9 Mothers: The $200M That Warped the Batch

MilesOkay, so my phone hasn't stopped buzzing since Tuesday.

GrantSame group chat?

MilesSame group chat. Somebody dropped a number in there and just left it. No context.

GrantThat's the S '26 batch for you. Total silence, then chaos.

MilesTwo hundred million.

GrantWait, as a valuation? Before Demo Day even happens?

MilesDays before. And this is Miles with Grant. We spend every week following one startup through one decision, and this week, the decision isn't even ours to make. It's everyone else's.

GrantBecause that number just sits there.

MilesRight. It's nearly double whatever the batch's previous high mark was. Nobody's naming the company yet.

GrantNot yet.

MilesBut you don't need the name to feel what it's doing. Every founder in that batch checking their own term sheet against a ghost.

GrantA single data point warping forty other negotiations.

MilesBefore we say who it is, though, I wanna sit in the room this thing actually dropped into.

GrantThe group chat.

MilesThe group chat, Bookface, X, the whole feedback loop. Picture the room for a second. Demo day isn't some Zoom call. It's invite-only, roughly fifteen hundred investors and press packed into San Francisco for one day.

GrantFifteen hundred people who all know each other's Twitter handles.

MilesBasically, yeah.

GrantSo when one number lands in that room-

MilesIt doesn't stay contained. It becomes the thing everybody's whispering about between pitches.

GrantRight, because there's nowhere for it to go. You're standing shoulder to shoulder with the same eleven to twelve investors all day.

MilesAnd this isn't even the first time a YC batch has had a number like that ripple through the room. TechCrunch talked to VCs about the spring batch, a different cohort, months before this one, and some of those startups were already commanding valuations over a hundred and seventy-five million.

GrantWait, seriously? A hundred and seventy-five? And that was considered the outlier back then?

MilesThat was the ceiling. That was the number people gasped at.

GrantSo the two hundred million figure this batch is dealing with isn't just breaking their own record.

MilesIt's breaking the record of the batch before them, too. It's not one data point drifting up a little. It's a jump.

GrantHuh. So walk me through the room again. What actually happens mechanically when that number drops?

MilesSo you've got a founder who just walked offstage. Solid pitch, decent traction, nothing crazy, and the investor they're talking to five minutes later has just heard about the two hundred million dollar company two booths down.

GrantAnd now every conversation after that gets measured against it.

MilesEven consciously. Nobody says it out loud, but the founder can feel it, like their normal round suddenly needs to justify why it isn't that.

GrantThat's the part that gets me. It's not even a competing offer. It's a rumor sitting three tables away.

MilesAnd it changes body language. Investors start asking different questions. Founders start second-guessing numbers they were completely confident in that morning.

GrantI bet watching deals get weird for way dumber reasons than that.

MilesYeah, I bet.

GrantI mean, professionals are supposed to be immune to that stuff. Turns out, they're not.

MilesNobody's immune to it. You could run the numbers cold all day, but the one figure you stick in your head by six PM is the biggest scary print you heard at breakfast.

GrantIt's just attention. The loudest number in the room becomes the reference point whether anyone consciously picks it or not.

MilesAnd once it's the reference point, every other pitch in that room gets read against it, fair or not.

GrantBut here's what I keep coming back to. That room only lasts for a day. Fifteen hundred people go home that night.

MilesTrue.

GrantSo where does the number actually do its damage? Because it's not like it dies when the room empties.

MilesNo, it doesn't. It follows people onto their phones.

GrantGive me a picture of that. What's that actually look like at eleven PM?

MilesSomebody's rereading their own term sheet on their phone in bed, comparing a normal cap table to a number they only ever heard secondhand.

GrantRight, and that number doesn't stay in the room. It follows founders straight onto Bookface.

MilesSo picture the actual thread. Someone posts their term sheet, twenty million cap, totally solid seed.

GrantNormal deal.

MilesCompletely normal. And under it, somebody drops a laughing emoji. Not mean, just reflexive.

GrantBecause next to two hundred, twenty looks like a rounding error.

MilesRight, and that founder didn't do anything wrong. Their number didn't move. Everyone else's read of that number moved.

GrantThat's the whole thing about anchoring, though. It's not that the twenty million deal got worse. It's that the reference point everyone's using got hijacked.

MilesI saw this constantly in institutional finance. You'd get one wild trade print, and for the next hour, every other desk is repricing off that one tick, even when it's not a real market.

GrantEven when the tick is basically noise.

MilesEspecially then. Nobody wants to be the guy who sold at yesterday's price. It's such a specific kind of stupid, too.

GrantYou know the number's noise, and you still can't unsee it.

MilesYou can't. I've watched a desk reprice a bond over a single erroneous print that got corrected an hour later, and the damage from that hour still cost people real money.

GrantSo the correction doesn't even matter once the panic's already spent.

MilesExactly, and nobody sends a follow-up post walking back the vibe. The laughing emoji just stays there.

GrantAnd on X, it's worse because it's not just the batch. It's every VC's timeline.

MilesSo now every founder closing a normal seed is getting DMs asking why they're not raising at the same number as the guy two doors down.

GrantWho's doing a completely different business.

MilesDoesn't matter. The comparison culture doesn't care what the business is. It just cares what the number is.

GrantThere's actually a piece from earlier in the year, a Substack called Lobster Capital, covering the batch before this one, the Winter cohort.

MilesDifferent batch, different quarter.

GrantTotally separate outlier, and the headline on it was basically about investors paying double for that batch's own standout round.

MilesWait, so this isn't even the first time a YC batch had one number warp the room?

GrantThat's my point. It happened months earlier. Different founders, different sector, same mechanism. One deal breaks the ceiling and suddenly double is just the going rate people quote in their heads.

MilesSo it's not a one-off glitch. It's a pattern the batches keep producing.

GrantEvery cycle, somebody sets a new ceiling, and everybody downstream recalibrates against it, whether it applies to them or not.

MilesAnd nobody's benchmarking against their own numbers anymore.

GrantNo, they're benchmarking against the headline.

MilesWhich is a terrible way to run a fundraise.

GrantIt's a terrible way to run anything. I watched traders blow up accounts doing exactly this.

MilesIt's wild that the fix is so simple though. Just look at your own numbers instead of the group chat.

GrantSimple to say. Much harder when the group chat is the only room you've got.

MilesOkay. Okay, we've danced around it enough. Let's say who this actually is. It's Nine Mothers.

GrantNine Mothers.

MilesYeah. Counter-drone defense company founded back in twenty twenty-four. YC's own company page has them listed. Three founders, Roman Khomenko, Bogdan Pyzh, and Russell Smith.

GrantWait, counter-drone, like actual hardware shooting stuff out of the sky?

MilesDetection and takedown systems based out of Austin. Nineteen employees.

GrantNineteen people carrying a two hundred... carrying that number on their shoulders.

MilesCareful.

GrantI mean, nineteen people is a small team for something this size in the headlines.

MilesIt is, but that's sort of the pattern in defense tech. Small technical teams, big contracts, since the customer is buying capability, not headcount.

GrantRight. Right. No relitigating. So who's buying counter-drone tech from a nineteen-person company?

MilesThe government. That's the whole thing. Inc. reported they've already got Department of War sales traction.

GrantGovernment contracts. Not a Series A lead investor, not a growth fund, an actual purchase order from the Pentagon.

MilesAnd Inc. reported something else that explains a lot. Garry Tan personally advised this company.

GrantThe YC CEO.

MilesYeah. He said on the TBPN podcast that YC's doubling down on defense tech right now, partly because there are former tech executives sitting inside Government agencies these days, and they're a lot more comfortable buying from a startup than the old guard was.

GrantSo the buyer side changed. That's... Okay, that's actually the real story, isn't it? It's not that founders got smarter at pitching defense contracts. It's that the people signing the checks used to work at companies like theirs.

MilesExactly. And in that same piece, Russell Smith says his inspiration was Palmer Luckey, Anduril. Of course. I mean, it tracks. Luckey built a company that sells missile interceptors and border towers to the Government at defense industry margins, not consumer app margins.

GrantSo when this batch's founders are staring at the two hundred number and thinking, "Why is my seed a fifth of that?"

MilesThey're comparing themselves to a company that isn't playing the same game at all.

GrantIt's not venture math, it's defense procurement math. Government contracts, compliance requirements, sales cycles that run through agencies instead of app stores.

MilesAnd that's before you even get to the timeline difference. A government sales cycle doesn't move at startup speed, whatever the ultimate check size ends up being.

GrantRight. Procurement doesn't care that you raised a seed round six months ago. It moves at its own pace regardless of how hot your cap table looks.

MilesCompliance alone probably eats months that a normal seed startup would spend just building product.

GrantWhich is almost funny because from the outside, it looks like defense tech is just printing money overnight, when actually it's the opposite. Long, grinding sales cycles that happen to end in enormous numbers when they land.

MilesA consumer op founder raising three million doesn't have a Department of War contact. Nine Mothers does.

GrantSo the comp was broken from the start.

MilesStructurally, yeah. You can't put a fintech seed and a counter-drone company with government sales on the same spreadsheet and expect the numbers to mean anything.

GrantWhich is almost worse, right? Because it means the panic wasn't even really about a comparable company. It was about a headline that happened to be in the same batch.

MilesSame Slack channel, completely different business.

GrantSo if that comp doesn't apply to you, what does? Let's look at the batch, not the headline.

MilesOkay, so forget the outlier for a second. Somebody did an actual breakdown of this whole class, not just the one company. Yeah, I saw that. It's covering the whole W26 batch, right, before this one. About two hundred companies total.

GrantHeavy B2B, heavy hard tech. Not what people picture when they think seed stage social app.

MilesThat tilt matters too. Hard tech means robotics, energy, defense-adjacent stuff, not just another SaaS dashboard.

GrantWhich is a different fundraising environment altogether. Different diligence, different check sizes, different timelines.

MilesRight. So if you're a founder building a normal product and everyone around you is hard tech, you're already comparing apples to satellites.

GrantSatellites. Sure.

MilesBut that's exactly why batch-wide framing matters more than any single valuation, whether it's an outlier or the batch's own strong average.

GrantWalk me through what evaluated actually means here, because Rebel Fund isn't YC. This is an outside firm doing their own scoring.

MilesRight. It's their own system, built off years of watching YC companies play out. So it's not official, but it's consistent across every batch they've looked at.

GrantConsistent is the word I care about. A headline valuation isn't consistent with anything. It's one negotiation, one Tuesday afternoon.

MilesRight. Right. And buried in there is a Rebel Fund number that I think is actually more useful than the headline number everyone's still talking about.

GrantGo.

MilesRebel Fund said thirty-five percent of that batch scored in the top twenty percent of every YC company they've ever evaluated.

GrantWait, top twenty percent of everything? Like going back to the earliest batches?

MilesTheir whole evaluated history, yeah. So roughly one in three companies in that batch is outperforming eighty percent of everyone who's ever come through the program.

GrantThat's a batch-wide signal. That's not one company getting bid up because of who backs it.

MilesExactly. And it's from an earlier cohort, not even the one we're talking about tonight, which tells you this isn't a one-time fluke. It's showing up class after class.

GrantSo if I'm a founder freaking out about a headline valuation, is this what I should actually be checking, where my company sits against Rebel Fund's grading instead of whether the outlier closed?

MilesThat's a real number you can act on. A single valuation is one data point about one negotiation between one founder and one investor.

GrantWhereas thirty-five percent scoring top tier, that's talking about hundreds of companies evaluated the same way.

MilesSame methodology, same yardstick, no single negotiation skewing it.

GrantThousands, though. That batch is only two hundred companies. You're talking about the whole history of YC evaluations.

MilesRight, the accumulated history. Yeah, every batch since the beginning scored the same way.

GrantSo on Demo Day night, everyone staring at one number that's basically gossip when the batch already has a real signal sitting in that data.

MilesAnd that signal doesn't come from one hot round or one lucky introduction. It comes from evaluating founders across dozens of batches.

GrantOkay, but even if you're in that thirty-five percent-

MilesEven that's still just a grade at the moment of the pitch. Even top twenty percent at Demo Day doesn't tell you who's actually still standing, still growing, still winning five years from now.

GrantSo what predicts that?

MilesSo if the top twenty percent grade doesn't tell you who wins in five years, what does?

GrantThere's a piece out there that actually tracked this, two hundred and seventy-three YC companies, twenty years of outcomes.

MilesAnd? The R-squared between Demo Day buzz and where a company ends up is point zero nine.

GrantWait, under ten percent?

MilesUnder ten percent, meaning Demo Day noise barely predicts anything past the pitch.

GrantGive me a name. Who slipped through quiet?

MilesAirbnb, DoorDash, Coinbase, Reddit. None of them were the loudest company in their batch.

GrantHuh. So the founder everyone's applauding on stage isn't the one you'd bet on.

MilesNot reliably, no.

GrantThat's gotta be a hard thing to sit with if you're the quiet company in the batch nobody's talking about.

MilesProbably, but apparently it's the better seat to be in, statistically speaking.

GrantBetter odds, worse Twitter mentions.

MilesI'll take better odds every time.

GrantI've got an old one that fits right into this. Michael Seibel talks about his own round. This is back when he was running Justin.tv before it became Twitch.

MilesRight, two thousand and seven.

GrantYeah. He raised a hundred and fifty grand at a three and a half million dollar valuation. Three angel meetings, done.

MilesThree meetings.

GrantCompare that to today. Typical batch company's landing six to fifteen million on its cap, and that's before anyone's heard of a two hundred million dollar outlier two doors down.

MilesSo the whole scale shifted, but the thing that actually mattered, whether the company built something people needed, that part didn't move at all.

GrantNope. Same test as always been.

MilesIt's almost reassuring, honestly. The test didn't get harder, it just got more expensive to enter.

GrantWhich is its own problem, but it's a different problem than the one everyone's panicking about tonight.

MilesWhich is kind of the whole point of this episode. Everybody in that group chat is staring at one company's number like it's a scoreboard.

GrantAnd it isn't. It's one data point from one industry that doesn't even sell to the same customers as ninety-five percent of the batch.

MilesMeanwhile, the actual predictor, whether people use the thing, whether revenue holds up eighteen months out, nobody's arguing about that on Bookface.

GrantBecause it's boring. Nobody screenshots a retention curve.

MilesYou'd think somebody would try just for the novelty of it.

GrantSomebody should. It'll probably get about four likes and a shrug.

MilesNo, they screenshot the valuation.

GrantRight, but if you're one of those founders raising a normal three or four million right now, the number worth carrying isn't the two hundred million.

MilesIt's whether you'd still be standing in five years if nobody had ever heard your name at Demo Day.

GrantWhich is a good place to leave founders sitting with this one.

MilesYeah. Look, the number that gets whispered around the batch, it's not your number. Benchmark against what your own cohort is actually building, not whoever's raising from the Department of War this month.

GrantYeah. Track your retention, your revenue, your customers actually calling you back. Five years out, that's what shows up.

MilesOkay, that's the note. Good one.

GrantI mean it, though. Every batch has its own version of this number, some outlier that makes everyone else's round feel small for about a week, and then everyone moves on and builds anyway.

MilesThat's the actual through line of the whole episode, isn't it? The headline changes every batch, the advice never does.

GrantPretty much. Nobody remembers the valuation five years later. They remember whether the thing worked.

MilesIf you're a first-year founder with a story like this, a decision that scared you, send it our way. We wanna hear it.

GrantYeah, we mean that.

MilesAnd if you liked this episode, leave us a review. It actually helps people find the show.

GrantSee you next time.