Assume the plan is wrong, then find out cheaply
The second sentence carries the weight. Once a founder accepts that the plan contains mistakes, speed is simply the cheapest way to find them.
You need to fail quickly. If you think you know what’s going to work in any aspect of your startup, you are wrong.
Chuck Gordon co-founded SpareFoot, an Austin company that helps people find and book self-storage units. The business began in 2008 with a different plan: letting ordinary people rent out spare room in garages and basements. That version did not take off, and the company turned to listing professional storage facilities, which is what grew. His line reads like a summary of that experience, stated as strongly as possible.
The strength is deliberate. “In any aspect” covers the product, the price, the customer, the marketing channel and the hiring plan. Founders usually accept that some of their assumptions are wrong while privately believing that the important ones are right. Gordon’s version removes that comfort. If everything is presumed wrong until tested, the only question is how fast and how cheaply the tests can be run.
That changes behavior in useful ways. A founder who assumes the pricing is wrong tries two prices in the first month instead of defending one for a year. A founder who assumes the target customer is wrong talks to three different kinds of buyer before committing the sales budget. A founder who assumes the marketing channel is wrong spends a few hundred dollars on each of several channels before spending thousands on one. Each quick failure narrows the field, and the company arrives at something that works with money still in the bank.
The absolute form also overreaches. Founders who have worked in an industry for years do know some things. Someone who ran a storage facility knows how operators think about vacancies, and treating that knowledge as worthless would slow the company down. The point is to hold it as a strong hypothesis, tested first where it matters most, not to throw it away.
Failing quickly has its own trap. A test that is too small or too hasty can kill an idea that needed more time. Some products take months to show their value: anything with a learning curve, anything that depends on network effects, anything sold to institutions with slow buying cycles. A two-week experiment on a product like that will usually fail, and the failure says more about the test than about the idea. Founders who pride themselves on failing fast can abandon good ideas as readily as bad ones.
The better discipline is to decide in advance what a fair test looks like: how long it runs, how many people it reaches, what result would count as a signal. Writing those criteria down before the results arrive keeps them from being bent to fit. Then fail as quickly as that test allows, and no faster.