Filip Hráček / text /

Performance vs Preference

One of the most powerful ideas that I keep coming back to is a principle called “Performance vs Preference.” I believe that anyone who’s building anything (be it a programming language, an app, a game, or a piece of furniture) should be aware of this.

There’s a book called Universal Principles of Design by William Lidwell, Kritina Holden, and Jill Butler, first published in 2003. I bought a tiny pocket version of the book for $13 at Books Inc. and I've kept it close to my workspace ever since. It’s full of great stuff that I love to revisit. But the section on “Performance vs Preference” probably stuck with me the most.

Since the two pages are nowhere to be found on the internet, let me reproduce them here.

Performance vs. Preference

Increasing performance does not necessarily increase desirability.

People often believe that a functionally superior design—the proverbial “better mousetrap”—is good design. This is not necessarily correct.

The reasons people favor one design over another is a combination of many factors, and may have nothing to do with performance, but with preference.

Preferences may be based on innate tendencies, cultural biases, aesthetic or emotional considerations, or legacy practices and conventions.

Success in design is multivariate. Consider both performance and preference factors in design to maximize the probability of success. Beware the trap of creating a superior product in one dimension, but having it fail due to neglect of other dimensions.

Worse is better

I find this a really insightful concept, similar in a way to the old "worse is better" discussion. Back in the late 1980s, some Lisp people were perplexed by the fact that people preferred the C programming language even though it was (from their perspective) obviously worse than Lisp.

People still fall for this same fallacy, even 36 years later.

Dvorak is “obviously” better than QWERTY.
Functional programming is “obviously” better than imperative programming.
One IDE is “obviously” better than another.

Well, no. To quote the book, “success in design is multivariate”. There are several dimensions to look at, not just one better-worse line.

And, crucially, one of the dimensions is what people already know. What they like. Their pre-existing preference. Even if you can prove that your thing is plainly superior, better thought out, more advanced, more performant, whatever, you still can’t expect it to be automatically more popular. People won’t drop everything they’re doing in order to learn your new, better ways.

They might, sure. Especially if your thing is significantly better. But don’t be as surprised as the Lisp guys of yore if they don’t.

MAYA

Another related term (also covered in the book) is MAYA, which stands for Most Advanced Yet Acceptable. This was coined by Raymond Loewy (1893-1986), the legendary designer behind things like the Coca-Cola bottle or the Shell Oil logo. He said:

“The adult public’s taste is not necessarily ready to accept the logical solutions to their requirements if the solution implies too vast a departure from what they have been conditioned into accepting as the norm.” (source)

There you go. Even if your thing is a “logical solution to [people’s] requirements”, they might not be ready to accept it. Always weigh in people’s current preferences, what they’re used to, what’s “the norm”.

I can’t tell you how often I come back to both Performance vs Preference and MAYA. Internet discussions are full of well-meaning people who clearly haven’t heard about either, and it shows.

— Filip Hráček
January 2025