I must build "what I want" and then "what other people want."

timdaub argues that founders must first build what they personally want, and only then build what others want, because building solely for others' demands drains motivation. He says YCombinator's "build what people want" motto doesn't account for personal motivation, and after years of working on other people's priorities he quit freelancing to focus on his own project. In that project, he writes that customers mostly state obvious, stated preferences, and that they often want a "faster horse carriage" when the team is building a car. He also describes Coinbase CEO Brian Armstrong, speaking at a16z's startup school, who considered quitting over employees' push to refocus the company on value-driven goals before deciding he must first like what he works on. Timdaub says he still wants users and sees "build what people want" as useful once the founder's own motivation is secured.