This is Dalton plus Michael. Today we're going to talk about some non-obvious insights we've had on how you actually build a very large company. I'd like to explore how the advice we give pre–product-market fit is slightly different from the advice we give post–product-market fit. Sometimes founders don't get this right.

To start with a disclaimer: this is a nuanced topic. This is a fantastic example of something I'm happy to speak about on video, but would never in a million years put on X or social media — nuance doesn't work there. Even writing blog posts, it had to be very long. This is super nuanced. The nuance here: pre-PMF, you just need an idea and you just need to make something that people want. So much of the advice is to stay out of Midwitland. So much of the advice is to quiet your mind from all the distractions and super fancy strategizing you think you're doing. Just go do the thing. That's how I would summarize lots of the pre-PMF advice: just go do the thing.

Less strategy, more help.

More just do things. Right? And those people win. Again, here comes the nuance.

Uh-oh.

What we'll totally acknowledge is that post-PMF, if all you do is that same pre-PMF strategy, you can end up at a local maximum.

Yes. More common than I ever thought. I think that there's this false belief: if I can make $10 million, I can make $100 million, I can make $500 million. Actually it often starts at $1 million. The thing that makes me $1 million can make me $1 billion. What's unfortunate is how easy it is to verify that's not true. You can look at a lot of software companies and ask: what did they do to make their first million? And is that how they're making all their money now? For just as many examples as you'd find in the positive, you'd find in the negative. Google monetizing search — monetizing search. Microsoft — completely different.

Yeah.

Completely different. Let's get into the nuance here.

I'd set this up using the Facebook example. I like to use them because everyone now knows it's a big company, but we remember when it wasn't. At the time, when it was a small startup, it had the best social network for college students. It basically had a monopoly for people with a .edu email address. At the time, you had to be a student at one of the universities they supported — otherwise you couldn't use it. Part of what made it cool was there were no parents on it, no kids on it.

It was a feature.

It was a feature that was super unique. It also restricted you to only your university. They built a super sticky product, but at the time the conventional wisdom was: this is some college-kid niche social network.

How can that be...

How can that be a big business? It's by its very definition niche. If we look back in time, the strategic decision that was not hill climbing — that was not hitting a local maximum — was to first allow high school students to come in.

No, first it was the companies that a lot of the Ivy League kids went to work at.

That was before high school kids.

That was before high school kids. Yeah. It was Accenture.

Oh, sure.

Yeah. I remember some of the first companies that came up. Yeah.

So Accenture was in.

I think Accenture got in before some high school colleges.

Departure from the original vision. Then high school kids — that was a really big departure. And then the decision to allow all people to sign up: I would argue that's a great example of a big strategic bet that was not related to just hill climbing what they were doing. Then there was the big strategic bet to do mobile — betting the farm on the mobile app.

Because all the money they were making was on web at the time.

Right. Then there was the Facebook app store strategic bet. There was the Facebook phone strategic bet, which didn't work out. There's...

Instagram, which is basically the company now.

Yes, it saved the company. With hindsight, you can see all these times where the founder had to make a decision that was not hill climbing — not making the product better. It was actually making a crazy bet and betting the company on something. With hindsight, those ones worked out. But remember back in 2005, 2006: there were a bunch of VC-funded social networks — hi5, Tagged. For whatever reason, those companies didn't make these types of moves, or to the extent they made strategic moves, those moves did not work out.

A takeaway from this story: if you're going to build a very large and successful social network, probably the best way to monetize it is through advertising. You have to figure out how to be huge. There are a lot of other network effects that come from being huge. But if we just look at the business perspective, the value to an advertiser of everyone being in one place — and the amount of premium you can charge — is massive. You don't get that value if you're the seventh-place player.

I'm sure there are a lot of other product reasons why this whole path made a lot of sense, but the business reason effectively enabled Facebook to make the revenue it does. Sometimes when we're talking to startups, I get the impression they care far more about what they believe their next fundraising milestone is than what number they'd have to hit to IPO, or to be a $50 billion valuation company in the public market, or a $100 billion company. Sometimes you're inadvertently making choices that reduce your chances of actually building a valuable company.

One of the areas we see this a lot: companies that have to bundle. Look at HR tech. Whatever reason — why would you want to buy 17 pieces of HR software? The market wants bundles. We see companies all the time that build a best-in-class product and then struggle with: do I do more? Do I bundle? We're seeing this for bigger players now. If you're Zoom, if you're Slack — you had these decisions. We got to a certain size. Do we build the bundle? If we don't, can we really slow the next Salesforce? So it's part of a bundle.

These are conversations we never want to have with a pre–product-market fit company.

Yeah. Again, let's use a specific example. Let's say someone wants to build a new Cursor competitor. Pre-PMF, it's: have you talked to your users? Does anyone want this? Is this better than Cursor? It's just the fundamentals. Any midwit strategizing about whatever is irrelevant if you have no users.

Your product is worse if you're sad.

If no one will use it, you have a problem. If Claude Code is just way better...

Game over.

Game over.

Yes.

But if you're sitting here right now watching this video and you've raised a bunch of money or a Series A, and you're the seventh most popular Cursor or Claude Code competitor, and you've done no strategic thinking about what the plan is — I think you're going to have a bad time.

Yes. There's another misconception that leads people to believe being the seventh-best player is okay: "Well, if the winner is generating $500 million in revenue and we're generating $50 million, we're just worth one-tenth of the winner." I don't think I've ever seen this thought backed up. I think the assumption is you're going to trend toward zero revenue. I guess there are some markets where there are six winners, but 😬. So what's interesting in your example is this: if you're the seventh-place player, you must be obsessed with winning the whole game.

Or just having to move. There's got to be a move.

Or playing a different game — one we can win. We encounter a lot of YC founders and founders in general who say: "But isn't $50 million revenue sufficient? Haven't I won? Isn't the finish line behind me? I got here. If I got to 50, it's a guarantee I'm going to get to 500."

One of the things I say to a lot of founders: what do you think is the failure rate for a Series B company? I think in their minds before I ask the question, they think every Series B company wins. After I ask it, they're like, "I guess mathematically that wouldn't work." When they hear stuff like 80%, 90%, they're like, "Oh."

What's funny about this nuance point we're making: I could see a lot of people going, "Yeah, totally guys, I agree with you. But why aren't you doing all that during the batch?" I don't think people recognize how this style of thinking is literally poison when you're trying to just ship a product — but it's absolutely necessary.

I'd attack it this way. Most of the time in pre–product-market fit, you're still trying to learn what the customer wants. I would almost argue you're still in the fundamental "let me learn what the hell's going on" phase. When you're trying to make these high strategic moves, it's very helpful that you know some stuff. It's very helpful that you have a mental model of the customer, of their needs, of why your product helps them or not. And these are things you're still figuring out pre–product-market fit. This kind of thinking creates paralysis. Instead of writing code today and talking to users, you spend eight hours talking about strategy.

To go back to the Facebook example: all of the strategic work Zuckerberg did was super helpful — for a guy with a hundred million MAU. Cool, you've got the keys. You get to spend a lot of time pontificating. Some guy sitting in his apartment strategizing about the future of social networking — that's not someone in a good position to do this type of analysis.

When Zuck's got 2,000 Harvard kids — that's not the time. I love the point you made at the beginning: this gives us a little bit more time to flesh these ideas out.

Another idea you introduced to me — a basic finance idea I didn't fully embrace at first — is comps. The number of times I'm talking to founders and I'm like, "What company do you comp yourself to?" Because if you're trying to think strategically, what does good look like? Where are we aiming for? The number of times people will say, "Oh, this company has a billion dollar valuation. I'm comping myself to them." I always laugh now. You have a sea of public companies you could use. It's like: I don't want to learn anything about playing basketball from Michael Jordan because I like this guy who plays on Duke right now. Okay, the Duke guy's great. Maybe he'll make the NBA. Can we talk about Michael Jordan for a second? "Oh, but Michael Jordan is old." But he won. Pick a newer winner. Let's talk about a company that's hit $100 billion. Can we just talk about someone who's really put points on the board — and we're not judging them on what they might accomplish in the future? Founders resist this. Those comps are cool.

In my experience, going public seemed so foreign. I didn't know anyone that did. It felt like talking about an astronaut who walked on the moon. It just didn't seem real. Now it seems completely real — but we're in the bubble. I think that's my steelman on why people resist.

I hear that. We've gotten to do a number of YC talks with Tony from DoorDash. I think he was thinking earlier than anyone ever believed: I want to do delivery better than Amazon. Before he ever had the right to think that way, I think he was thinking that way. That's just a choice. He could have only comped himself to Uber and Postmates. You choose who your heroes are. It's more exciting — scarier for sure, but more exciting to compete against the big, big players.

All right, last dangerous topic. I love talking about this. During early-stage fundraising, we often make fun of investors who say, "What are you going to do when Google builds your thing?" It's often a sign of an unsophisticated investor.

I think it's because we heard that so much when we were founders.

There's no right answer. If Mark decides to directly compete with you and ignore everything else, how would you win? It almost feels like an unfair question. But there is a stage of company where the Googles, the Amazons, the Salesforces, the Microsofts will take interest.

I heard a great story from a YC company doing a CRM. They said, "Have you ever heard of Microsoft Dynamics?" I was like, "I don't know the full range of Microsoft products." They said, "Oh yeah, it's their CRM that's taking customers out of Salesforce's hands left and right." I'm sitting here thinking: how many of our founders think Salesforce is number one? And Microsoft's like, "Oh, this is the time. In the last five years, let's start. We could do that."

Or a classic example: AWS. For damn near a decade they were the only serious player. Then Google and Microsoft said, "Oh yeah, we could do that." The crazy thing: they can hire some of the people who built your thing and ask them, "Hey, now that you've built it, how would you do it better the next time?" So what seems like a dumb question from an early-stage founder — how should a later-stage founder think about these big guys? — because they do compete.

I think it's worth thinking about from your customer's perspective. Go through the laundry list. How hard would it be for someone to switch? Are there non-obvious reasons they wouldn't want to switch? Being really honest with yourself about those things is probably a good idea. In the case of Slack, from what I've heard — and this is not first-party, just third-party — a lot more people were willing to use Teams and other stuff than Slack expected. So it wasn't as sticky as they had hoped, which is part of why they needed to get acquired. I believe Microsoft actually did damage to Slack, and that's why they got acquired. That's my understanding.

I talked to somebody who was a product lead at Microsoft. He said, "Look, we could bundle Teams. We could effectively almost default roll it out to all Microsoft folks. For every one of those deals, Zoom would have to get people to adopt it, upsell them, figure out some IT thing, do all the contracting. We're talking: this is the next part of the suite."

A question that is completely irrelevant to most YC founders for their whole career: if you're a product like Slack with the amount of takeoff you had, or Zoom with the amount of takeoff you had — was there a way to compete against Office? One of the things I often think about: there is a point where you might actually have to go at these big guys' throats.

One more time, back to the Facebook thing. Google decided to kill Facebook with Google Plus. They put half the company on it. It totally failed — but not because they didn't try.

No, no.

It was a very serious effort by Google to bury Facebook. It just didn't work out.

I think about the ocean and different-sized fish. When you're a fish this small, the great white doesn't care. But as you get bigger, you've got bigger competitors. I almost wish more founders thought of that as empowering: "Hey, now I get to compete with the big guys." It's not like the big guys have the better product. Often, by competing with the big guys, you can bring a better product to market. I'll use DoorDash as an example. I've been an Amazon user for 25 years. I would have said Amazon was top five in my life for maybe 20 of those years. Something started happening about five years ago where I would do an Amazon search.

I know. It's hard not to buy junk. You have to really try not to buy trash on Amazon. The whole thing is set up to buy crap.

With the route button, yes.

You almost need to discover outside of Amazon, deep link to the thing you want to buy. Otherwise, a lot of it is bad.

I'm sure some PM is gonna get sent this video. The email from the boss: "Have you seen this? Please consider, you know, investor guys talking about it."

The other thing is the sponsored products. Google taught me that sponsored stuff's at the top — scroll past it and then you get the real stuff for a while, and you should be able to tell the difference between ads and non-ads. Amazon's taught me I don't even know what's going on. Is that an ad? Is that not an ad? Are those reviews fake? Are they real? I have to have a plug-in to tell me whether the reviews are real or not.

On the other hand, DoorDash — I can't even believe I'm saying this — DoorDash fulfills from local stores. It's very hard for random spam-shipping products to get distribution in local physical stores. When I order on DoorDash, I feel safer.

You get the thing.

Yeah. It's a brand I've probably heard of. It's going to work. It gets there faster. I feel such a rich douchebag sitting here — I don't need these sneakers in an hour. I don't. But I'm not even using DoorDash because I need them in an hour. I use it because I trust it. I feel uncomfortable. I got all the Amazon stuff. I got Echo throughout my whole house. But it's very interesting: DoorDash is competing against Amazon. They've decided we're going to take on the big boys. The big boys maybe are more vulnerable than they look.

What's funny to cap us off: that was always the plan. The comps — you said this a minute ago — but the plan was always to imagine the size. The logistics business for all products was bigger than food only.

But first — you know what you have to do? You had to get someone a fucking burrito in Palo Alto. You don't get to have this conversation without: how do I get someone a burrito in Palo Alto?

So use this advice correctly. Don't come back and yell at us. We don't want to hear from your pre-launch company: "How do I take on Google?" But if you do find yourself in a bit of a rut in your post–product-market fit — maybe you've got to be more ambitious. Which is counterintuitive.