An idea kept off the market runs short of air
Real users react in ways no team can predict, so time spent polishing in private is time spent guessing. Some products cannot afford that kind of learning.
Usage is like oxygen for ideas. You can never fully anticipate how an audience is going to react to something you’ve created until it’s out there.
WordPress 2.0 came out at the end of 2005. The next major version did not arrive until early 2007, more than a year later, as the project kept adding to a release that was never quite finished. Matt Mullenweg, who started WordPress and runs Automattic, the company behind its hosted service, wrote last November in an essay on releasing early that the long gap taught him how costly it is to hold work back. Features sat unused, feedback never came, and the team spent months refining things users might not have wanted.
His comparison is to breathing. An idea shown to no one cannot find out what is wrong with it. The team can test, discuss and polish, but every judgment is a guess about how strangers will react. Once the product is in use, the guessing stops. Some features nobody touches; others get used in ways nobody planned.
The line holds for most software, especially software that can be updated quickly. A web service can ship a change in the morning and remove it in the afternoon. The cost of releasing something imperfect is small, and the cost of learning slowly is large.
It holds less well where a flawed release causes lasting harm. Software that handles money, medical data or safety systems should not reach users with known defects, because the users bear the cost of the learning. A product launched loudly to the press also gets one chance at a first impression, and a rough version can fix a poor reputation for years.
The distinction is between usage and launch. Mullenweg’s argument is about getting real people using the thing, which can happen quietly, with a small group who know it is early. That supplies the oxygen without staking the reputation. Most products can find some group willing to use them early, and the ones that cannot tend to be the ones that should wait.