Most Business Mistakes are not Fatal

An Important Lesson

Scott Alexander writes,

Musk is a paradox. He spearheaded the creation of the world’s most advanced rockets, which suggests that he is smart. He’s the richest man on Earth, which suggests that he makes good business decisions. But we constantly see this smart, good-business-decision-making person make seemingly stupid business decisions. He picks unnecessary fights with regulators. Files junk lawsuits he can’t possibly win. Abuses indispensable employees. Renames one of the most recognizable brands ever.

…The main answer to the paradox of “how does he succeed while making so many bad decisions?” is that he’s the most focused person in the world.

Nope. The main answer is that in situations with a lot of ambiguity, you do not have to meet the standard of making zero mistakes. You just have to make less costly mistakes than the other guy.

Did Bill Gates make mistakes? Clippy, anyone?

Gates was slow to “get” the Internet, but he still won the competition with Netscape. It wasn’t because he bundled worse products with his near-monopoly operating system. Netscape had the worst products. I know, because I was running a Web business at the time. We would try to take advantage of a new standard that had been proposed by Netscape—and find that it worked on Microsoft’s browser and not on Netscape’s. We tried to use the Netscape server for our business, and it continually crashed under heavy load (Netscape itself never used its own server for its web site).

Speaking of my web business, it was somewhat of a success, but the story I told about it is A Sequence of Miscalculations. If a few bad decisions are enough to ruin a company, then our business should have been a total flop.

As I wrote here, the game lasts more than one move. In chess (or, in my case, Othello), the best players make mistakes. We joke that the winner of the game is the one who makes the next-to-last mistake. As a highly-rated player, when you take on a lower-rated player, you can count on that player to make some costly mistakes (and if by some chance he doesn’t, you will suffer a rare upset).

Or think about war. Military historians can find countless mistakes made by generals on both sides in a conflict. In real time, nobody is able to see through the fog of war.

Two Types of Errors

For many decisions, errors can go either way. You could sink money into a company that fails. Or you could miss out on investing in a company that takes off.

Suppose you post something to Facebook that might get it in trouble with an important constituency, such as the European regulatory body. If Facebook fails to take it down and they are penalized, that is a Type I error. If Facebook takes it down and it was innocent and non-controversial, that is a Type II error.

A basic lesson of statistics is that the more you try to avoid Type I errors, the more likely it is that you will make Type II errors, and vice-versa. People who forget this tend to flounder.

For example, in mortgage lending, a Type I error is making a loan with a higher risk of default than is desirable. A Type II error is turning down a loan with a reasonably low risk of default. In the early 2000s, politicians accused lenders of making Type II errors. Lenders were allegedly depriving deserving families, especially minorities, of the opportunity to own a home.

When the housing market crashed in 2006-2008, the politicians turned around and blamed lenders for making Type I errors. They were allegedly guilty of making predatory loans to families, especially minorities, who could not possibly have handled the payments.

Since 2008, mortgage credit has become tight, in response to the punishments and regulations that were handed down at that time. Now, with lenders afraid to make Type I errors, it is likely that Type II errors have become prevalent.

Rules are pearls

Bureaucracy and rules are a reaction to error. Just as a pearl grows inside an oyster around an irritant, a rule grows in response to a mistake. How many people have had to take their shoes off at the airport and have them run through a scanner because of one incident? Small irritant, big pearl.

Elon Musk has what he calls an “algorithm” that strikes me as a defense against bureaucratic overgrowth. The WSJ writes,

Musk urges his employees to:

• Question every requirement

• Delete any part or process you can 

• Simplify and optimize 

• Accelerate cycle time

• Automate

All of these points run counter to bureaucratic stagnation and aversion to mistakes.

Of course, rules and bureaucracies do serve a useful purpose, or else they would not survive. They can prevent middle managers from making mistakes that are fatal to the organization. They can provide a sense of security to important constituents, including top executives, Board members, and regulators.

You should be careful about questioning how a decision-maker can succeed while still making mistakes. Do not hold a decision-maker to the standard of making zero mistakes. To succeed, it is sufficient to make less costly mistakes than those that someone else would make when faced with an equally complicated array of choices.

substacks referenced above:

@

Book Review: Elon Musk