5 Critical & Company-Altering Learnings from B2B Startups | Y Combinator MD Anu Hariharan

YC is university for startups. It's well known as the best accelerator in the world. YC is still the best shot at starting a company. Think of the accelerator as the undergraduate program and continuity is the graduate school. Now we have a full machine all the way to IPO. What makes YC is the community, is the founder community. We want to help that young founder, help them build the confidence that they can do it. It's very hard to compare YC to a venture fund. We actually know our founders from day one. What we look for is how fast does the founder move. Enable entrepreneurship globally. That's a mission. There are just way more founders starting amazing companies everywhere in the world. YC needs to scale. So just a minute on YC. I think many of you know YC is, you know, well known as one of the best accelerators. I help lead our growth fund, the Continuity Fund, along with Ali Rohgani. We launched it 7 years ago at YC, and we— since then on, we continue to support YC companies well after the accelerator program, both through batch programming, helping CEOs learn how to scale, as well as through investments. So what are the 5 critical learnings for founders? And I actually went back with our team and reflected on all our notes of some of the most successful YC companies. So the first learning that I want to focus on is a very important one, but something that you probably have heard many times. Is the product good enough? Many startups usually assume fundraising as a signal or an endorsement that you have a good product. That's the number one mistake most founders do. So we see that at YC Demo Day. Companies from the batch raise $1.5 million, they think they have product market fit. Or they go on to raise a Series A 18 months later and they think they have product market fit. But most of them actually don't have a good enough product. And so when you start scaling before you have a good product, it's quite deadly. You know, Tom Blomfield, who is now a visiting partner but a two-time YC founder, said this very eloquently. If you hire too fast before you have product-market fit, now you have a bunch of people whose jobs depend on you not changing the product. And that's why your product is not going to be good enough. Versus when you have very few people really focused on, is the product good enough, then you have a shot at making it really better. And the money you raised will last you longer. And so you have more time to experiment. So the example that I want to give you here is one from Mutiny. I don't know how many of you are familiar with Mutiny. It helps you personalize websites. It's a B2B startup that went through YC a few years ago, started by Jaleh, who is one of the co-founders. So she said, pre-lockdown, the product was amazing, right? A lot of— because what happened pre-lockdown? It was a hot market. All the startups were funded very well. So if you were a B2B startup or even a B2C startup with a lot of cash, you were OK with your teams buying as many SaaS tools as possible, spending as much money because in your P&L, it didn't really impact that much. And so a lot of B2B startups saw good revenue growth, high willingness to pay. Then the lockdown happened. And what happened? Suddenly, these startups that did not see much churn saw 15% churn. 15%. So was the product really good enough? Was it as valuable? No, because if it was really valuable, they would pay for it no matter what the market cycle is, right? So the way you test for is the product good enough— this is where I'll break it down— you have to ask really 3 questions. There are multiple questions, and there are nuances. But for most companies, it boils down to 3 questions. So what are the 3 questions? One, you should go ask your customer, "If my product goes away tomorrow, how disappointed will you be? Somewhat disappointed or very disappointed?" And if at least 40% of your customers don't say very disappointed, you probably don't have a good enough product. It's just moderately appealing. So we say a rule of thumb is at least 40% of your customers will be very disappointed if your product goes away. So that's one way to test. Second way to test is you should know yourself how often you expect your customer to use the product. So in the case of Mutiny, Jalay said, I expect my customers to experiment with the product or run experimentations at least a few times a month. If it's a few times a month and you go back and study your customer data and go and figure out all those customers who are not using the product that often, they are at a high degree of risk no matter what they tell you. And so the second question you have to ask is usage. How often do you expect the customers to use it? And how does that actually compare to reality? The third trick you can use is growth. But here's the tricky part about growth. You can buy growth so it looks good. So the way to really look at growth is if you have a self-serve model, which quite a few B2B startups do. Is your inbound through the roof? If your product is really good, you probably are not going to be able— you don't have enough people on your team to respond to inbound queries. That's when you have strong product-market fit. If it's once a week or maybe like 5 inbounds a month, probably not good enough. If you have an outbound sales motion, especially especially as a young B2B startup, then you should really look at pipeline and conversion. So one of the things Jelly did was, am I at least converting 20% of my customers when I do a demo? Because if it's far less than 20, maybe it's not that good enough. Because think about it. If you're doing 100 demos a month, that's 20 customers. And if you can't even convert 20, then you have to ask yourself, Is the product really good? What's it missing? What would it take for it to become valuable? And by the way, this step is so important throughout the life of your startup. Another mistake I see companies do is they will hit a certain product-market fit, and then they've gone on to scale, and they forgot to ask this question to their earliest customers. So if there's one thing I'd like you all to take away is make sure you ask this question every year to every team member, to your sales team, if you have customer service reps, is the product really good enough? And find tangible ways to measure. That's the number one mistake we see most founders do, especially post-seed or Series A. The second mistake, which we've seen pretty much 100% of YC companies do, including the founder who's going to talk after me, which is Josh at Gusto, you should ask him, is they all underprice. Because think about it. You worked so hard to get a product out of the door. You finally have 10 or 15 customers. And you're like, wow, jeez, I want to grow it to 100 customers. So maybe I should just charge them like $5 a month. Right? That's the number 2 mistake we see. Now, people often ask me, OK, are startups on average underpricing by 30%, 40%? Let me give you a story. Peter Reinhardt from Segment. I'm sure many of you are aware of Segment. If not, you should look it up. It's an absolutely amazing product to collect, unify, and keep your customer data in one place. They pretty much graduated from MIT and through a random pivot landed on this idea. And their first module was an Analytics.js module. And they come from the open source mentality. So when they hit product market fit, they had enough customers using their product weekly, consistently. But they were paying nothing because these guys didn't ask for anything. They didn't charge for it. And so— but they had $100,000 from YC. And, you know, no matter— even though it was just 3 co-founders and maybe 1 or 2 engineers, the money was running out. And so at some point they said, "Wow, we have to charge something." But they felt so bad to charge, so they profusely apologized, sent very long apologetic emails asking for $10 a month. In fact, I think their first price point was $5 a month until one customer wrote back to them and said, "You better charge more because otherwise I think you're not serious enough to build a company, and then I don't want to put all my data on this platform which is not serious enough to build a business." That was the wake-up call for Peter. He's like, wow, me not charging, they assumed I'm not here to build a company, and they don't want to trust us with their customer data. But they still kept it for a while, $10 a month, so $120 a year, until they hired a sales advisor. And the sales advisor looked into their product and said, you should be charging $120,000 a year for this product. That's how valuable your product is. Peter said, there's no way. No way I'm charging $120,000. So the challenge he gave him was, OK, the next sales meeting you're going to do, which was Nat Friedman at the time. He was the CEO of Chameron. You're going to ask him for $120,000 a year. And if you don't, I'll quit as your sales advisor. That's what it took to convince Peter to make the ask. Peter made the ask. Nat said, I'll pay you $12,000 a year. And they settled at $18,000 a year after that meeting. That was 150 times the original price Peter had set with the $10 per month. That's how much B2B startups underprice. And this is true. You can see that with Gusto. They underpriced the first 2 years. It's very nerve-wracking. But I often say the price is often the way to test if truly you've built something valuable. Because if you have, and your customer is going to be very disappointed about you disappearing, they actually won't blink twice. And so the best way I often tell founders to test is increase your price by 30%. And if your customer's not blinking twice, then you probably have underpriced a lot. And you can keep testing it, especially with your early customers, until you land at a price point that makes sense. But know that I can tell with a very high degree of confidence that most B2B startups at YC have underpriced. And so there's a long way ways to go before you can charge where your customer is going to cringe. The third most common mistake I see, especially at the earlier stages, companies do is they can rarely serve all B2B customers well. B2B is so broad. When you say I'm a B2B startup, who are you really selling to? Tech companies is broad. You have startups with less than 50 employees. You have You have mid-market startups with 50 to 100, 200 employees. And you have late-stage startups like Stripe. And then you have public companies like Airbnb. You really need to be nuanced. You can't sell to everyone. This is another mistake I often see founders do, especially when they go right after YC Demo Day. Oh, we have signed up 8 customers or 10 customers. 3 look like this, 5 look like others, and 2 from a different— and they all have different product needs. And so I think it's very— at the earliest stages, it's OK to iterate. But you have to land pretty quickly on what's the segment. So my partner Michael Siebel has a very easy phrase. Don't try and say it's SMB or startup. Figure out who are your easy customers. As it is, it's so difficult to build a startup. So when you're starting out, figure out who are your easy customers. So figure out for whom— whatever bare-bones MVP you've built is more than enough, and they're very happy. And you start with them. For Brex, it was startups because founders couldn't get a credit card that easily. So instead of going and solving for all the people in the world, the first segment was startups. For Gusto, it was SMB. The payroll software that they built could service any company with less than 10 employees. It's not SMB, any kind of SMB. As long as you have less than 10 employees, our payroll software will work. It doesn't work for others. So figure out what your niche is. And remember, as you scale, it's really hard to, again, service multiple segments. So I'll use the example of Gusto. So for the first entire year, I was actually shocked at the stat. Can anyone guess how many customers Gusto had? Any guesses? 2,000? 50? That's a wide range. What's that? 1 million. Okay, from your mouth to God's ears. We still have to hit a billion. But they had 100, 100, NPS of 80, 30% month-on-month growth. Of course, you're coming off a small base, so growth rates are really good. But 100. Why only 100? Well, A, it's really difficult to reach SMBs. They're not online. Now they are more online, but they were not online when Gusto started. Second, you cannot make a mistake in payroll. You can't say move fast and break things if your paycheck isn't arrived. Guess what your customer is going to say? And payroll is a very difficult software to build. So that's why they took the time to build. And they wanted to make sure that the 100 customers that they worked with were delighted by their service. Now, 2 years in, they knew that they should roll benefits as they were scaling their customer segments and supporting companies with less than 25 employees. It took them almost a year to launch benefits. It didn't take off. Why? Any guesses? Go ahead. People just thought them as payroll. Well, initially it was ZenPayroll, but they changed the name to Gusto, which helped them. Any other guesses? Different what? Sorry? Different profile. Actually, it's more than different profile. It is a different profile, but the customer that wanted benefits wanted more. They just didn't want payroll plus benefits. They wanted 401(k). They wanted— so your sales, it's not like your sales team was underperforming. The customer segment, their needs are different. So therefore, it's really important that when you're scaling as a B2B startup, Every startup segment that you focus on, you got to come from the customer lens. What does that customer need? And it may be fundamentally different from the product you're offering. You may not have the time to build all the features. And that's OK. Take the time to do it. But be deliberate. If you assume that just adding a feature will help you scale to the next segment, that doesn't work. That doesn't work 8 out of 10 times. You may get lucky with 1 or 2, but that doesn't work 8 out of 10 times. The fourth common mistake— this is actually one of my favorites— is early sales should not be delegated. I see so many startups when they hit $1.5 million ARR, they're on their way to hire the VP of Sales. And if I ask them to explain how exactly How exactly are you going to tell the salesperson? How exactly do you expect the salesperson to go acquire customers? The founder's response would be, well, that's what I'm hoping they would tell me. That's not their job. You hire a salesperson when your sales call and demo calls become extremely boring and repetitive. Because that's when you've figured out what your go-to-market motion is. Remember in the early days, the point number 1 I made? Is your product good enough? You don't even know that. You have to do 100 to 200 customer calls, customer conversion to know whether your product is good enough. When your 150th demo call is boring and you know exactly how it's going to go and you know they're going to convert, that's when you go hire a sales leader. Usually, I mean, Christina at Vanta actually did this extremely well. She said she did it up to the point point where she knew that week which customer calls were going to convert, how the demo call was going to go, what questions they were going to ask, and what her answers were going to be. And she had 1 to 2 sales SDRs or sales account— one of 2 sales reps helping her. And she knew even how to train them. That's when she understood that it was critical to hire a VP of sales. Because we've now cracked— we found a way to repeatedly sell the product. And we know exactly how to do this, that we can now hire a VP of sales who can build up the sales team, help scale up the sales motion so that we can reach thousands of customers. If you do this step prematurely, it's not your salesperson's fault that you couldn't grow. The most common mistake is you hire a VP of sales. They've cranked up your team. You have all these sales accountants, but your growth is slow. And the root cause of that is not that the sales motion isn't working. Your product is not appealing enough. So I often tell founders, one of our most critical advice at YC is do the sales till it becomes so boring that you know how to predict. That's when you hire your first sales leader. And then the last learning Which, last but not the least, is the most important. Plan to reach default alive. I have never understood why so many B2B startups are so reliant on external investment funding. There is no— I mean, there is some reason for some of them. But actually, if you look at it critically and closely, Many of them should be able to scale without any funding. I usually say if you have hit $5 to $10 million ARR, you cannot get killed. You should be default alive. And so what is default alive? If you are not familiar with this term, I highly recommend that you read Paul Graham's blog post. This is something that YC takes very seriously. In fact, when you graduate from the YC batch, it's part of your monthly investor update. If you answer this as default debt, it's kind of shaming you into how to get this to default alive. So what's default alive? If you are 8 to 9 months or even 15 months into building your startup and you have decent revenue scale, then the question is, okay, how soon before you become profitable? Now, why is this important? Especially some founders will tell me, "Hey, there are investors waiting to give me money. Why should I care about this?" Well, most people are not familiar with this term, which is called the fatal pinch. Does anyone know? We've actually talked about this quite publicly, but I don't know how many of you know the fatal pinch. Anyone here? How many of you know it? Raise your hand if you know what is the fatal pinch. Okay, 2. 2 of you know. So let me tell you what the fatal pinch is. The fatal pinch is you raised a bunch of money, you hired a bunch of your team members, but your growth is less than 10% month on month, and you don't have enough runway or the time to experiment. And no one's going to give you more money. That's the fatal pinch. Versus if you have raised money, spent the least amount on hiring, and you're really finding a way to make money— remember, charge, don't underprice, charge enough so that it supports your CAC— you should be on your way to at least break even. And so we often encourage startups to get to default alive because then, you know what, even most investors will be willing to give you money because you don't need money. That's when investors give you money, when you don't need money. And you found a sustainable way to acquire customers and support them with their own— with their own price that they pay, the revenues that you're making. If not, you don't have a business model. The problem with not focusing on Default Alive is you have not figured out the business model. That's the segment problem. It's all free. You know, by the way, it's not the only company. There's another company that went through YC called PostHog. It was one of the most successful open source product analytics launch that they did through YC. Top number one on Hacker News. 30,000 signups. James was extremely confident that when he turns on the paid subscription, at least 5% will convert. Guess how many converted? You said 2. 30. 30 people paid. And so then he went back to iterating on the product and figuring out who are the easy customers, how can I make sure I built the most valuable thing for them, and then if you build the most valuable, they'll pay and price appropriately, right? And so when you do those three things, you will be on path to default alive. And so the success case, I know it sounds like a survivor bias, but Zapier, Wade from Zapier, never believed in fundraising. And I think that's why he hit default alive with $1.5 million. He's never had to raise money. He doesn't need to because he prints cash. He uses that cash to invest in the company and then scale. And I strongly believe— at YC, we often say we wonder why so many B2B startups are fundraising. And they're in this fundraising mode every 12 months. They're back boasting their valuations and the numbers. If in fact, I'd say you should be boasting your revenue number. And your breakeven, not the valuation. That's a vanity metric. That doesn't indicate anything. So I would say the fifth mistake most people do is plan to reach default alive because it doesn't matter whether it's 2021 or 2022 or 2023. You are in control of your destiny. And it can be done. It may mean that you grow slightly slower because you're not able to invest in go-to-market. Ahead of time, and that's okay. That's perfectly okay. The key is, do you have a product that lasts for decades for customers to use? And that's all that matters. So in summary, if there are 5 things— again, only 5 things— that you can take away from this presentation, I'd say the number 1 is the product Is that good enough? And make sure you ask this question every 6 months for the life of your company. You are the only person as a founder who can do that. No one will ask that as often as you do. Second, price appropriately. Know that almost all B2B startups are underpricing. So go try an experiment. Go raise the prices by 30%. And see who flinches and who is okay with it. 3, companies can rarely serve all B2B customers well. So focus, especially early, on the easy customer. Iterate to figure out who your easy customer is and just try to acquire those easy customers. 4, do not delegate early sales. May— try— founders should do sales until it gets it's so repetitive and boring that it's so predictable, that's when you hire your first sales leader. And last but not the least, plan to reach default alive. Try to bootstrap the company or get to break even with the initial investment so that you never have to rely on external investors to scale your company. Thank you, everyone.