Y Combinator co-founder Paul Graham says most universities answer the question of how to prepare students for entrepreneurship the wrong way.
In a recent essay titled How Universities Should Prepare Founders, Graham argues that schools do not need more entrepreneurship courses or more innovation centers. He says universities should focus on two things instead: helping students believe that starting a company is something they themselves can do, and giving them more time to work on their own projects.
What YC looks for in founders
Graham writes that the traits that best predict whether someone could become a strong founder are not GPA, results in business competitions, or the number of startup classes they have taken. In his view, the signals are much simpler: whether a person is good at making things, and whether they have formed the habit of constantly making things.
He says YC has spent 20 years observing what kinds of people are more likely to become good founders. When YC partners select teams, he says, they look for people who are very good at building and who are used to building continuously.
The reason, as Graham frames it, is that the hardest part of a startup is still the product: knowing what is worth making and being able to make it.
For students who want to start a company later, Graham says a more useful form of preparation may be deep study in fields such as computer science, mechanical engineering, or molecular biology, rather than creating a separate track of entrepreneurship classes.
He also makes clear that “building” is not limited to engineers writing code. Design, science, art, and other kinds of creative work can all become a founder’s core advantage. He points to Steve Jobs studying calligraphy, an experience that later influenced the Mac’s fonts and desktop publishing capabilities.
Because of that, Graham says he would rather encourage young people to pursue powerful ideas and bodies of knowledge worth understanding deeply than steer them into a designated entrepreneurship major.
Don’t go hunting for startup ideas too early
One of Graham’s more counterintuitive points is that students should not deliberately search for startup ideas too early.
He argues that the best startup ideas often do not look like startup ideas at first. Many huge companies began as side projects that founders simply found interesting. If people screen ideas from day one by asking whether something could become a $1 billion company, those ideas that seem strange at first may get discarded immediately.
His advice is straightforward: instead of thinking about entrepreneurship while making nothing, keep making things that genuinely interest you, without worrying first about whether they can become a startup.
Graham says that is also why YC cares so much about what applicants have built and pays little attention to GPA. A project, in his account, solves at least three startup problems at once.
- Actually making something is one of the most effective ways to understand a field deeply.
- Working on a project together is the best way to tell whether someone would be a good co-founder.
- Many of the best startup ideas grow naturally out of side projects.
Universities may need more blank space, not more classes
Graham also criticizes the highly structured nature of university schedules. If schools want the most ambitious students to create more, he says, they may actually need to make students a little less busy.
He uses Harvard’s Reading Period as an example, writing that both Microsoft and Meta are connected to it. Reading Period is the gap between the end of classes and the start of final exams. Students are supposed to use the period to study, but it also creates a rare condition: everyone is still on campus, yet there is no assignment due the next day.
Graham says that kind of time, when students are not being chased by deadlines, gives them a real chance to immerse themselves in their own projects.
He adds, half-jokingly, that just a few weeks of that kind of gap helped produce two trillion-dollar companies. If Harvard’s Reading Period were twice as long, he writes, it is hard to know what US GDP might look like.
Graham acknowledges the main objection. If students are given more free time, many of them may simply waste it. Even so, he says the trade-off is worth it. The least motivated students can only get so much worse, while the most ambitious and creative students may produce outsized positive results once they have more freedom.
That kind of system may not improve the median student, he says, but it could significantly raise the upper bound of what students as a whole are able to produce.
Why business plan competitions may teach the wrong lesson
Graham is especially blunt about conventional entrepreneurship education on campus. He argues that entrepreneurship is like chemistry lab work or painting: it can only really be learned through practice.
If a university is going to offer a course on how to start a company, he says, the only format that makes sense is one in which students actually start companies. In some respects, that is what YC itself does.
But he says a full-time student cannot realistically take three or four classes while also running a high-growth startup at full intensity. If the company starts working, the student may eventually have to choose between the company and school.
That is why, in Graham’s view, universities cannot truly teach students how to be founders through classroom instruction alone.
He singles out business plan competitions in particular. In those contests, students form teams, come up with a startup idea, prepare slides, and pitch to mock investors. Graham says the exercise is not just of limited use. It can backfire by teaching young founders that the most important part of entrepreneurship is persuading investors.
Speaking as an investor, Graham says fundraising is only a “necessary evil.” The people founders really need to persuade are users, and users are not won over by a story but by a product.
His conclusion is direct: the core of entrepreneurship is not building a narrative investors want to hear. It is building a product that users actually want to use.
Rather than thinking about entrepreneurship while doing nothing, Graham says students should make things first and think about entrepreneurship later.