In a Startup, There Is No Test to Game
The Y Combinator founder says top students arrive trained to ace tests and charm investors, but users only care if the product works.
YC Root Access · Lecture 3 - Before the Startup (Paul Graham)Paul Graham, the investor advising young founders on what comes before a startup, was describing a pattern he had seen over and over. New founders copy the outward steps of building a company, raising money and renting offices, without making anything people want. He said his team had a nickname for it: "Playing house."
Then he offered his explanation for why young founders do it. He argued they have spent their schooling learning to outsmart the way they are measured, and college rewards it.
He confessed he had done it himself. In many classes, he said, only a couple dozen concepts were suitable for test questions, so he did not master the material. He tried to predict the questions and prepare the answers in advance. He remembered a class on Automata Theory, the study of abstract computing machines, where only a few questions even made sense to ask.
That habit, he said, becomes a founder's first impulse. New founders ask for shortcuts to persuade backers, and then shortcuts to expand fast. When founders hear the phrase growth hacks, Graham said to mentally translate it to "bullshit," because real expansion comes from building something users love and then telling them about it.
His line for that moment was this: "Starting a startup is where gaming the system stops working."
What he meant, he explained, is that there is no supervisor to mislead. At a large employer, he said, advancement can come from currying favor with managers or feigning late night productivity. In a startup there are just customers, and customers judge only whether the software fulfills their need. They are, he said, like sharks too dull to deceive, reacting only to food or no food.
He added one limit. A persuasive speaker may mislead backers through an early round or two. But he said that only squanders the founder's own time, because a venture without customers is doomed. A creator ignorant about pitching but with a product people adore, shown by a climbing usage chart, will collect cash more easily than a polished operator with stagnant numbers.
He closed that thought by saying he found it thrilling that some parts of life are not won by outsmarting the rules.
Questions people ask
If gaming the system does not work, what actually makes a startup grow?
Graham's answer is simple: make something users really love, then tell them about it. In practice that often means talking to users directly, watching how they use the product, and fixing what frustrates them. Growth follows when people want the product enough to keep using it and tell others.
Why can founders fool investors but not users?
Investors judge a promise about the future, so confidence and polish can carry a meeting or two. Users judge the product in front of them every day, by whether it solves their problem. If it does not, they simply stop using it, no matter how good the pitch was.
Does this mean fundraising skill does not matter?
It matters, but Graham argues it matters far less than having users who love the product. He said someone who knows nothing about fundraising but has made something users love will raise money more easily than a polished pitcher whose usage chart is flat.