The SAFE Trap: What a W26 Founder Signed at 2am

Miles: Two in the morning, Parking lot, and she's staring at a term sheet on her phone thinking, okay, $40 million cap, we did it. Grant: And she signed it. Miles: She signed it. Welcome to Year One. I'm Miles. Grant: I'm Grant, and that founder, Six months later, she's sitting across from a Series A lead who wants 18 to 20%, and the math just doesn't work anymore. Miles: That's wild, Right? Because on Demo Day, everything felt like momentum. According to Lobster Caps, 14 W26 companies walked into Demo Day at a million ARR or more, three times the prior batch. Grant: So the room is electric. Investors are moving fast, and that speed is exactly where the trap is. Miles: Right, so today we follow one W26 founder through the decision that nearly broke her cap table. We're going to walk through the actual dilution math, three stacked post-money SAFEs quietly locking up thirty percent before a Series A even starts. Grant: And then we get into the MFN clause, the one buried in YC standard deal. One cap concession to close a reluctant investor. And it cascades in ways she never modeled. Miles: Here's the thing—she's not alone. A YC partner is going to break down the two flavors of cap mistakes they see again and again post Demo Day. Grant: And we end where she is now, cap table on the table, founders at sixty three percent, a Series A lead circling, still unresolved. Miles: No clean ending. Not yet. Let's get into it. We're just going to let her talk for a second. Grant: Yeah, don't touch anything. Miles: Walk us back to what you actually thought you were signing. That's where we're starting today. A W26 founder fresh off Demo Day staring at a number that felt enormous and turned out to be a very specific kind of math she hadn't run yet. Grant: Right. And with context here matters. According to multiple sources covering the W26 batch, the standard round coming out of Demo Day. was $4 million on a $40 million post-money valuation. That's the default. One analyst noted that three years ago in W23, that same default was $2 million on $20 million. The entry price basically doubled in Miles: In three years. Grant: three years. So the number in front of her is not unusual. It's the going rate, which, you know, makes it feel legitimate. Miles: Here's the thing, though. A $40 million cap sounds like a ceiling. revealing what it actually is is a denominator and that's a very different conversation yeah Grant: nodding along and that's what she didn't know yet she heard 40 million and thought valuation she didn't think dilution and Miles: and to be fair most first-time founders don't i mean come on you just pitched 200 people in a room over two days someone hands you a term sheet at midnight you're running on fumes and adrenaline Grant: a parking lot Miles: And a parking lot, of course you sign it. Grant: The pressure at Demo Day is real. According to sources tracking the W26 batch, competitive seed rounds were closing within hours of companies finishing their pitches. Investors were moving fast to lock in spots before someone else did. Miles: So the whole environment is designed to create urgency, which is not necessarily sinister. That's just how price discovery works when you've got a room full of capital chasing a limited number of spots. rough spots. But for a first-time founder, that pressure is exactly where the misreading happens. Grant: And the misreading isn't about the number being wrong. It's about not knowing what the number actually means for the cap table. Miles: Which is what she's about to explain. So the question I keep coming back to before we get into the mechanics is when did she realize the math wasn't what she thought? Okay, so the founder signed at $40 million. Now let's get into what that actually meant on paper. Grant, walk me through the basic math. Grant: Right. Post-money SAFE arithmetic is almost insultingly simple. Investment divided by cap equals ownership. $500,000 on a $10 million cap, that's exactly 5% locked in at signing. Miles: And not simplicity is where people get tricked. Grant: Totally. Miles: Mm-hmm. Grant: According to SheetVentures 2026 SAFE cap analysis. The post money cap only SAFE is now the market standard, no ambiguity, no renegotiation, you sign, you're done. Miles: So back to our founder-they didn't sign one SAFE at Demo Day. Grant: No, they signed three. Miles: Three separate SAFEs. Grant: Three different caps, different investors, same general window. And here's where I always push founders: did you run the stacking math before the third check cleared? Miles: And the answer is? Grant: I think you know the answer. Miles: So let's do the scenario. Say you've got three SAFEs totaling one and a half million dollars against a five million dollar post-money cap. Walk us through what that looks like. Grant: Each SAFE independently locks in its slice. Five hundred thousand on a five million cap is ten percent. Three of those? Thirty percent committed before Series A even starts. Miles: Thirty percent gone before a single Series A term sheet. Grant: And that's before the option pool expansion your Series A lead is going to demand." The Startup Law Blog put it bluntly: "Startup lawyers see founders give away forty percent of their company pre Series A because they didn't model the stacking. Miles: That's wild, right? The math is so simple individually- five hundred divided by five million-but nobody actually adds them up. Grant: Right, and the second question I'd ask this founder, the one that nobody asks at two in the morning in a parking lot, Speaker 3: is: Grant: Is who named the cat first? Miles: Oh, that's the one! Did you walk in with a number or did the investor? Grant: Because if the investor named the number and you just nodded, you didn't negotiate, you capitulated. Miles: Did you open a spreadsheet before you signed? Grant: That's the only question that matters. Sheed Ventures' guidance is direct on this: CAPS producing more than twenty five to thirty percent dilution before a priced round are considered punishing; most founders don't discover they've crossed that line Miles: Wow! Grant: until they're serious. Series A lawyer emails them a cap table summary. Miles: And at that point, it's done-you can't unsign the SAFEs. Grant: Locked in, which is why the next piece of this story matters a lot, because buried in at least one of those documents was a clause the founder barely registered-the MFN, the most favored nation provision. And that's where the math gets genuinely weird- YC's own standard deal, per their published terms? Split five hundred thousand into two instruments-a post-money SAFE for seven per cent., and an uncapped MFN SAFE for the rest. Miles: That MFN clause can reach back and touch every deal you sign after it. Grant: So you could sign a lower cap later to close a tough investor, and suddenly your earlier MFN holders get to convert at that lower cap too. Miles: Cascading dilution, higher than anything you originally modeled. And that's exactly what we need to pull apart next. So here's what I want to push on. Before we move forward I want to go back to the moment you actually signed the YC docs-the most favored nation SAFE-the three hundred seventy five K piece, uncapped. Had you ever heard the phrase "most favored nation" before that day? Grant: No-just no Miles: I mean, I skimmed it. My lawyer sent a summary. It said 'uncapped, most favored nation provision, standard YC terms.' I thought 'standard means fine,' right? Grant: Right; and that's the thing, because "standard" is doing a lot of heavy lifting in that sense. The YC deal is well documented; according to YC's own published terms, that three hundred and seventy five K SAFE converts on the terms of the lowest cap SAFE you sign between the. Between a batch start date and your priced round-so every SAFE you close after YC, that most favored nation SAFE is watching. Miles: And I signed three of them after batch. Grant: Yeah, so Walk me through the last one-the one where you drop the Cap to Close a tough investor. Miles: We needed to Close. The guy had been dragging for six weeks. I came down from Fifteen million to Twelve million. He signed the same day. Grant: And the moment that Twelve million Cap hit the Cap table... table, why YC's Three hundred seventy five K most Favored Nation SAFE could elect to convert at Twelve million, not Fifteen million, not Forty million-Twelve. Miles: I didn't model that. Grant: Nobody does the first time. Kruze Consulting actually flagged this pattern in a March, two thousand twenty six piece calling it "hidden dilution creep." You plan for one Cap, you make one concession to Close one investor, and every most Favored Nation SAFE follows that lower number down. The cascade isn't dramatic in any single moment, it just lands when you convert. Miles: Wait, so what should I have done? Walked away from the investor? Grant: Or structured it differently-non economic terms, a longer pay in schedule, something that doesn't touch the cap, because the cap is the number every most favored nation safe is anchored to. Miles: I didn't even know to ask that question. Grant: That's the part that gets me. Startup Counsel, experienced folks, have said on the record that 48-hour expiration pressure on term sheets is largely psychological, not a hard legal deadline. But first-time founders treat it like a bomb with a timer, so you skip the review that would catch exactly this. Miles: Completely. I had forty-eight hours, felt like four, you know what I mean? Nobody was going back to the docks. Grant: And that's where the most favored nation clause lives-in the space between forty eight hours and the legal review that didn't happen. Miles: I could have used more time and less Parking lot energy. Grant: Yeah-now flip that on its head. What does this look like from someone who's watched a few hundred founders sign the same docs? That's the conversation we need to have next. Miles: So that's the MFN cascade. Now I want to bring in a voice who has watched this exact story play out hundreds of times. Grant: And I mean hundreds. We sat down with a young Carta partner, and honestly, within 60 seconds, they were finishing our sentences. Miles: Yeah, they'd heard it. Grant: So I asked them straight up, when founders come out of demo day and they're setting their SAFE caps, what's the mistake you keep seeing? Miles: And they didn't pause. Grant: Not even a little. They said there were two flavors-founder who names a number that's too low because they're terrified of not closing, Miles: And the founder who accepts a number that's too high because they're afraid of looking like they don't believe in their own company. Wait, that's the whole thing, right? Both moves come from fear. Grant: Exactly-one is fear of rejection, the other is fear of looking weak. Neither one is actually about the math. Miles: And the math is the part that matters downstream. Grant: So I pushed on that. on that. I said, OK, what does the data actually say about where founders end up when they set the cap wrong? And they pointed me at something VC Cafe wrote back in February, citing Carta. Graduation rate from C to A1 has improved from 17 percent to 30 percent over the last couple of years. Miles: That sounds like good news. Grant: It is until you hear the rest. The revenue benchmarks to actually clear A1 are higher than they've ever been. than they've ever been. So you're graduating more, but the bar you have to clear is way steeper. Miles: So setting an aggressive cap today creates direct down round risk at A1. Grant: That's what the partner said, and they framed it really clearly: according to Eqvista's 2026 fundraising report, AI startups are commanding roughly a 42% valuation premium over non-AI peers, median pre-money around 7... seventeen point nine million dollars Miles: Which sounds great. Grant: until you raise above that median because now you've set a milestone clock you didn't mean to set. Miles: So founders are using the cap as a confidence signal. Grant: That's the pattern. The partner said it so plainly, founders treat the SAFE cap as a signal of how much they believe in the company, not as a financial instrument with real downstream consequences. Miles: I mean, come on, those are two completely different things. Grant: They are; and no one tells first timers; you walk out of Demo Day, someone's offering you a number, and you're thinking about what the number says about you, not what it does to your cap table in eighteen months. Miles: So what do the founders who get it right actually do differently? Grant: The partner's answer was almost boring. They model the Series A before they sign the seed. They ask, "If a lead wants fifteen to twenty percent at Series A, Speaker 3: what will it take to get them there? What will it take to get them there? Grant: What's left for everyone else? Then they work backward. Miles: Right, the cap is a denominator, not a trophy. Grant: Which brings us back to this founder, because all of that math, those graduation rates, that AI premium, that downstream pressure, Miles: Yeah. Grant: it's all theoretical until you see what it actually looks like on one real cap table. Miles: And we have exactly that—what the dust settled into. So where does this founder actually stand right now, six months out from Demo Day? What does the cap table look like? Okay, so here's the honest picture. Why is the own 7% on that post money SAFE? The seed investors, three of them, are sitting on roughly 22% combined once those SAFEs convert, and the founders are at about 63%. Which sounds okay, right? Until, until a Series A lead walks in wanting eight... 18 to 20 percent, then you run the math. And there's barely room. Barely room. You need to carve out a fresh option pool, probably 10 percent before the Series A closes. That comes out of the founder's first. So we're talking mid 40s on founder ownership by close. That's survivable, but it's thin for a first time founder, you know, that's the number that has to carry you to a B/C exit. Right. And according to Eqvista, the median Series A for an AI company right now has pre-money somewhere in the 40 to 50. 50 million dollar range. So the dilution math works on paper if you hit the milestones. If you miss? Downround, and the MFN clause bites again. Exactly. That's the thing nobody says out loud. The MFN risk doesn't disappear after the seed closes. So I asked the founder, straight up, what's the one conversation you wish you'd had before you signed? What did they say? They said they wish they'd called a lawyer about the MFN SAFE sp- If specifically, not a general startup attorney, someone who had actually read those docs on 50 Y Batch companies. Because the YC standard deal is not generic, it has specific mechanics, Grant: Mm-hmm. Miles: and they just didn't know what they didn't know. That's wild, right? Because the document's public. Anyone can read the YC SAFE templates online, but reading it and understanding how it interacts with your other SAFEs are two completely different things. They also said they wish they'd had one honest conversation with another W26 founder before Demo Day. Just, what cap did you set and why? Peer Intel, the stuff you can't get from a blog post. VC Cafe actually wrote about this back in February. Fewer startups are getting funded in 2026, but the ones that do are raising on tighter instruments with bigger checks, which means if your cap table's messy going in, Speaker 4: Right. Miles: you have less runway to fix it before a Series A lead asks the hard questions. So is the Series A conversation happening for this founder right now, or... Speaker 5: Still unfolding. They've had two intro calls-nothing on paper yet. Miles: So we'll leave it there. Speaker 5: The founder's last line to me-and I want to keep it exactly as they said it-'I understood the cap after I signed it; that's not the order it should happen in. Miles: So, um, that parking lot story. Signing at two a m thinking you'd made it? Speaker 5: Yeah; and six months later staring at a cap table that's basically already spoken for before Series A shows up. Miles: That's the thing about a post-money SAFE, it sounds like a ceiling, it's actually a denominator. Speaker 5: Right. And the MFN clause? That's the part nobody reads until it costs them. Miles: The entry price on W26 basically doubled. (Pulled from W23, according to Lobster Caps' recap. Four million on a 40M post-money. The pressure to sign fast is real.) Speaker 5: Which is exactly why reading the docs matters more than ever, even the boring clauses—especially those!" (thoughtfully) "If this episode made you slow down before you countersign anything, that's the whole point. Miles: Know a YC founder in Year One who'd tell their story? Reach us at yearone at heymeado.com. And if this helped, leave a review. Speaker 5: Yeah. Miles: Seriously, it matters. Speaker 5: Thanks for listening. We'll see you next week. Miles: Warmly, take care, everyone.