Being called crazy is common, and it proves nothing
Skepticism follows most new ideas, good and bad alike. It is something to endure, never something to count as evidence.
When you innovate, you’ve got to be prepared for people telling you that you are nuts.
In the 1970s the relational database was an idea in research papers by Edgar Codd at IBM. Many practitioners thought it would be too slow for real business use, and IBM itself was cautious about turning it into a product. Larry Ellison and his partners built a commercial version anyway, and Oracle brought a relational database to market before IBM did. It became the foundation of one of the largest software companies in the world.
Ellison’s line comes from that kind of experience. A new idea, by definition, runs against what experienced people believe, and experienced people say so. Anyone building something new will hear that the market does not exist, that the technology will not work, or that the incumbents will crush them. Founders who expect this are less likely to be knocked off course by it.
The line needs a counterweight, and Ellison’s own record provides one. In the mid-1990s he championed the network computer, a cheap machine with no hard drive that would run software from the network. Critics said it was a poor fit for how people actually used their PCs. They were right about that period; the product did not catch on, and falling PC prices undercut it. Being called nuts turned out to be accurate feedback.
The point is that ridicule is not a signal in either direction. Most ideas that are called crazy are crazy, and a few are early. The work is in telling them apart, and that means asking what, specifically, the critics believe will fail. An objection based on habit (“nobody does it that way”) is weak. An objection based on cost, physics or customer behavior deserves an answer, with data if possible. Oracle could point to Codd’s work and to working software. The network computer’s case depended on customers wanting to give up local storage, which they did not want yet.
Timing complicates the picture further. Some ideas called crazy are simply early. Cheap machines that rely on the network for their software may yet make sense once connections are fast and reliable everywhere. Being wrong for a decade and being wrong forever look identical to the company that runs out of money in the meantime.
Endure the reaction, then weigh the arguments. A founder who uses the line to dismiss every criticism has kept the courage and thrown away the information.