Talking about a project is not starting it
Describing a plan to friends feels like progress. There is some evidence that it substitutes for progress, and also some cases where the conversation is the first step.
Time’s a wasting - don’t talk about wanting to do something, just start doing it.
Talking about a plan produces a small reward of its own. Friends nod, someone says it sounds great, and the speaker leaves the conversation feeling as if a step has been taken. In 2009 the psychologist Peter Gollwitzer and colleagues published experiments suggesting that this is more than a feeling: people who announced an intention tied to their identity, and had it acknowledged by others, then worked on it less than people who kept it to themselves. The acknowledgment seemed to satisfy part of the goal in advance.
Garrett Murray, a designer and developer who founded the iOS studio Karbon in 2010 and co-produced the Maniacal Rage podcast, is credited with the line above. No earlier source for it has turned up, but it fits the way independent developers of that period tended to work: small products, built in the evenings, shipped before anyone had a chance to discuss them at length.
Starting also changes the conversation itself. A person who has built a rough version of something talks about it differently: about the specific part that broke, the screen that confused a tester, the question a first user asked. That kind of talk is useful to everyone involved, where talk about wanting to build something is mostly pleasant.
The advice is sound for a common kind of project. A short film, a side application, a blog, a first prototype: in each of these the cost of starting is a few hours, and nothing learned from talking about it matches what is learned from a first rough version. People who spend months describing a project they have not begun usually discover, when they finally begin, that the hard parts were not the ones they had been discussing.
Talking is sometimes the work, though. A founder testing whether anyone will pay for a service learns more from twenty conversations with potential customers than from building it first. Anyone who needs a co-founder, a license or a loan has to talk before doing, and skipping that step leads to building something nobody asked for or starting something that cannot legally continue. The line is aimed at talk that replaces action, not talk that informs it.
A simple test separates the two. Useful conversation changes what gets built. Conversation that leaves the plan exactly as it was, and mainly produces encouragement, is the kind the line warns against. The remedy in that case is to set a date, however near, and produce something small enough to show by then.