On good design

I wanted to write something up as a reaction to Dieter Rams’s famous ten principles of “good design”—or at least what I thought they said, because after rereading them, I disagreed less than I remembered. The principles are fine, but I still find them too focused on the outward interface of a product, even its looks. This might do for a radio, but I don’t think it’s a wide enough view for the world we find ourselves in today—not for the complexities of networked software, not for coming to grips with the second- and third-order effects of the things we make on our twitchy, new, overconnected planetary nervous system. So here’s my stab at it.

Good design serves people.

A good design acts only on behalf of the user—and only at their bidding. It is not a beachhead in a market land grab. It is not a way to connect a company to a user for purposes of extraction. It is not a flywheel in a network effects play for dominance. It does not try to coerce, cajole, or manipulate the user. It does not try to anticipate, and thus have undue influence upon, human needs. It is not showy to the detriment of being able to use the thing.

Good design is a show of respect for the user: for their intent, their battery life, and their budget.

Good design is good fit.

A form exists not by itself but as part of an ensemble with everything around it, and so a form is only as good as the quality of its fit to its environment. Thus “design” is not some property self-evident in a single object. No one thing is a self-contained universe of pure user experience. Something is well-designed only when it fits well into the ecosystem of people, practices, and other forms that surround it.

To this end, a good design embraces standards. Standards plug an object into an ecosystem, opening up generative possibilities.

Good fit includes sustainability, because a form’s fit to the context of our planet is its ultimate fit. Systems that use an outsized amount of energy relative to the benefit they give society are bad fit. Gadgets with planned obsolescence are bad fit.

Good design is holistic.

Good design is not merely good graphic design. It’s not just about the visual. Good design pervades the entirety of a thing, its internals and externals, and how they relate to many kinds of people: end users, people around the end users, operators, repairers, builders.

Under this heading go admin-facing screens, software architecture, runbooks, and repair manuals.

Good design is open.

Good design doesn’t close the door on uses that the designer couldn’t imagine ahead of time. It doesn’t second-guess you or try to be “smart” (that is, to optimize for a single kind of use, at the price of getting it wrong when you step outside of the designers’ imaginations). It has maintenance hatches and a transparent implementation, ensuring that anyone can pick it up and work on it, maybe even the person using it.

This is also how you build for longevity; see the Right to Repair movement.

Good design builds on what came before.

Good design can be innovative, but it usually doesn’t have to be. Useful innovations advance practice for everyone and get reused time and again; that’s the definition of progress. Good designs make use of the successes of the past.

Good design is long-lived.

Dieter Rams is dead-on here—you should get many years of use from a single good design. The careful designer prevents churn. A good design has high return on investment. (Again, this ties back into the environment.)

The way to build for longevity is to build for transparency. When you pile up abstractions too high, you lock yourself into a particular set of assumptions from a particular moment in time, because abstractions create dependencies. They’re hard to undo. You must be willing to endure some duplication to avoid the wrong abstraction.

As such, a careful designer is averse to dependencies. He or she tries to ensure that the dependencies they take on are going to be around for a long time.

Good design works with the grain of its materials.

Working with your materials means hewing closely to what they naturally do and not piling up artifices that distort and denature them.

What you get from a cutting-edge kind of build may be remarkable (“wow, I didn’t know X could do Y.”) But this kind of design is more likely to wind up as a gimmicky technical demo than something people want to use. When you push your materials too far, they yield no more flexibility to build upon.