NRR Cliff: 98% to 121% in Two Quarters, No New Sales

Derek SimmonsHi, welcome back to ARR Autopsy. I'm Derek, and sitting across from me as always, Elena. And Derek, I gotta say, the number you sent me this morning kept me up.

Elena Reyes98.

Derek Simmons98% NRR, not the rosy 106% blended number every deck quotes, the actual median for companies sitting between 1 and 10 million ARR.

Elena ReyesYeah, yeah. According to ProductQuants new benchmark piece. That's the real floor most founders are standing on.

Derek SimmonsWow. And our guest today started exactly there, 98%. Two quarters later, 121. No new logos.

Elena ReyesWait, none?

Derek SimmonsZero. Just fixed what was already inside the base. Okay, so today we're going back to 4.2 million ARR, 11 people on the team, a two-person CS group trying to hold the whole thing together, and there's something ugly hiding under that 98%.

Elena ReyesWe're not spoiling it yet. Oh, you're going to want to hear this one. Three levers, one pricing shakeup, and a whale account that almost torpedoed the whole story. I've already got my advisor hat half on. I want to know if that 121% is real or if it's three accounts carrying the whole team. Classic problem. And then when it worked, the operation cracked. Fixing retention broke something entirely elsewhere. We'll get there. All right, let's start where every founder starts. Stuck. Ninety-eight% stuck. Ninety-eight%

Derek SimmonsNinety-eight% what?

Elena ReyesMedian net revenue retention for companies sitting between $1 and $10 million in ARR.

Derek SimmonsWait, that's it? Everyone quotes 106.

Elena ReyesYeah, Productquant ran the numbers. All B2B SaaS median NRR is 106%, but that's blended across every stage. Plot twist. So the good number everybody benchmarks against doesn't even apply.

Derek Simmonseven apply to most companies raising money.

Elena ReyesWhich is exactly the trap. Founders walk into a Series A pitch, board tells them they need 110, 120 NRR, and meanwhile their actual cohort median is Ninety-eight.

Derek SimmonsNo wonder NRR turns into the diligence landmine every single time.

Elena ReyesEvery time.

Derek SimmonsOkay, let me put on my advisor hat for a second. The spread underneath that gets worse the more you segment it, right? Get this, DigitalApplied's 2026 data has Enterprise SaaS

Elena ReyesSaaS at 118 median NRR. SMB 97. Twenty-one points. That's not a rounding error. That's a different business model. And it vanishes the second somebody quotes you one blended stat like it means something universal.

Derek SimmonsRight. So where does today's founder land in all this?

Elena ReyesLeaning in. Started at ninety-eight percent NRR. Textbook median. Nothing alarming on paper.

Derek SimmonsAnd...

Elena ReyesTwo quarters later, One-twenty-one percent. without adding a single new logo.

Derek SimmonsHold on, 23 points of NRR with zero new acquisition?

Elena ReyesYeah.

Derek SimmonsZero. So if 98 isn't a five-alarm fire, what made this person go touch the number at all?

Elena ReyesBuilding on that one 21% tease, let's rewind. Where exactly were they stuck?

Derek Simmons4.2 million ARR, 11 people total, and customer success was two humans splitting renewals and support tickets.

Elena ReyesTwo people covering retention on 4.2 million? Help me understand. How many quarters did that 98 just sit there?

Derek SimmonsThree straight, and it's not like they sat on their hands. Quarterly business reviews and annual price increase that didn't spook anyone?

Elena ReyesPlus a CS hire.

Derek SimmonsRight, but no health scoring process behind it. No usage thresholds, no red flags, just reactive check-ins whenever an account complained.

Elena ReyesSo we worked hard on retention actually meant.

Derek SimmonsOn a random Tuesday, one person scanning a spreadsheet hoping nobody's usage quietly dropped off. That's the whole system.

Elena ReyesSo that's it?

Derek SimmonsThat's it. And the Ninety-eight never moved.

Elena ReyesOkay, here's what bugs me. Ninety-eight% NRR includes expansion dollars. What was gross retention doing underneath that number?

Derek SimmonsNobody was tracking it. When they finally pulled it, gross retention was sitting high Eighty-seven-ish.

Elena ReyesEighty-seven?

Derek SimmonsEighty-seven. A handful of expansion accounts were quietly covering for real churn happening in the rest of the base. Peace.

Elena ReyesSo the ninety-eight was basically a mask.

Derek SimmonsFiscallions stage-by-stage breakdown puts private B2B SaaS median gross retention around eighty-eight to ninety-two percent, so eighty-seven wasn't wildly off-benchmark, it was just invisible to this team.

Elena ReyesAnd an NRR number floating just above one hundred usually means expansion is barely outrunning churn underneath it.

Derek SimmonsExactly the trap they were in, one soft renewal quarter away from- from NRR dropping under ninety.

Elena ReyesKnock desk. So what does this mean for anyone listening at ninety-eight percent with new logos still coming in?

Derek SimmonsPull your GRR before you feel good about anything. High eighty-s GRR with NRR near one hundred just means expansion is covering for churn underneath.

Elena ReyesWhich is the diagnosis that flipped everything for this team.

Derek SimmonsAnd once they saw that number, they didn't chase upsells first.

Elena ReyesThree specific moves in a very particular order. Order and the sequencing is what actually surprised me. Okay, so building on that GRR mess, the founder didn't chase upsells, stop the leak first, then build the ladder. That's the whole sequencing. I like that. So walk me through lever one. Dunning optimization, just fixing failed payment retries. ProductQuant's piece on NRR benchmarks flags this is worth one to three points of NRR, and it shipped in under two weeks. Two weeks for that kind of lift? Yeah, that's silent involuntary churn. Cards expiring, banks flagging the charge, nobody's canceling on purpose, the invoice just fails and dies quietly.

Derek SimmonsWhich never even shows up as a support ticket.

Elena ReyesExactly, it just evaporates off the books. Lever two took longer: proactive usage monitoring, right-sizing seats before renewal instead of waiting.

Derek Simmonswaiting for the downgrade email.

Elena ReyesOkay, but right-sizing is just giving customers a reason to pay you less?

Derek SimmonsSure, term, sure. But Elena, think about it. If you shrink the invoice before renewal, you're negotiating from a position of trust, not defending a number nobody believes anymore.

Elena ReyesFair. Fewer surprise cancellations because you got ahead of the conversation.

Derek SimmonsRight. Now, level three, the one everybody wants to talk about, expansion. They shifted parts of... It's of pricing toward usage-based triggers.

Elena ReyesWait for it.

Derek SimmonsM3's analysis on this pegs usage-based models at 115 to 130% NRR versus 95 to 105 for flat rate plans.

Elena ReyesThat's a 30-point swing just from how you bill someone.

Derek SimmonsThat's the number that made me sit up.

Elena ReyesOkay, but that's one analysis, not gospel. I'd trade it as directional.

Derek SimmonsFair, fair. But here's the part I actually respect. In fact,

Elena Reyesthey didn't lead with Lever three.

Derek SimmonsRight. They fixed Levers one and two first.

Elena ReyesGRR moved from that high eighties number into the low nineties before anyone touched pricing.

Derek SimmonsSo the Lever had a floor under it before they started climbing!

Elena ReyesThat's the move. Most founders do it backwards-chase expansion revenue to paper over churn they haven't fixed.

Derek SimmonsWhich just delays the reckoning.

Elena ReyesAnd makes it bigger when it hits.

Derek SimmonsSo three levers, two quarters landed at one twenty one.

Elena ReyesOkay, but let me push on that number for a second. ProductQuant has enterprise NRR at one eighteen, mid market at one oh eight, and SMB at ninety-seven.

Derek SimmonsRight, so one-twenty-one clears enterprise tier benchmarks at an eleven person company.

Elena ReyesWhich is exactly what makes me suspicious-is that broad based, or is that three big accounts doing all the lifting?

Derek SimmonsHuh. Good instinct.

Elena ReyesBecause usage-based expansion loves to concentrate in your biggest accounts. If two whales carry the number... Exactly.

Derek SimmonsThe other ninety-eight customers could be flat or worse.

Elena ReyesSo before we call 121 real, I want to see it sliced by account size.

Derek SimmonsWhich apparently is exactly what this founder did next.

Elena ReyesBuilding on that suspicion, here's the trap that wrecks cap tables:

Speaker 3Wait for it...

Elena ReyesA nine million ARR company reported a shiny one-eighteen percent ARR, enterprise-grade, on paper.

Speaker 3And?

Elena ReyesTwo accounts drove sixty one percent of that expansion. Strip them out, forward ARR craters to ninety seven percent.

Speaker 3So the textbook number was two whales in a trench coat.

Elena ReyesPretty much. Investors caught on. The screen now is a twenty plus point gap between gross retention and net retention at Series A and Series B.

Speaker 3Because expansions covering for churn nobody fixed.

Elena ReyesThat's the pattern. So when our founder handed me one hundred and twenty one percent, I didn't take the headline, I asked for the cohort breakdown.

Speaker 3Line by line.

Elena ReyesLine by line rank every account by expansion dollars. See what the top two or three are carrying; top two accounts were fourteen per cent. of the expansion, not sixty one.

Speaker 3Okay, that's actually broad based.

Elena ReyesWhale-adjusted, the number only slipped to one hundred sixteen per cent., still Enterprise-grade tier.

Speaker 3That's the figure that survives due diligence.

Elena ReyesRight-spread across dozens of accounts upgrading tiers, not two logos propping up a deck.

Speaker 3I want to get specific on something, though. Not all the movement inside that number That number is the same kind of movement.

Elena ReyesMeaning...

Speaker 3There's Packaging churn and there's Product churn and people lump them together constantly.

Elena ReyesWalk me through it.

Speaker 3Packaging churn is a customer downgrading because you jammed them into the wrong tier, wrong seats, wrong bundle. That's a pricing miss.

Elena ReyesFixable with a form, basically.

Speaker 3Pretty much. Product churn is someone leaving because the thing didn't work for them. That's the scary version.

Elena ReyesWhich one was driving the founder's original churn?

Speaker 3Packaging almost entirely, which, help me understand, is the good version of this problem.

Elena ReyesBecause you don't rebuild the product.

Speaker 3You rebuild the pricing page: different fire drill entirely.

Elena ReyesSo the one twenty one holds, the churn source is diagnosed, but fixing it broke something else. The team that built this motion is drowning in it now.

Speaker 3Of course it is. Nothing's ever just solved.

Derek SimmonsBuilding on that packaging fix turns out solving churn broke something else entirely.

Elena ReyesOh, this is the part where the wind turns into a headache.

Derek SimmonsTwo person CS, remember? Back at four point two million. They fixed retention, but now every account needs a health score and proactive outreach.

Elena ReyesSo the fix that raised NRR is also drowning the team running it.

Derek SimmonsExactly. Manual reviews on every expanding account. That doesn't scale with two humans.

Elena ReyesHow many accounts are we talking that they're trying to touch by hand?

Derek SimmonsEnough that expansion conversations backed up. Renewal timing slipped because CS couldn't reach accounts. reach accounts before the window closed.

Elena ReyesSo the growth engine stalls right as it starts working.

Derek SimmonsRight. So they killed the calendar-based QBR and built usage-triggered upgrade prompts instead.

Elena ReyesIn-app, not a rap on a Zoom call.

Derek SimmonsAccount hits a usage ceiling, the system surfaces the upgrade in that moment. Nobody has to notice it first.

Elena ReyesProductGrowth's research on expansion triggers backs that up. Prompts fired at the moment of need convert way higher than a schedule. A scheduled review nobody asked for.

Derek SimmonsBecause the QBR sits on your calendar, the usage prompt sits on theirs.

Elena ReyesDoes that replace CS or just take pressure off? Takes pressure off. CS still owns the relationship, they just stopped being the trigger for every single upsell. Fair. So what's the takeaway for someone staring at their own NRR right now?

Derek SimmonsCompare yourself to your segment, not the blended average. Productquant's breakdown makes that case directly. A flat one hundred per cent target is the wrong target depending on where you sit.

Elena ReyesAnd fix gross retention before you chase expansion.

Derek SimmonsThat's the sequencing this founder followed, dunning, then seat right-sizing, then usage pricing, in that order.

Elena ReyesAnd if the automation doesn't hold up once usage patterns get messy?

Derek SimmonsThen you're back to two people drowning in manual outreach. So if retention just clicked for you and supports underwater... build the trigger before you build the headcount. Thanks for watching!

Elena ReyesOkay, so before we let you go, that GRR reveal, the gap between the high 80s and the shiny NRR number, that's the one that's sticking with me.

Speaker 3Right.

Elena ReyesExpansion quietly covering for churn. Sneaky but effective diagnosis. The takeaway? Stop measuring against one blended NRR number. Check your segment first. Exactly what Digitalapplieds breakdown was built for.

Speaker 3And watching that founder go from stuck. Stuck to broad-based expansion in two quarters without a single new logo.

Speaker 4Mm-hmm.

Speaker 3still gets me. Great one. If this saved you from a bad renewal bet, send it to a founder who needs it. Subscribe on YouTube, wherever you listen, and leave us a review. It genuinely helps. We want more founders showing us their real numbers. Thanks for tuning in, everyone. We'll see you next time. Bye, all.