Dalton Caldwell - All About Pivoting

How's everybody doing? I'm Dalton. I'm a partner at Y Combinator. Um, in addition, I'm the head of admissions, um, which is our selection process for the companies that get into YC. I am here to talk about pivoting. Um, yeah, let's talk all about pivoting. Cool. All right, here's some stuff we're going to cover: what the heck is a pivot, why you should pivot, when you should pivot, and evaluating ideas to pivot to. So, we're going to try to cover all the bases here.

All right, let's talk about the term pivot. Um, this is one of those terms that if I'm in a cafe and I hear someone talking about pivoting, I roll my eyes because it's one of those words that I associate with annoying startup people. Um, and so, let's just explain what we mean here. It just means changing your idea. That's all it means.

And technically, if we want to be really technical, I would call a true pivot where it's like a real company, and you have lots of users, and you've raised money, and you're like, "We're gonna shut this thing down and do something different." So, the most famous example is Slack. They raised money and had like a hundred employees for this video game called Glitch. I was a beta user, and then they just shut it down and did something crazy. I would call that a pivot. I think that's a valid use of it. Probably what most of the folks here are doing, I interchangeably call it pivoting, but you should just call it changing your idea.

Um, it should feel really lightweight when you're at the earliest stages of your company, especially pre-launch or very near after launch. Changing your ideas constantly is kind of the norm, and I wouldn't think that this is some huge thing. It should feel lightweight. And frankly, if you're not in a state where you're rapidly changing ideas or assumptions over and over and over again in quick succession, you are likely doing it wrong. You are likely moving too slow. And so, this is just like part of it: change your idea constantly.

Um, trying to find exactly the right version of your idea is something you should be doing in the beginning. Let's talk about why you should pivot. So, the main reason I would argue is opportunity cost. And the definition of opportunity cost is that the loss of potential gain from other alternatives where one alternative is chosen.

So, in other words, you can only really work on one thing at a time. Sometimes people try to violate that rule, but that's a whole different topic. And so, by working on something that's not working, and you have evidence that it's not working, you are taking opportunity costs to not be doing something else. It's as simple as that, right? I don't know, I tried to write some pseudo-code here for you about it; you know, it's kind of a joke, um, but like how well things are working divided by the number of months of concerted full-time effort.

If that number is less than excitement to work on something else plus confidence you can find something better, you should pivot. And so, the key thing to if you like look at this equation, like what am I really trying to say? It's that if you've worked on something for months and months and months and months, and it's not happening, like that's a pretty good signal, right? Like that's what drives this equation that I put here the most is the number of months you worked on something and it's not working.

Um, and so if you're throwing good money after bad, good time after bad, and it's not happening for you, that is a pretty darn good signal. But if it's really, really early, and you've only been working on something a couple of weeks, it's less obvious.

Let's give you some good reasons to pivot: I hate working on it. All right, it's not growing. It's just not working. I'm keeping doing the thing, and nothing is happening. I'm not a good fit to be working on this idea. Like the more you learn about this, the more you realize that you are just not the right person for the idea.

Another one is, I'm relying on an external factor outside of my control to make my startup take off. So, a couple of examples of that are like relying on mainstream virtual reality headset adoption. That's a good one; you know, like any day now, the new Facebook thing's gonna take off, and that's when my VR app's going to take off, or like relying on mainstream crypto adoption; things like that.

Um, those are like totally out of your control, and if you're just sitting here hoping someone else does something good, and then your startup will work, man, you should definitely pivot. Another one is just where you're out of ideas on how to make the thing work. You're just—you've thrown everything against the wall to make your current idea work, and you're actually kind of out of ideas. That's usually a sign you should pivot.

Good reasons to not pivot: Um, you're trying to run away from doing hard work. Sometimes you see people where they build a product, and right when it gets to sales time, they pivot, and they do that over and over again. That's probably not a good reason to pivot; it's just someone trying to dodge the sales part. Um, so watch out for that.

Um, also, another reason not to pivot is if you're the type of person that changes their idea over and over and over and over again, like chronically, um, and you're detecting yourself doing that, it's good to actually follow something through all the way. So, you just want to watch out for that happening.

Also, um, a good reason not to pivot is that, like, you know, you just hear there's some hot new thing, and you want to pivot because you write a TechCrunch article because someone raised money for a hot new thing; that's not a great reason to pivot.

Let me give you some reasons on why people take too long to pivot, and the reason I mention this is I would argue, on average, most people take too long to pivot, right? I think that's usually, like, if I had to put people in buckets, more people take too long than the reverse.

And so why do they take too long? Loss of version; when you feel like you've invested in something, you have a really hard time letting go of it, and that's loss of version. You can research this. Um, have a little bit of traction; like you have a few users or like one customer, you're like, "Maybe it'll work," that will make you not pivot.

Um, and that's rough. Um, another thing is, like, people are very polite in the world, and they have a hard time telling you the truth that they don't like the thing that you're working on. Um, and so a lot of founders get confused about politeness, um, and getting that confused with traction.

And so you're never going to really—I mean, maybe I shouldn't say never—most likely, you're not going to have people being like, "Your idea is horrible; give up now! I will never be a customer!" They're not going to tell you that to your face. They're like, "Oh, this is great; maybe if you add a few more features and come back, you know, we'll take a look at it," right?

Oh, that's dangerous because you can end up doing that iteratively over and over and over again for, like, years and never actually get customers. So watch out for politeness and getting the politeness confused with traction. Another one is a fear of admitting weakness or defeat. If you pivot, you're giving up in some way.

Another one is putting blame on why things aren't working on customers or investors. So this is where you're like, "I'm not wrong; the world is wrong! You know, like, no one gets this; this is like way ahead of its time!" Um, things like that are not a great sign.

Um, and usually people realize that maybe it is kind of on them versus on the external world, but it takes a long time to get there. And then finally, there's a lot of inspirational stuff out there that's just like, "If you just believe hard enough and keep going, you know, you'll eventually be recognized, and everything will be great."

Um, and so there's a lot of those inspirational messages out there, and that can kind of be counterproductive. I don't know, this is super eligible, but basically this is where I talk about the little bit of traction thing. I call it the uncanny valley of product-market fit.

Um, and you know what's weird? I noticed this at YC as a YC partner: someone that gets into YC with an idea that's just a total fail, like immediately, is actually a huge advantage than someone that has like a little bit of traction. Isn't that weird? Like, a total fail? The founders can declare bankruptcy on the idea immediately and just work on other stuff with no regrets and no like second guessing.

Oh, should I pivot? You know, they're like, "Wow, that was really bad!" And having that freedom to immediately throw off the old idea and work on a new one is actually a weird advantage. Isn't that counterintuitive? And so you just want to really watch out for a little bit of traction because I've seen that be a trap that has captured a lot of really talented people for long periods of time.

And then let's just talk about the anecdotes. Um, so anecdotes about, you know, stories of people that just kept doing what they were doing, and it didn't work, and then five years later, you know, it was great.

I think those are cool, and they're inspirational, and I like them, and they're all true, you know, they're anecdotes. But it's kind of like anecdotes about people that, you know, played the lottery every day for like five years, and then they won the lottery, and now they're like really happy. And you know, that's super cool, but that is not actionable for you, right?

There's nothing you can do with this anecdote about someone that like just kept believing strong enough. Um, like I would much rather give people advice to play the statistics of this and to take accountability for like their own actions in the world than just like hope and dream that you might be one of the anecdotes too.

And so if you decide to pivot or don't decide to pivot, just remember you decide; it's your life. And if it works out good or bad, that's on you. And so often, I get founders that want to push this decision on me as a YC partner or other authority figures about, "Is this idea working or not? When should I know when to give up?"

And ultimately, all we can do is give you guidance, but this is exactly in the class of problems that is on the founder and never something you should look to an authority figure to tell you how to decide product-market fit.

I'm sure you know we're talking about this a lot. It's been discussed a lot in startup school, and it's because it is so important. Most people will never get it; probably most people watching this don't have it, even if they think they do. Um, and you know you have it when growth is not your biggest problem; it's other stuff.

And one good reason to pivot is you just get more shots on goal to try to get this elusive thing right. Like if you made something and you launched it, and it's like, "Man, not really working," um, a dang good reason to pivot is you get another roll of the dice; you get another shot.

And so, I've just seen people that use these opportunities really well. It's much easier to be lucky when you get like half a dozen shots on goal than one, right? And so just if we're just playing the statistics of how do I get product-market fit, taking several high-quality shots—notice I say high quality—you can't just constantly pivot through stuff, another launch it.

But if you do a full awesome product iteration of making something, completing it, shipping it, giving it to people, following all the advice you're looking here, and you can give yourself multiple of those shots, I would argue you are creating luck for yourself, and the odds that you actually hit something that works are much higher than someone that only ever takes one shot.

Okay, let's talk about how to find a better idea. Um, here's the advice that I give people during the batch when they are looking for an idea to pivot to: um, find something that you're more excited about, and that makes you more optimistic about the world, and more just generally more excited to wake up in the morning and work on the thing and not less excited.

And there's a corollary here, and this, I think, is counterintuitive: often, choosing what is perceived as a harder idea is more, uh, is good. And so I see a lot of people where they're like working on some, like, ad network, ad tech thing to like affiliate something, ad targeting. I don't know; those like never work, and it's because they're not inspirational, and no one cares.

And so when you see people go from that into something that's really, really exciting, right? Like, "Oh, I want to help small business owners do X," or I had a company in this last batch, um, and they did, uh, they started with, um, this ad tech thing that they were all very bored with, and they pivoted to Robinhood for India.

And the moment they pivoted to that, the founders, they were like—they lit up! And every conversation I had with them, they were like so excited about their idea, and it was contagious talking to them. Um, and it was almost like a different, like—the amount of change; it was the same founders, but what changed is they found an idea that was real, that they were excited about, and not just some boring ad tech crap no one cares about that stuff.

Okay, so if you're doing something that seems like a good startup idea because you're writing TechCrunch, people raise money for it, but it's not inspiring to you, and you're like, "Man, this is pretty lame," you probably want to find something that's more exciting to you.

Another thing to do on a finding a better idea is make an honest assessment of what you're good and bad at. This is hard, but you want to be really self-aware of what you're good and bad at, and play to those strengths.

And another thing here for finding a good idea is, especially if it's a pivot, find something that you can very quickly build and validate and not something that takes like a year or two of R&D. Right? Often, like, if you pivot from one thing that's impossible to ship into another thing that's impossible to ship, not good. Ideally, you find something that's way easier to ship, like really fast. Highly recommended.

Quick note here, caveat: it's totally okay to work on an idea where you're not going to raise venture capital. All good! Most businesses in the world don't require venture capital. And so, all of your—if what you're doing is consuming lots of content on how to start a company like this and it's all kind of like venture capital-focused stuff, um, and you're not going to do something that raises venture capital, you can kind of get blown off course idea-wise.

And so just be self-aware about this and realize that if you do want to raise venture capital, the idea does matter a lot. And there's a constant, like, recurring theme I see is people that are trying to raise money for something that is definitely not venture capital fundable, and they get really frustrated; everyone gets frustrated with that.

It just—it doesn't make sense for all businesses. And so just be self-aware of this when you're choosing an idea of like, "Is this something that at least hypothetically is VC fundable?"

Just to give you some rule of thumb, what does that mean? Um, I don't think there's a guidebook for what venture scale means, but here's some rules of thumb: can I imagine this business generating hundreds of millions or billions in net revenue per year? That sounds venture capital fundamental.

Um, can I imagine the revenue growth to get to those numbers to happen in like less than 10 years or five years? Like, can it happen really fast? Can I imagine this thing that I'm doing as a publicly traded company someday? Can I picture it? Can I visualize it? Kevin's first lecture talks about this, but these are all things—like if you just can't see any of these, that's not a great sign.

And other key properties is usually technology is a key component, and usually the founders build the technology at least in the early days for something to be VC fundable. Um, you want to see really high margins— not for everything, but again just a rule of thumb: software margins—you know, 80% gross margin, 70% gross margins—really high margins to some of you we wanna see.

And then, um, just it's funny, like I think a lot of people learn about fundraising from Shark Tank, and I don't think a lot of that stuff is venture fundable, just in case you were wondering. Um, it's fundable for people that want to put product stuff together, but I don't know if you know, I think you'd have a hard time raising money for venture capitals for the majority of that stuff.

But hey, like I said earlier, it's your call; it's your dream. So, figure it out. When's the best time to pivot? As soon as these things happen: you've launched and trying to get users for weeks or months, and you feel hopeless. It feels hopeless; you should probably pivot when the idea is impossible to get started with.

Like, "Cool, once we raise 100 million dollars, then we will build a prototype." You should definitely pivot if you're one of those people, because unless you have 100 million dollars, you are in a chicken and egg unsolvable issue.

And here's another one: you know when you're hurt, it's not going to work. A lot of founders know secretly that their thing isn't working, but they want to keep up this front to everyone in the world that it is working. And they think they can fool people into funding them or working with them. If you can't convince yourself and you know this isn't really working, man, is that a good time to pivot, right?

Like, you know more than anyone else about your business; you gotta listen to yourself. And it's sometimes not stuff you want to hear from yourself. Um, okay, let's talk about other epiphany stuff. If you pivot over and over and over again, it causes whiplash. Whiplash is very bad because it causes founders to give up and not want to work on this anymore, and that actually kills the company.

Weirdly, it's more deadly to your company to get whiplash and get sad than to work on a bad idea, because if you're having fun working on a bad idea, you won't give up, and then conceivably, you can maybe make it work. If you get really sad and hate your life while you're working on your startup, you will definitely not succeed, and it's because you will give up.

And so this is weird; like it's kind of better to work on an idea that's not the best one if you're really having fun. Um, and then you just want to be in a happy medium. Some founders pivot way too much, and they'll probably watch this lecture and then like pivot once a day for six weeks—don't do that!

And some people just work on the same idea for five years, and it's not working, and they just are really obstinate about it. Just find the happy medium like everything else in life. Another thing is, it's really hard to have employees and be pivoting, so don't do it. It just makes it slower, and it makes them really sad.

And so, trying to scale up like a team and taking all the advice here about how to scale a team while you don't even know what your idea is or you think you're going to change your idea is definitely not best practices. It only slows you down. I would only add people to the team after you know your idea is working and you have confidence; otherwise, it's just all downside.

Okay, let's go to a different thing here. I made this up for this lecture, and I came up with a subjective notion of a quality score just to give you a few criteria to evaluate an idea. Okay, and so what I did is, there’s you know four key things that I think you could use to evaluate the quality of an idea, and you could give them a one to ten and then you average those together to give an overall quality score.

Okay, so let me just go into detail on these. How big of an idea it seems to be? And again, this is the stuff I talked about earlier: is this an obvious publicly traded company? Like, um, I don't know, Tesla? Hey, it's a new car company, there's lots of publicly traded car companies; that seems pretty big. A new bank: there's lots of banks that are publicly traded; that seems pretty big.

Uh, the opposite of the spectrum is like, um, I don't know, a franchise subway? I'm gonna buy a Subway franchise, you know? Um, you know, that's—that's the under the spectrum or like, "Oh, I'm gonna import some stuff and sell it on eBay," you know, probably a little less obvious how it's big.

Um, the second founder-market fit is really, really key. So again, a 10 out of 10 would be, "Hey, I was on the self-driving car team when I was an undergrad, and then I worked on self-driving cars my whole career, and so now I'm gonna do a self-driving car startup!" Yeah, that's a ten out of ten. A zero out of ten is like, "I'm going to do some kind of advanced AI startup, and I don't know how to program." That's a zero out of 10 right there; don't recommend that.

How easy is it to get started on? LDA is actually, I think, undersold. I don't know if people realize this is actually really key, um, is ideas that are easy to start are highly recommended. And there's so many really good ideas out there that never work because the founders can't figure out how to get started, and then someone in the future does the same idea and has a much better way to get started, and then they win; and then people get bitter because they're like, "I had that idea!"

So, I would argue getting to find an idea that's easy to get started with, um, is just as important as the idea itself. And finally, early market feedback from customers: this just means do people just want it immediately in a sales straightforward, or is it just impossible?

So let me give you some examples of companies that I worked with and advised at YC that you may or may not have heard of. Brex is one; they're in YC117, and they're—I funded them to do a different idea, um, and talked to them a lot during their pivot, and they pivoted during the batch, and they got product-market fit pretty quickly.

And they have now raised hundreds of millions and worth billions in two years. So it's like one of those rare outlier stories. I'm not saying this is common, but holy cow! I got a front row seat to watching kind of the most epic pivot of all time.

I think maybe there are other ones, but I was like, "Wow, that really worked!" So let's talk about the before and after here. What they started with, they had this idea for a new VR hardware headset that you would use to do work or something.

Um, and so if you—here's the scores I just put in, you know, I put these in now—these I didn't do these at the time, but how big does it seem? Yeah, it seemed like medium big; like VR headsets, I don't—there's no publicly traded ones, but that seems like the future, right?

Um, founder-market fit, uh, I gave a one out of ten. These guys knew nothing about hardware or optics or any of the things associated; they were just fintech software folks. So, they had literally no idea what they were doing, and they were very upfront about that.

Um, how easy it was to get started? That's only a two, and the reason is they had to hire hardware people to build a prototype—not a good sign. They needed millions of dollars to build a prototype—not a good sign. And it would take like years and years and years of manufacturing, so this was pretty bad. And then finally, early market feedback: they went and talked to people and they're like, "Do you want to use this?" And no one wanted to use it. So this was bad.

Okay, so overall score, two and a half out of ten. Okay, post-pivot: credit card for startups. That seemed pretty big, there's tons of publicly traded comps of like fintech companies. You know, again, even in startup plan, there's Square, there's Stripe. I mean, there's lots of examples of that, so I'd give that a 10 out of 10.

I did at the time, founder-market fit 10 out of 10. The reason is they had started a fintech company earlier in Brazil, and it was successful, and they sold it for like 20 or 30 million, so they knew exactly what they were doing, and they could write all the code themselves, and they could ship it themselves, and they had all these existing relationships. Man, is that such a better fit!

Um, how easy was it to get started? I'm only giving this a three out of ten because they actually—it's hard to launch a new credit card product. This is one where if you don't have founder-market fit, I would not necessarily try this at home.

Um, there was something about the founder-market fit that made it easy. And then finally, the early market feedback, I'm giving an eight out of ten, and it's because in the batch, they were like, "Hey, do you have a credit card? Do you want to be a customer?" And people were like, "Yes!" And that was a whole sales process like I witnessed it multiple times.

So if that's your sales process, good! It's like a three-sentence sales process, and people say yes. That is good! Let's talk about Retool. They're winner 17, and they're a really good SaaS company, and honestly, everyone should check it out and maybe use it. It's a tool to build internal tools; it's awesome.

Um, I'm encouraging all the YC startups to use this to build internal management pages; it's actually like a really great product. So, um, that's what Retool is; you should look at it. Pre-pivot was Venmo for the UK.

Um, how big it seems? Seems pretty big; like Venmo is big. Um, founder-market fit three out of ten; they didn't really know anything about fintech, but they managed to get a launch; there was something there. How easy was it to get started? I'm giving that a seven out of ten; it's because they already launched; they'd already had a bunch of users then, so that was pretty impressive.

And you know, early market feedback though was only three out of ten, and the reason is no one wanted to pay, and they were losing money on every transaction. So there's all these reasons this wasn't really working, and so they decided to pivot.

Even though this is an example of the insidious sword attraction, right? They had enough traction that it was like hard to decide to give up on this because they had users, and it was like kind of working, so that was a tricky one. Anyway, Retool post-pivot is the no-code internal tools builder. I think it seems like a ten out of ten because 80% of all software built is for internal consumption, not external consumption like a big companies. And so that seems like a big market to me.

Founder-market fit 10 out of 10; one of the founders had made something like this at his college internship, and so he had a very good idea of what the product should be and had relevant experience. Um, it was easy to get started; they built it in two weeks and got their first customers. Great!

And early market feedback, only given a 5 out of 10. People were interested in it, but they weren't sure they wanted to trust some new startup to this thing. So if this one did not fly off the proverbial shelves, they managed to get some users, but it was not immediate obvious product-market fit; they had to build it out a bit. So that was a great pivot.

Next, we got Magic. They were in winner 15; they pivoted during the batch, and what's great about them is they built a profitable and sustainable company. Yay! I wish we all could do that, right?

Um, how did they do it? They started off with this idea for a blood pressure coach where there was an app and you enter your blood pressure, and it would tell you how to lower your blood pressure. Um, didn't seem like a huge idea—uh, two out of ten.

Final market fit: They didn't know anything about health whatsoever or much along those lines, so that's probably just a two out of ten. How was he—was he started? Eight out of ten, because they built the app really quickly, and they got users really quickly, and they followed all the advice; good for them!

And then the early market feedback was like not good; everyone was really polite, and no one used the dang thing. Like, their usage was just horrible, so this one was again was kind of obvious they should pivot. They realized it on their own pretty quickly, and then they built a bunch of prototypes.

They actually built the Magic prototype in a weekend, no joke! This is actually true; not an exaggeration. And they put it on Hacker News, and it went to number one and got like 2,000 upvotes, and it blew up overnight. I don't know if anyone remembers this thing, but it went like crazy viral.

Um, and they got all this press; like it was the most—it was an example of a shots-on-goal thing I was talking about earlier where they didn't know—like they had like five ideas, and there's no way I would have known; they would have known that this would be the thing that would capture the world's imagination, but it did.

And so not obvious founder-market fit on this one either, but it was very easy to get started. It got really good early market feedback, and it seems pretty big, so that was great. It also inadvertently inspired tons and tons of clones.

So I would argue that anyone that's doing chatbot stuff is sort of like a direct descendant about when this thing blew up and everyone's like, "Oh, chatbots for the future!" Uh, was sort of like second-order effects from this going so crazy viral and inspiring so many people.

Segment, they're in YC summer 11; they pivoted a bunch of times including years after the batch. It took them a really long time to get it going; they didn't even pivot during the batch, and now they're worth billions, and it's a really good company.

Um, it's a top data infrastructure company. So pre-pivot, they had this thing that was a classroom feedback tool where like you would give it to students in a classroom, and they could say if it was if they were lost or not. I think was what the feature was.

Um, so if you were confused, you would like push a button, and it would tell the professor. Okay, didn't seem huge, um, in my opinion; final market fit, well, they were students, and they were young, so I would give that like a five out of ten. They weren't experts on education, but they understood the audience.

Um, it was easy to get started; they built it really fast, and the early market feedback was actually decent, where professors liked it, and they got a bunch of schools to adopt this thing. So their sales were successful, but ultimately this was not, you know, it took them on the order of years to discard this one.

Um, and yeah, this was—it's good that they pivoted because now they're doing Segment again, which everyone should use. Uh, it's a data collection tool that one did not seem obviously big. Have anyone here used Segment? It's this JavaScript thing you put on your page, and it connects to other tools.

It was less obvious how that would be a publicly traded company at the time, but the final market fit was great because they built so much analytics stuff. They were world experts on analytics already when they built it.

It was easy to get started because they literally built this thing and open-sourced it and gave it away for free, and people begged for them to support it; like the market begged for this product to exist. They didn't think this was the company; they just had this like thing on GitHub, and so many people were obsessed with this thing that they built. They're like, "Well, I guess the market's telling us we should make this the company," and they did it.

So that I would say is easy to get started and really good market feedback. You know, with the market begging for something, and that you don't even think it's good enough to convince you it's good, that's a pretty good sign. And so that was a good pivot.

Okay, in summary, I tried to give you those examples just to give you real, real, real-life examples of stuff that I personally worked on just to show, like, the before and after and how these decisions are really made. And so changing your idea is part of a startup; the sooner, the better because of the opportunity costs and the shots-on-goal type of stuff.

And when you're considering changing your ideas, especially in the early stages, it shouldn't feel like a big deal; you should probably do it all the time.

And following best practices is recommended. Hopefully, I gave you some good pointers on what those best practices are, but if you're really scientific about this, you can dramatically increase the odds that your startup will work. Great! That's it for me. Thanks!