Every board deck has the same slide. NRR sits in the top-right corner, usually green, usually above 100%, and usually treated as proof that the business is compounding. For a PE-backed CEO or an operating partner reviewing a portfolio company, that number is often the single most trusted signal in the room.
It shouldn’t be. Not because NRR is a bad metric, but because NRR above 100% is not one story. It is at least three, and they lead to entirely different conclusions about the health of the business.
An operating partner who takes the blended number at face value is approving a growth story without knowing which of the three stories they’re actually being told, or whether it’s more than one at once.
| Component | What the Board Sees | What It Can Hide |
|---|---|---|
| Source | NRR > 100% | Contractual escalators doing the work, not earned expansion — no organic growth engine |
| Concentration | NRR > 100% | One or two whale accounts masking flat or shrinking performance everywhere else |
| GRR-Masking | NRR > 100% | Base erosion offset by expansion elsewhere — the business is getting smaller and more concentrated at the same time |
The Escalator Problem
The first version of a healthy NRR number is the one everyone assumes: Customers are buying more because the product is delivering more value. Seats expand, usage grows, cross-sell lands. That’s earned growth, and it’s the version worth paying a premium for at exit.
The second version looks identical on the board slide and means something very different. Price escalators, CPI clauses, and seat minimums built into the contract accrue automatically, with no CS effort and no fresh conviction from the customer. A 4% annual increase baked into a three-year contract shows up in NRR the same way a genuine upsell does. Finance tracks the contract terms. The customer success or revenue team tracks usage and expansion. Those two systems rarely talk to each other, and the blended NRR figure erases the distinction before it ever reaches the board.
A company running mostly on escalators can post a strong NRR number for years with no organic growth engine underneath it. The number is real. The story it implies is not.
The one question that surfaces it: Break out NRR by contractual increase versus new or expanded usage. What’s the split?
The Whale Problem
The second failure mode hides in distribution rather than source. A blended NRR number treats every dollar of expansion the same, whether it came from a handful of accounts or was spread across the base. That means a small number of large accounts expanding aggressively can lift the entire portfolio company’s NRR even while most of the customer base is flat, shrinking, or churning underneath.
This is the version an operating partner is least likely to catch, because NRR is almost always reported at the company level, not the account-distribution level. One whale’s renewal, timed well, can carry the whole quarter’s number. The board sees strength. What they don’t see is that the strength lives in three accounts, not three hundred, and that concentration is its own risk heading into a renewal cycle or a market downturn.
The one question that surfaces it: What does NRR look like with the top three accounts excluded?
The Number That Lies Best
The third version is the most dangerous, because it’s the only one where the board number looks best while the business is doing worst.
Gross revenue retention measures whether the existing base is holding. Net revenue retention measures the base plus expansion. When GRR is declining, meaning accounts are churning or shrinking, and NRR is still healthy or even rising, the only explanation is that expansion elsewhere is offsetting the losses. The board sees one green number and reads it as growth. What’s actually happening is a business that is getting smaller and more concentrated at the same time, with the erosion fully buried inside a metric built to reward compounding, not decay.
By the time GRR decline shows up in NRR, the underlying problem has usually been building for two or three quarters. NRR isn’t just failing to catch it. It’s actively burying it.
The one question that surfaces it: Put GRR and NRR on the same slide, same period. If GRR is declining while NRR holds or rises, which accounts are doing the offsetting?
That’s not a one-time question — it’s the recurring diagnostic our Outcome-Based CS Framework runs at the portfolio level: see how it works →
Three Stories, One Number
None of this is an argument against NRR. It’s the most useful retention metric a board has, and no operating partner should stop tracking it. The argument is narrower: A single blended number cannot tell you which of three very different businesses is producing it, and only one of the three is the growth story it’s usually assumed to be.
The diagnostic is three questions, not a new reporting system. Source, concentration, and GRR-masking. Any portfolio company CFO can answer all three from data they already have. The only requirement is knowing to ask.
The next portfolio review is the place to ask it: which of these three stories is producing this company’s NRR? Our Outcome-Based CS Framework gives operating partners the diagnostic to find out — see the framework →
Post 7 in this series: What good looks like at $25M ARR versus $75M ARR — and why the same GRR number means something different at each stage.
Andrea Mulligan is a B2B SaaS executive and advisor with 30 years of experience building Customer Success, Professional Services, and GTM organizations. She works with PE-backed and growth-stage companies on CS transformation, revenue retention strategy, and post-sale model design. Start a conversation →