Skip to the page

Plainly Put

One sentence about building things, taken apart

HomeExecution

Shipped work is the most credible reference a person has

A record of things built and shipped persuades better than any self-description. Some valuable work, though, makes no sound at all.

Your work speaks for you.

Michael Lopp

Michael Lopp has written for years under the name Rands, on a blog called Rands in Repose and in two books about managing engineers, Managing Humans and Being Geek. His subject is the culture of people who build software, and this four-word line captures one of its core beliefs: what a person has made matters more than what they say about it.

In a good engineering team the belief is close to true. Code is visible. Colleagues can run it and see whether it breaks. A developer who ships a clean, reliable feature earns more respect than one who talks well in meetings, and a portfolio of real projects is the most persuasive thing a candidate can bring to an interview. The same holds for designers, writers and anyone whose output can be examined directly. For people starting out, the advice is practical: finish things and put them where others can see them.

The line also works as a check on self-promotion. A founder who spends more time describing the product than improving it is relying on words to do work that the product should do. Users do not read the press release; they use the thing, and it either speaks for itself or it does not.

The limits appear as organizations grow. In a company of five hundred, most people never see most of the work. Credit tends to go to those whose work is easy to notice and easy to explain, such as a launch, a demo or a big customer. Much of the most valuable work is neither. Keeping systems running, fixing slow processes, mentoring new hires and preventing disasters that never happen all leave little to point at. People who do that work well can be passed over while louder colleagues are promoted.

Work also needs context to be understood. A small change that made the site twice as fast will look like a tiny commit to anyone who does not know what it did. Someone has to explain it, and if the person who did it will not, a good manager should.

So the line is right about what counts and too hopeful about who notices. Let the work carry the argument, and make sure the right people know where to look.