Failing at something beats never trying it, most of the time
Over a long life, people regret what they did not try. That holds only when the attempt was survivable.
The only thing worse than starting something and failing… is not starting something.
In the 1990s the psychologists Thomas Gilovich and Victoria Medvec asked people about their regrets and found a consistent pattern. In the short run people regret mistakes they made. Looking back over years, they regret far more the things they never tried. Seth Godin’s line compresses that finding into a rule for anyone deciding whether to begin.
Godin has spent the past few years, and much of his new book Linchpin, arguing that the hard part of making things is shipping them, and that fear dressed up as prudence stops most projects before they exist. A failed attempt at least produces something: a lesson, a skill, a contact, a clearer sense of what not to do next time. An idea never started produces only the question of how it would have gone.
The line holds best where starting is cheap. A weekend prototype, a short newsletter, a small shop online, a first sale to a single customer. None of these risks much, and each teaches more than months of planning.
It holds worst where failure is expensive or lasting. Starting a company with borrowed money, a personal guarantee or a family’s savings is a different decision from starting a side project, and the regret research says little about it. Someone with no dependents can absorb a failure that would break a household. The line assumes the failure is survivable, and for some starts it is not.
There is also a way of starting that is worse than not starting: starting a dozen things and finishing none. Each start feels like progress while the pile of abandoned work grows. That is a failure mode the line does not warn about, and it is common among people who took the line to heart: the habit of beginning is easy to build, the habit of finishing much harder.
The useful version is to make starts cheap. Decide before beginning what the attempt may cost in money and time, write the limit down, and stop when it is reached. A start with a known ceiling is almost always better than none.