unsustainable rate.
In a term sheet, this often shows up as a lower EV/revenue multiple, more aggressive earn-out structures tied to retention holding steady, or simply a pass, if the churn looks like a product problem rather than a segment-fit problem.

The term sheet difference between these two companies can be substantial, even though a founder reading only the NRR line would present them identically. That's the whole argument for showing GRR alongside NRR every time: it's the difference between "trust the headline" and "show your work."

The 2026 Benchmark Reality Check

Think your GRR/NRR gap looks bad? Check the market average first. Aleph's 2025 SaaS metrics benchmarking puts median GRR at 84% and median NRR at 102% — an 18-point gap. That gap is the clearest evidence you'll find that expansion revenue is doing serious heavy lifting across the industry right now. The "median" SaaS company loses 16% of its revenue base annually to churn and contraction, and only grows net revenue because expansion covers most of that hole.

Put plainly: a lot of what gets presented as "growth" in SaaS pitch decks is actually "recovery." Not necessarily damning — expansion-led growth is a legitimate, often capital-efficient motion — but it changes how you should read a standalone NRR figure in 2026. A 102% NRR sitting on an 84% GRR is the market median, not an outlier. Match that, and you're not underperforming. But NRR at 102% with GRR stuck at 70%? You're below median on the metric that actually predicts durability, even though your growth number looks fine on paper.

Benchmarks also vary a lot by business model, and treating them as one-size-fits-all trips up plenty of founders comparing themselves to the wrong peer set.

  • Enterprise, sales-led SaaS typically runs higher GRR (90%+) because contracts are sticky, multi-year, and hard to churn out of mid-term. NRR expansion tends to come from seat growth and module upsells sold by a dedicated success or sales team.
  • SMB and PLG-led SaaS tends to run lower GRR (sometimes 75-85%) because smaller customers churn more easily and face less switching friction. NRR can still look healthy if usage-based pricing or seat expansion scales with customer growth — but the GRR floor is genuinely lower by category, not by execution failure.

If you're an SMB PLG company benchmarking your 80% GRR against enterprise peers at 93%, you're measuring against the wrong comparison set. Investors who specialise in your category already know this. Founders who don't account for it in their own narrative end up either overclaiming or underselling their position.

What VCs Want vs What PE Firms Want in Diligence

GRR and NRR get asked for in every serious diligence process. But growth-stage VCs and private equity/growth equity buyers weight them differently — and knowing which audience you're in front of changes what you lead with.

VC diligence (Series A-C) is mostly forward-looking. VCs are underwriting a growth trajectory and a market, so NRR carries real weight as a signal of product-market fit and expansion motion — can this thing compound without constant new-logo spend? But a GRR materially below category benchmark is a red flag precisely because it's early evidence of a retention problem that only gets harder to fix at scale. Early-stage VCs will tolerate lower absolute NRR more than later-stage buyers will, but they're allergic to weak GRR trending the wrong way — it suggests the core product hasn't found its retention floor yet.

PE and growth equity diligence (buyouts, platform investments, growth equity checks into profitable-ish businesses) leans much harder on GRR. These buyers are underwriting cash flow durability, not just a growth story, and GRR is a far better predictor of what revenue survives a leadership change, a price increase, or a recession-driven budget freeze. PE firms will often build their base case off GRR alone and treat NRR-driven expansion as upside, not baseline — the opposite of how a lot of founders mentally weight the two metrics. Walk into a PE process still leading with NRR as your primary metric, and you're speaking the wrong investor's language.

Neither group ignores the other metric — that's the whole point here — but knowing which one is the headline for your audience saves you a round of follow-up questions and signals you actually understand how your own business gets underwritten.

Fixing Low GRR Without Accidentally Making It Worse

Here's the trap: a team spots weak GRR, panics about the growth number, and throws more resource at expansion — more upsell campaigns, more aggressive seat-expansion pricing, more sales comp tied to upgrades. NRR ticks up. GRR keeps sliding. Nobody notices until the next fundraise, when an investor asks why churn's been quietly getting worse for four straight quarters while the board slide only ever showed NRR.

Fixing GRR means treating it as a product and customer success problem, not a sales problem. The levers are genuinely different from NRR's.

  1. Segment your churn before you touch anything else. Separate "bad fit" churn (wrong ICP, wrong use case) from "lost to a competitor" churn from "failed to deliver value in time" churn. Each needs a different fix, and lumping them into one churn rate hides which lever actually matters.
  2. Audit time-to-value, not just onboarding completion. A customer who technically finished onboarding but never hit their first meaningful outcome is a churn risk wearing a disguise. This is usually where GRR leaks start.
  3. Put customer success incentives on retention, not expansion, for at-risk accounts. If CS comp is tied to upsell revenue, reps will rationally chase expansion in healthy accounts and under-invest in saving shaky ones — exactly backwards from what GRR needs.
  4. Run a pricing and packaging review focused on contraction, specifically. Contraction (customers downgrading, not fully churning) is often the quieter half of the GRR leak, and gets way less attention than full churn. If downgrades cluster in one tier or feature set, that's a product signal, not a pricing one.
  5. Don't suppress expansion to "protect optics." The goal isn't to slow NRR growth — it's to stop treating expansion as a substitute for fixing retention. Both can improve together. They're not actually in tension once you stop using one to paper over the other.

Founders who get this right tend to report GRR and NRR together internally, monthly, at the cohort level, long before any investor asks for it. By the time it shows up in a data room, it's not a scramble — it's a slide they already have.

Building a Board Slide Investors Can't Poke Holes In

Most retention slides are a single trailing NRR number with a trend arrow. That's not enough anymore, and it invites exactly the follow-up questions you're trying to avoid. A board-and-investor-ready retention slide needs three things on it, together — not scattered across three appendix pages.

1. GRR and NRR side by side, trended over at least 6-8 quarters. Not a snapshot. A trend. One quarter of 95% GRR means nothing if the three before it were 82%, 85%, 88%. The trajectory is the signal, not the point-in-time number.

2. Cohort-level breakdown, not just blended company-wide figures. Show how your 2023 cohort, 2024 cohort, and 2025 cohort have each retained and expanded over time. This is what separates a defensible retention story from a number that just happens to look fine this quarter because of timing. Investors increasingly expect this view by default in diligence — don't make them ask for it.

3. A one-line explanation of the gap. If your NRR is 118% and GRR is 88%, annotate it: "30-point gap driven primarily by seat expansion in top-quartile accounts, concentrated in [segment]." That sentence does more work than any amount of extra chart formatting — it shows you understand your own numbers, not just report them.

This slide format does double duty. Internally, it forces your team to track GRR with the same rigor as NRR, which tends to surface problems earlier. Externally, it signals to any investor reading it that you're not hiding behind the flattering number — which, counterintuitively, often makes your NRR figure more credible, not less.

Frequently Asked Questions

Can GRR ever be higher than NRR?

No. NRR always includes everything GRR includes (starting revenue minus churn and contraction) plus expansion revenue on top. Since expansion revenue can't be negative, NRR is mathematically always greater than or equal to GRR — they're only ever equal in the rare case where a company has zero expansion revenue in the period.

What's considered a "good" GRR for a SaaS company in 2026?

Above 90% is generally considered strong, and anything above 95% is excellent, though the right benchmark depends heavily on segment. Enterprise, sales-led SaaS companies often run GRR in the low-to-mid 90s, while SMB and PLG-led businesses more commonly sit in the 75-85% range due to higher natural churn in smaller accounts — against a 2025 market median of 84% reported by Aleph.

Why do some investors care more about GRR than NRR?

GRR strips out expansion revenue, so it isolates pure retention and churn signal without being flattered by upsells. Private equity and growth equity buyers in particular lean on GRR because they're underwriting cash flow durability — what revenue survives a leadership change or budget cut — rather than a forward growth narrative, which is more of a venture capital lens.

Is a declining GRR always a red flag, even if NRR is rising?

Generally yes, and it's one of the fastest ways to lose credibility with a sophisticated investor. Rising NRR paired with falling GRR usually means expansion revenue from existing accounts is masking an accelerating churn problem, which is a much riskier retention profile than it appears on the surface.

How often should SaaS companies report GRR and NRR internally?

Monthly, at the cohort level, not just as a single trailing twelve-month figure. Tracking both metrics together on a regular cadence helps teams catch revenue leaks early, size customer success investment correctly, and walk into a fundraise or diligence process with a trend they already understand rather than a number they're seeing for the first time.

If you're building your next board deck or fundraising data room, don't lead with NRR alone and hope nobody asks the obvious follow-up. Pull your GRR and NRR trends for the last six to eight quarters, break them out by cohort, and write the one sentence that explains the gap between them. That single slide will save you more diligence calls than any amount of polish on the rest of the deck.