Paul Graham: illogical startup

Stanford course CS183B: How to start a startup . Started in 2012 under the leadership of Peter Thiel. In the fall of 2014, a new series of lectures by leading entrepreneurs and experts of Y Combinator took place:



It is good to be a father, because when you need to give someone advice, you can ask yourself: “What would I say to my own children?” You will notice that this question can give your thoughts the right direction. My children are still small - my two-year-old son, today, when asked what will happen to him in a year, said: "I will be a bat." The correct answer would be: "I will be three years old," but the option with the bat is much more interesting. And although my children are small, I already know that I will tell them about startups if they go to college - and that’s exactly what I’m going to tell you now. You will literally hear what I would say to my own children, since most of you are young enough to be my son and daughter.

Startups - things are very illogical - I really don’t really know why.Maybe it’s simply because knowledge about them has not yet been introduced into our culture, but, whatever the reason, statues are not an area where you can constantly rely on intuition. In this sense, they are like skiing.

Did any of you guys learn to ski as an adult? When you first start skiing and want to slow down, the first thought you have is to lean back because this instinctive movement “works” in all other situations. But try to ski like this, and you will fly head over heels from the mountain. As I found out from personal experience, an important element of learning to ski is the ability to suppress this impulse. Over time, this will become a habit, but at first, going down the mountain, you just try to keep a list of things in your head: you need to alternate the leading leg, try to follow the s-shaped curves, do not “fall” onto the inside of the foot and all that jazz.

Startups are just as unnatural as skiing, and for startups there is also a similar list that you need to remember to get started. Today I will tell you about the first points in this list of illogical things that you need to memorize in order to prevent your instincts from misleading you.


1. Do not believe your instincts

I just named the first item on this list: startups are so strange that if you follow your instincts, they [instincts] will lead you astray. If you only remember this thing, then before you make a mistake, you can stop and think.

When I ran Y Combinator, we joked that our task was to tell the founders things that they would ignore - and it really was. Over and over again, YC partners warned the founders of the mistakes that they were ready to make, and over and over again, startupers ignored these warnings, and then returned with the words: “It’s a pity that I didn’t listen to you then.” And now, among the co-owners of the startup, there is some strange character, and the founders can not do anything about it.


However, you can trust your instincts when it comes to people. Until recently, your life was a little like a startup, but the way you interact with people in ordinary life and the way this interaction happens in the business world are things of the same order. In fact, one of the big mistakes that the founders make is that they do not fully trust their intuition about people. They meet someone who, it seems, makes a good impression, but deep down they suspect something is amiss, and subsequently it all crawls out, and the founders can only say:

“You know, I suspected that something would go wrong with this guy, but I didn't pay attention to it because he made such a good impression.”


There are representatives of a special type of thinking in business, it is typical for founders with a technical education, which, it seems to me, all of you have. You think that business is such a somewhat unpleasant activity. Therefore, when you meet people who seem smart, but slightly unpleasant, you think: “OK, this must be normal for business,” but that’s not so.

Just pick people the same way you choose friends. This is one of those rare occasions where you can indulge yourself. Work with people who you like, and whom you could respect, and whom you know long enough to be sure of, because there are a lot of people who can successfully pretend to be “suitable.” Just wait for the situation when your interests diverge, and you will see everything for yourself.

2. Do not try to become a startup expert

The next illogical moment - it may disappoint you somewhat - is that in order to succeed in a startup, you do not need to be an expert on startups. This distinguishes our classes from any others that you can attend.

If you study French, you can speak French at the end of your studies. You will learn, you will not necessarily speak exactly like a born Frenchman, although it is quite similar. Our classes may teach you to understand startups, but this is not what you need to know. To succeed in a startup, you do not need to be an expert on startups, you need to be an expert on your users .


Mark Zuckerberg did not succeed in Facebook because he knew about startups, he succeeded even though he did not understand anything at all about startups - Facebook was first registered in Florida [and not in Delaware]. Even you guys understand that this is no good. He succeeded, despite the fact that he did not understand anything in startups, because he was very well versed in his users.

Most of you do not know the mechanics of attracting investment from business angels, right? If you worry about it, stop it. I am ready to say that Mark didn’t know that either. Even if he drew attention to the fact that Ron Conway wrote him a check for a large amount, now he probably does not remember about it at all.

In fact, it doesn’t concern me that studying the mechanics of creating a startup in detail is essentially pointless, but that it can be dangerous in some way, since another typical mistake of young founders is the desire to go through all the known stages of creating a startup. They get a fairly sane idea, they attract investments to add value to their project, and then they rent an office somewhere on South Market Street in San Francisco [the place where the offices of many successful startups are located - GitHub, Dropbox, AirBnB], they’re gaining a crowd of acquaintances into a team, and suddenly they realize that the project is slowly going down to the bottom because, imitating all the external attributes of a startup, they ignored the only truly important thing: they did not create the product people needed . 


We saw that this happens so often - I mean a situation where people are addicted to the process of creating a startup - that we even came up with the name for this: "A house for games." In the end, I realized why this is happening: the reason why young founders seek to copy the external attributes of a startup is that they have been taught this very thing all their life to the present moment.


And it is not surprising that after they have been effectively trained to play such games all their lives, first of all, the young founders who create the startup try to calculate the rules of this new game. What kind of "extracurricular work" do startups do, what do I need to do? It so happened that a measure of the success of startups is the amount of funds raised - and this is another mistake of the founders. They always want to know what the catch is, what tricks are there to help convince investors. And we are forced to tell them that the best way to convince investors is to establish a really working fast-growing startup , and just tell them about it.

Then the startups say: well, and what are the special techniques for ensuring rapid growth - this issue is compounded by the presence of a term such as “Growth Hacks”. Wherever you hear someone discussing growth hacking, just mentally replace this word with “bullshit”, because, as we constantly tell the founders, the only way to make your project grow is to do something that like users and tell them about it . That's all you need, that's all growth hacking.

A lot of conversations between partners from YC and the founders begin with the founders saying: "Like me ...", and the partners answer: "Just ...". Why does everyone love to make things so confusing? For many years, being puzzled by this question, I eventually understood the reason: people are looking for “tricks”, tricky moves in everything. They are trained to seek them out.

3. Do not try to outsmart the system.

And this is the third illogical thing you need to know about startups: starting to work on a startup, you can no longer "find loopholes in the system." You can continue to beat the system if you work for a large company. Depending on how bad things are going there, you can succeed with the help of someone's protection, you can impress others as a productive employee by sending letters late at night or - if you're a smart guy - by changing the time settings on your computer - no one checks the contents of the headers block, right? Yes, I like the audience that understands my jokes. If I said this at a business school, they would ask me: “What? Headers? What does the headline have to do with it? ” Oh hell, I just realized that they were recording me on video. From this moment I will tell everything strictly according to plan.

So, in startups, this approach will not work. There is no boss who can be deceived - how to deceive the one who is not? There are only your users, and it is important for all of them that your product does what they want to get, right? They are like sharks, sharks are too stupid to try to hold them; you cannot wave a red rag and trick them - in their understanding, there is either meat or it is not. The danger - especially for you - lies in the fact that unfair play to a certain extent works against investors.


If you know very well what you are talking about, then you can fool an investor by one, maybe two rounds of financing, but doing so is not in your interests. I mean that you are doing this to increase your own capital, so it’s dishonest to do this in relation to yourself. This is a waste of your own time: if a startup is doomed, all you do is kill time on it. Therefore, do not try to outsmart the system. In the field of startups, there are tricks in the same way as in other areas, but their significance and effect are an order of magnitude lower than solving real problems.

For someone who does not understand anything in raising capital, but who does what his users like, it will be much easier to get financing than for someone who knows all the book tricks, but whose product no one uses.

In a sense, it’s bad news for you to outwit the system in this area. This robs you of your most powerful weapon — in the end, you spent twenty years mastering it.

Nevertheless, it seems to me that existence in a world where, in order to win, there is no point in tricking, is very exciting. Studying in college, I would be delighted with the thought that in the world there are situations in which “short cuts” mean very little, or even mean nothing at all. And they are: this is one of the most important things that you should think about when planning your future. How to “win” at each stage of the work, and what will this gain give you?

This leads us to the fourth illogical truth about startups. If you create a startup, it will capture a huge part of your life, more than you can imagine. And if your startup succeeds, it will absorb your life for a long time, at least for several years, or even for a dozen years, and possibly for the whole time until the end of your career.

4. It will never be easier

It may seem to you that Larry Page lives a life that can only be envied, but in this life there are moments that you really don't make sense to envy. For him, everything looks as if at 25 he started to run as fast as he could, and since then he has not been able to rest for a second. Every day in the Google Empire, there is some kind of dirty trick that only the emperor can deal with, and he, as the emperor, really has to deal with it. If he goes on vacation for at least a week, a whole mountain of such dirty tricks accumulates behind him, and he has to rake it up without complaint, because, firstly, he, as the “father of the company”, cannot show a shadow of fear or weakness and secondly, if you are a billionaire who complains about the difficulties of life, no one will ever sympathize with you.


This all has a strange side effect: the difficulties of a successful startup’s life are hidden from the vast majority of those who have not yet succeeded. Go up to the man who just won the hundred meter race at the Olympics - he will desperately try to catch his breath. Larry Page does the same, but you will never see it.

Y Combinator has already funded several companies that can be called extremely successful, and in each case, the founders said the same thing: "It will never be easier." The causes of the problems are changing, so you may be worried about larger things, such as the delay in building your new London office, and not about a broken air conditioner in your one-room apartment, but the total amount of anxiety never decreases. He is probably only growing.

To start work on a successful startup is the same as deciding to have children: as if you are pushing a button and everything in your life is irrevocably changing. To be honest, children are, of course, wonderful, but if you need to remember one single thing from this lecture, remember this: there are a lot of things that are much easier to do before you have children - and many of these things will make you a better parent when you still have children. In affluent countries, most postpone the “push of a button” for a while, and I think that you are very familiar with this rather intimate procedure. But at the same time, when it comes to startups, many believe that they should be created while studying at college.


What you really need to know is what your user wants. You can’t learn this without creating a company, which means that in reality you can learn startups only from your own experience. In college, this will not work for the reason that I just talked about. A startup will absorb all your time. If you create a startup in college as a student, you will not be able to “sit on two chairs,” because if you create a startup, you will cease to be a student. Nominally you will be a student, but not for long.


Mark Zuckerberg will never be able to leisurely relax abroad. If he travels to another country, he either makes a de facto formal visit or hides incognito in a fashionable hotel, such as George V in Paris.

He will never be able to travel with a backpack behind Thailand - do people still do that? More people traveling around Thailand with backpacks? Oh, for the first time in my lecture, I see that the audience has shown enthusiasm. It was necessary in Thailand to tell all about it.

He can do what you cannot do — for example, take off a plane to travel abroad. Very big plane. But success squeezed out of his life happy accidents, insights, spontaneity. Facebook controls it to the same extent that it controls Facebook. 


Despite the fact that being under a certain pressure from a project that you consider to be the work of your whole life is not bad, insights have their advantages . Among other things, they give you more options for choosing the direction of “the work of your whole life”. In this case, you do not even need to pay for it.

You don’t sacrifice anything if you refrain from creating a startup at twenty-odd, because the probability of success in this enterprise will only increase if you wait. The likelihood that you will launch a third-party project in your twenty that will “shoot” like Facebook, after which you will be faced with the choice: to stay in the project or to leave and choose to stay, is astronomically small.

Because startups usually “shoot” only when the founders push them in every way. It is extremely stupid to try to achieve this at twenty-odd.


Should I do this at a different age? Creating a startup is quite difficult - if I have not yet clarified this point, then I will try again. So, creating a startup is very difficult . And if it is really too hard, what if you are not ready for such difficulties?

5. You never know for sure

The answer is contained in the fifth illogical truth about startups: you never know for sure. Thanks to the accumulated experience, you already imagine what prospects could open before you if you wanted to become mathematicians or professional footballers. This is not true for every audience. But in your life you were unlikely to often do something similar to creating a startup - unless your life before that moment was very, very strange.

I mean, creating a startup will make a big difference if you succeed. Therefore, you need to try to evaluate not what you are now, but what you could become. And who can give an exact answer to this? Certainly not me. For the past nine years, I have only been doing what I was trying to guess - this was my job (I wrote “predict” in the notes and said “guess” instead - a Freudian clause).

Seriously, it’s very easy in ten minutes to determine how smart a person is. Throw a pair of tennis balls over the net and see if he can beat them so that they do not fly away unclear where - this is not difficult. The most difficult and important part of my work was to predict how stubborn and ambitious the candidates would be.

In solving this problem, perhaps, there is no more experienced person than me. I can say that the expert understands this: not so much. My experience has taught me to remain completely unbiased about which startups from each set will become stars. Sometimes the founders think that they knew what the future holds for them. Some come, convinced that they are gutting the Y Combinator exactly like they did with the few tests that were completely distant from reality that they had encountered in life before. Others come, thinking that they got into Y Combinator by mistake, and hope that no one will know about it.


There is virtually no correlation between these approaches and the future of the startup. I read that the same is true for the army. Boastful recruits have no more chances to become professional fighters than quiet people, most likely for the same reason. The tasks that need to be solved at this stage are too different from everything that a person has encountered before. If you are scared to death by the prospect of creating a startup - you better not do it. Unless you are one of those people who excel at something by stepping over fear. Otherwise, if you are simply not sure whether you can succeed, the only way to verify this is to create a startup, not right now.


So, if you want to create a startup one day, what do you need to do in college now? There are two tasks that now need to be solved: come up with an idea and find co-founders. Both of these are achieved in the same way, which leads us to the sixth and final illogical truth about startups.

6. Do not try to extort an idea

You can come up with the idea of ​​a startup without thinking specifically about such ideas. I wrote an entire essay on this subject and do not want to repeat it here in its entirety. In short, the point is that if you consciously try to come up with the idea of ​​a startup, you will get not just a bad idea, but a bad idea that looks plausible. And this means that such an idea will mislead you and everyone around you. You will spend a lot of time before you realize that there is nothing good in your idea. You can come up with a good startup idea by stepping back a step. Instead of consciously making an effort on yourself and deciding what to do, you should turn to a state in which ideas come to you yourself.

In fact, this process proceeds so out of control of your consciousness that at first you won’t understand that the thought that came to your mind may be the idea of ​​a startup. Yahoo, Google, Facebook, Apple were created that way. At first, none of these companies was thought of as a company - they were just third-party projects. The best ideas, as a rule, arise as third-party projects , because they are so different from your usual affairs that your mind rejects them as ideas on which the company can be built.




Even Y Combinator is a thing that I started working on simply because it seemed interesting to me. I think I have some kind of internal compass that guides me. But you are not me, and I do not know what is going on in your head. Maybe if I study this question more deeply, I will come up with some kind of heuristic concept for defining really interesting ideas. But for now, all I can offer is a hopelessly unfounded statement. Incidentally, it contains the true meaning of the word "unfounded." So, the hopelessly unfounded statement sounds like this: if you are generally interested in interesting problems, then the energetic satisfaction of curiosity can be an excellent preparation for creating a startup, and at the same time a wonderful life principle.


Although I cannot give a general definition of an interesting problem, I can tell you about its large component. If you think technology is something that spreads like a fractal pattern, every detail of this pattern can be an interesting problem. Steam engines are not so interesting these days, although who knows. So, a way to surely turn to the perception of unconsciously emerging startup ideas is to start exploring the latest technological advances.

As Paul Buckheit said about this, “Live the future.”

And when you immerse yourself in this study, ideas that may seem prophetic to others will be obvious to you. You may not recognize startup ideas in them, but you will know that sooner or later something similar will happen.



The most optimal behavior strategy in college is that if you want to become a successful startup founder, you should not go to any new entrepreneurship training program. This is a classic version of university education for education. If you want to create a startup, you need to study serious things in college, and if you have this inborn curiosity and thirst for knowledge, you just need to follow your inclinations.

The only element of entrepreneurship - I can never say this word without a smile - which really matters - this is expertise in the chosen field. Larry Page would not be Larry Page if he had not become a search expert, and he became one because he was genuinely interested in this topic, and not because of some other hidden motives. A startup at its best is simply a desire to satisfy genuinely curiosity, and the best way to create it is to put curiosity first. Therefore, my last advice to young people who want to become founders of startups can be expressed in a nutshell: just learn .

Well, how much time do we have left? Eighteen minutes for your questions - not bad. Do you guys have any questions?

















[ Translation of Adora Chung’s next lecture ]