DESIGN INTELLIGENCEJuly 25, 2026·Mary · DEPIX Design Intelligence

QWERTY: The Worst Design That Won

Look down at the keyboard you are almost certainly reading this on. The top letter row spells QWERTY, an arrangement so familiar it feels inevitable, even natural. It is neither. It is one of the most consequential design decisions ever made, it was made for reasons that stopped applying over a century ago, and by most measures it is not very good. QWERTY is the canonical proof of an uncomfortable truth about the concept phase: the earliest decision doesn't just carry the most leverage - it can become permanent, whether or not it was ever right.

The layout was set in the early 1870s by Christopher Latham Sholes and his collaborators as they refined the first commercial typewriter. The exact rationale is debated, but the shape of it is clear: the arrangement solved a mechanical problem of that specific machine - keeping the type-bars of commonly paired letters apart so they were less likely to clash and jam. In other words, the keyboard layout we all still use was optimised for the failure mode of a mechanism that has not existed in mainstream use for fifty years. The constraint is long gone. The layout it produced is forever.

We know it is not optimal because people have built demonstrably better ones. The Dvorak layout, patented in 1936, puts the most common letters on the home row and balances the hands to reduce finger travel; Colemak is a modern, gentler re-optimisation. Studies on touch typing efficiency have gone back and forth on exactly how much faster these are, but almost nobody argues QWERTY is the best possible arrangement. And yet you type on QWERTY, your children will type on QWERTY, and the on-screen keyboard of a phone that contains no type-bars to jam faithfully reproduces a jam-avoidance layout from the age of steam. Why?

The answer is the most important idea the concept phase can teach: path dependence. Once a critical mass of people learned QWERTY, every downstream decision reinforced it. Typewriter makers standardised on it because that is what typists knew; typists learned it because that is what the machines had; schools taught it; every new device inherited it to be compatible with the last. This is a textbook network effect hardening into a classic case of lock-in: the cost of switching is not the layout itself - a phone could ship Dvorak tomorrow - it is the hundreds of millions of trained fingers and the universal expectation that any keyboard, anywhere, is QWERTY. The standard is the value now, and the standard is self-perpetuating regardless of the merit of what got standardised.

Here is why this matters far beyond keyboards, and why it should make every designer slightly nervous. We usually talk about the concept-phase decision as high-leverage: get it right and everything downstream inherits the clarity. QWERTY adds a darker corollary - the decision is often high-leverage and irreversible at the same time. A concept-phase choice that reaches critical adoption stops being a design and becomes an infrastructure, and infrastructure is defined by the fact that you can no longer change it cheaply. The keyboard is only the most visible example; the same dynamic governs technological lock-in everywhere, from file formats to plug shapes to programming languages to the width of railway tracks. In each case an early, half-arbitrary decision quietly became a permanent boundary on everything built after it.

The lesson is not "always get it perfect" - nobody can. The lesson is to know, at the concept phase, which decisions are the QWERTY kind: the ones that will be cheap to reverse next month, and the ones that will become load-bearing for a century the moment enough people adopt them. Those second decisions deserve a completely different level of scrutiny, because you are not choosing a feature, you are choosing a constraint your successors will inherit and curse. The tragedy of QWERTY is not that Sholes made a bad call; it is that a reasonable call for a 19th-century machine got frozen into the muscle memory of the entire species before anyone realised it had become un-undoable.

At Depix, this is the sharpest edge of the whole thesis. The concept-phase decision is where design has the most power - and, sometimes, where it quietly forfeits the power to ever choose again. The next time a decision feels small and early and reversible, ask the QWERTY question: what happens if this one wins?

Sources: