Counting heads is not the same as thinking

The sentence comes from Warren Buffett’s letter to the partners of his investment partnership reviewing 1965, the year the partnership took control of a struggling textile company called Berkshire Hathaway. Buffett was explaining how he decided. He took no comfort, he wrote, when important or loud or numerous people agreed with him, and no discomfort when they did not. When the facts could be checked and the reasoning was sound, he acted, whatever the consensus.

For an investor this is almost a job description. A share price is a running poll of what the market believes. Buying only what the crowd likes means paying the crowd’s price, and returns above the average can come only from being right when most people are wrong. Buffett has lived through several rounds of being called out of date, most recently in 1999, when he declined to buy technology shares and financial magazines asked what had gone wrong with him. The crash that followed answered the question.

Founders run polls constantly without calling them that. They ask friends whether an idea is good, count likes on a post, send a survey to a mailing list. The answers feel like evidence. Mostly they measure politeness, or how easily an idea can be understood in a sentence, which is a different thing from whether it will work.

Investors run polls of their own. A seed round that fills in a week because two well-known names have joined it is a vote, not an analysis, and the late arrivals are often trusting the early ones to have done the thinking. The same herd can leave just as quickly. A founder who reads a full round as proof that the business is sound has confused popularity with evidence, which is exactly the mistake Buffett was describing.

The exception is important. When the question is what people want, what they will pay or how they will vote, opinion is the evidence, and ignoring it is not independent thought but wishful thinking. A pricing decision made without asking any buyers is not braver than one based on a survey. It is simply less informed. Even then, what people say and what they do often differ, so the best evidence is behavior: preorders, sign-ups, repeat purchases.

Contrarianism is not thought either. Doing the opposite of the crowd is still letting the crowd decide. Buffett’s standard is harder than that: check the facts, reason from them, and then notice what others think only after the reasoning is done.