There is mythological figure invoked whenever someone needs to defend the indefensible: the average user. It haunts every product meeting of the last decades, appearing in whiteboards and specifications and justifications for questionable decisions but… he was never actually interviewed, nor observed in the wild, yet is still always present in aggregate statistics and has become the perfect excuse for what I would call, without euphemism, the systematic infantilization of software design. This can, and I personally believe it too, be expanded to other areas of design but I digress. “We did this for the average user”, they say, but what they’ve done is essentially remove options, hide settings, replace clear menus with ambiguous icons, transform functional interfaces into treasure hunt, all in service of this phantasmagorical entity that conveniently always wants exactly what the company decided to deliver anyway.
Let’s start with the obvious: there is no average user in the same way there is no person with 2.3 children. It’s a statistical abstraction that evaporates as soon as it’s attempted to be materialized. There are novice users, expert users, users with specific needs, users with peculiar workflows, users who employ a product in ways never imagined, etc. What there is not is someone perfectly positioned at the center of all these distributions, some platonic ideal of mediocrity who represents everyone by representing no one in particular. But admitting this would be inconvenient because if there’s no average user, then there’d need to be something infinitely more laborious: create systems that serve different levels of proficiency simultaneously, with layers of complexity that present simple interfaces by default but offer depth for those who need it, with real discoverability rather than merely polished empty surfaces. That takes work, and work is expensive. That, too, required careful thinking about information hierarchy, progressive disclosure, how to serve both the beginner and the expert without alienating either — much easier to point at the “average user” and declare, with the confidence of someone who’s never actually met them, that they don’t need this feature, don’t want complexity, or only desire simplicity.
There is a fundamental dishonesty in this conversation that needs exposing — when companies say “we simplified for the user”, what they frequently mean is “we simplified for ourselves”. Fewer options means less code to maintain and combinations to test, fewer configurations means that and less customer support overhead. It’s essentially an operational cost reduction dressed up as humanitarian concern, and the costume isn’t even that convincing if you look closely. Consider how Safari on iOS for years had no robust tab management and no extensions, because the average user doesn’t need that, presumably. But then you’d look at Android where Chrome had all of it from the beginning, and guess what? People used it. Regular people. Also, Windows 11 centralized the Start menu because “research showed” users preferred it that way — which users, exactly? The aesthetics department? The marketing team’s vision of modernity? Certainly not my workflow, nor that of countless professionals who used icon positioning as spatial navigation, a technique that goes back to the very foundations of graphical interfaces but was apparently too complex for our fictional average friend.
Part of the problem is that modern design has become obsessed with the first impression, and to dazzle quickly, you simplify, remove friction, make everything seem obvious and immediate, but products are enduring creations, and the user that was impressed with simplicity on day one will be, six months later, furiously clicking around because they can’t accomplish something that should be trivial because the new “simplified” interface never anticipated their specific use case since it was designed for someone who doesn’t exist. As Don Norman called it — affordances — a thing’s actual and perceived properties that indicate how it can be used, offering clear clues to possibilities without labels, well-designed software leverages these to age with the user, starting simple but revealing depth gradually and has a consistency that allows knowledge to transfer between contexts and has, above all, respect more than what a focus group said was “intuitive enough”. The “average user” excuse assumes the opposite: that people are static, incapable of learning; permanently beginners.
A case could be made that invoking the “average user” is almost always a symptom of design laziness because it abdicates the difficult task of creating truly adaptable systems in favor of a solution that’s “good enough” for most, a.k.a. another way of saying it’s not quite right for anyone. The real cost to “presuming” interfaces is the gradual erosion of agency.
Leave a Reply