Startups: The Very Beginning

This was the talk I gave at Y Combinator's Future Founders Conference for women. 

I've given a lot of advice to startups over the past 15 years, but today I'm going to focus on just the advice that matters in the earliest stages.

I'm going to start by talking about cofounders, because they are really the foundation of any startup. You need two things in the founding team that you’d need in any foundation of a building. The founders collectively need certain skills, just as the foundation of a building has to cover a certain space. But you also have to hold together the way the foundation of a building does.

You can always change the idea you are working on, but it's much harder to get rid of a cofounder. So you want to choose cofounders very carefully. They should probably be people you already have a pretty good relationship with: a friend, a classmate, a coworker, a spouse. Someone you've worked with on things and you know is effective and reliable. Also someone who shares roughly the same ambitions and moral outlook as you. Otherwise you'll probably be driven apart by the stress of the startup.

The younger you are, the easier it is to find cofounders. It's easy to meet talented people in a school setting. Also, as you get older, people become more constrained in their lives and careers, and have more financial obligations. So it's harder for them to start a startup with you.

The ideal cofounder relationships are organic. It shouldn't be the case where you say, “I have this great idea and want to start a startup,” and then have to hunt around for a cofounder you don't know very well. Ideally you're already friends and you have the idea together.

So if you think you might want to start a startup one day but aren't ready to do it right now, my advice is just to start collaborating on small projects with friends. Maybe one of these projects will become a startup, but it doesn't have to. The point is just to practice working together.

As well as holding together, the foundation of a building needs to cover a certain area. There are certain qualities founders need to have. But you don't all need to have all these qualities. You just need to have at least one person with each.

There’re a bunch of useful qualities in founders, but I’ve narrowed it down to the 3 most important. The most important quality is determination. There’s got to be at least one founder who's super determined. I've seen so many smart and talented people fail because they couldn't stick with it when things got tough. Startups are really hard and take a really long time. There has to be at least one founder who's just a tower of strength.

Part of being determined is being able to withstand rejection. People will think your idea is lame, customers won't be interested, investors will say no, reporters won't care. And they might not be polite about it either. But you can't let rejection discourage you. This is really hard for a lot of people.

The second thing you need in the founding team is domain expertise. At least one of the founders should be an expert in what you're working on. Which means if you're starting a software startup, at least one of the founders should be a programmer. You can get away with hiring programmers, but you're more likely to succeed if the founders themselves at least know how to program, even if they're not great at it.

I'm not saying that you can't be successful if you aren't a programmer. But if I had to suggest the best way to spend a year preparing to be a founder, it would be to learn the basics of coding. I've seen so many women who had great ideas for startups but couldn't implement them, and had to either outsource their initial version to a development shop, or worse, recruit a cofounder they didn't know just because that person could program. Time after time I've seen these companies blow up, when if the founder had only known enough programming to make a prototype, they probably would have gone on to succeed.

The third quality you need in the founders is the ability, or at least willingness, to sell. Someone has to be the face of the company—to sell the product to customers, and sell the company itself to investors, the press, and potential hires. At least one of you is going to have to sell. 

Those are the three essential qualities: determination, domain expertise, and ability to sell. There should be at least one founder with each. If you have all three, you could theoretically go it alone. But even then you'll find it easier with a good cofounder. It can be done as a single founder, but it’s so rare. 

So once you have the foundation of the building, the founding team, the next layer up is the idea you work on.

Y Combinator's motto is "Make something people want." That sounds really obvious, but it's because founders so often overlook this obvious point that we made it our motto. It's not enough to make something you believe other people want, or that they want, but not that much. They have to want your product enough to actually sign up, to keep coming back, and to tell their friends.  All when you're just a tiny startup.

A good way to ensure that you make something people want is to make something you yourself want. This isn’t the only route to success, but it's the safest one, and the one that the biggest startups usually take.  At YC, we love teams who are domain experts and have built something to solve a problem that they themselves had. Airbnb did this, Stripe did it, Dropbox did it. But the pattern doesn't end with YC.  Microsoft, Apple, Amazon, Google, and Facebook did it too. They all made something the founders themselves wanted to use.

So my advice to anyone trying to come up with a startup idea is to solve a problem that you yourself have. When you're part of the target market, you'll have insights about it that you wouldn't otherwise.

But remember that making something for yourself is just a heuristic to guide you in finding an idea. In the actual execution, you need to focus on users. You need to understand what they want, and be fanatically dedicated to making them happy. One very important piece of advice we give startups is to "do things that don't scale." That means to do so much for your early users that you couldn't possibly keep doing that much if you were bigger. That sounds like strange advice, but it works. Startups have momentum, so it makes sense that you have to make an extraordinary effort to get them rolling.

I don't know of a single case of a startup that felt they spent too much time talking to users. The way you stay on the right path in the early stages of a startup is to build stuff and talk to users. And nothing else.

A critical part of making users happy is to measure whether you actually are or not. Otherwise you can deceive yourself about how much users really like you. Especially if your first users are friends who don't want to hurt your feelings.  There's a famous sentence we often quote at Y Combinator: You make what you measure. Pick a number you want to grow, and focus on that.  The best number to measure is revenue, if you can. That's a really good test of whether users genuinely like what you've made.

During Y Combinator, the Airbnb founders had their revenue target taped to their bathroom mirror. And they hit it. Their target was to become what we call "ramen profitable," which means that you have enough money to pay your expenses if you live really cheaply. You’re eating ramen noodles, but at least you don’t have to quit and get jobs. Once you cross this line, you become a lot more confident. No one can stop you now.

Obviously the less money you spend, the easier it is to become ramen profitable. So it's very important to live cheaply at the beginning. Low expenses buy you time. And you’ll need more time than you think. Even if you have a great idea, you usually need at least a few months to work out the kinks.

The standard metaphor for this time that your initial capital buys you is "runway." You have a certain amount of space to take off, and if you don't, the company is dead. And this metaphor is not over-dramatic. It happens constantly. A startup will be working on a genuinely good idea, but spend too much in doing it, and run out of money and have to shut down.

The other thing that can kill otherwise promising startups is lack of focus. The most successful founders are fanatically focused on their product and their users. The reason focus is so much more important in startups than in most other kinds of work is that there are so many different things you could be working on, and so few people to do them. It's the opposite of a big company. So you have to choose only a few things to work on, and work on them very intensively.

When we do office hours with startups at Y Combinator, this kind of focus is usually the goal. What is the most important problem right now and how are you going to solve it? A lot of the advice we give in office hours is what not to work on. And these are often valid problems, just not the most important problem right now. 

Focus, as Steve Jobs famously said, is about saying no. So what you choose not to do is as important as what you choose to do. One very specific thing you should not do is to become a scenester. Startups are considered cool at the moment. If you wanted to, you could occupy all your time playing at being a startup founder without ever actually doing the difficult and unglamorous work of building things and talking to users. Needless to say this is the road to disaster. So if you ever find yourself doing anything that seems at all scenester-ish, tell yourself "No," and get back to real work.

All the advice I've given so far applies equally to all founders, but there's one more piece of advice I have for women specifically: Don't worry about being a woman. Yes, it may be a little harder for you as a female founder. But it's not going to be so much harder that it will make the difference between success and failure. Startups at least have this advantage over the corporate world: they are already so hard that the additional difficulty imposed by being a woman is rounding error in comparison.

Sure, there are some real obstacles women face as startup founders. But don't let yourself be intimidated or distracted by all the noise out there online about how it's harder for you as a woman. Unfortunately, the conversation around this topic is too often driven by people who have not actually built anything themselves. And they're often incentivized to make the problem of sexism seem bigger than it is. Activists, like all activists, tend to magnify the problem they're addressing. And reporters know that outrage is the best way to get clicks.  But the people making the real difference are the builders not the talkers. They’re the women out there who are quietly and successfully building their companies. Like the women speaking here today.

I want to end by emphasizing a point about startups that I think is especially important for women: there's not just one model of a startup founder. The press presents this image of a stock character of a founder. It’s always a young white guy. Whether they're glorifying it or attacking it, it's usually the same stock character. But when you look at the thousands of actual founders out there, as I have, there's a great deal of variety. I'm an example myself. So don't let yourself be deterred from starting a startup because you don't match the image presented in the press.  

If you can make something people want, if you can focus on delighting users, and you measure how much you delight them in revenue, then you can start a startup. That's the standard to hold yourself to, not the stock character founder you see in the press.