The Problem with Designing for the Average User

The average user does not exist. This is not a philosophical point; it is a measurement fact. When you describe the average height, income, and reading level of a population, the percentage of individuals who match all three figures simultaneously approaches zero. The average is a statistical artifact that describes no one, and designing for it produces products that fit everyone moderately well and serve no one exactly.

The US Air Force discovered this in 1950 when a researcher named Gilbert Daniels measured ten physical dimensions of over four thousand pilots and found that not a single pilot fell within the average range across all ten. Cockpits designed for the average pilot were, in a precise sense, designed for nobody. The solution — adjustable seats, adjustable pedals, adjustable everything — is now so obvious it seems like it was always the plan. It was not.

Software design still largely designs for an imagined average. The user is assumed to have a certain level of literacy, a certain willingness to read documentation, a certain tolerance for onboarding friction, and a certain baseline of prior experience with similar software. When products reach populations outside that assumed average — elderly users, non-native speakers, people with disabilities, people in lower-bandwidth environments — the failure is attributed to the users rather than the design assumption.

The alternative is not to design for every edge case simultaneously, which is impossible, but to design systems that can be configured and adapted rather than systems optimized for a single point. Adjustability costs more at the design stage and returns its investment in adoption reach and reduced support burden.

The companies that learn this tend to learn it from an expensive failure with a population they had not designed for. The lesson usually travels slowly.