Design is a decision, not a job title.

Somebody decided three tickboxes and one confirmation prompt were enough to move nine hundred million dollars. Nobody called it design.

Start with a mistake, and decide whose it was

Design is deciding what a system makes easy, hard and impossible

Every system already has those three properties, whether or not anybody chose them. A form has them. So does a contract, a timetable, a purchasing specification and the choice of which number a team is measured on.

The only question is whether somebody sat down and decided them on purpose. Mostly, nobody did.

This is wider than the everyday definition, deliberately. The test of a definition is what it lets you see. This one makes the Citibank screen a design decision, which it plainly was, and which no narrower definition can account for.

The decisions get made anyway, by whoever happens to be holding the pen.

A badly drawn screen isn’t needed for a disaster

From 1999, the British Post Office ran its branches on an accounting system called Horizon. It showed shortfalls of money that had never existed. That was a software defect, of the kind every large system has, and engineers could in principle have fixed it.

The disaster came from two decisions above the defect:

• Subpostmasters were made personally liable for any shortfall the system reported.
• In court, the computer’s record was treated as more reliable than the person in the dock.

No screen in that story was badly drawn. What ruined those lives was a decision about whose account of reality would count. Lawyers and executives took it, and none of them would have recognised it as a design question. It set what was easy, hard and impossible for hundreds of people, and nobody ever examined it on those terms.

So design is a kind of decision, not a profession

The word covers two things that overlap less than it suggests.

• A job, with schools, portfolios and a labour market.
• A role, held by whoever is actually deciding what a system makes easy, hard or impossible.

At Citibank the role was almost certainly held by a developer implementing a written requirement. At the Post Office, by lawyers and executives. Designers are people who sometimes get to make these decisions. In most organisations, they don’t.

Every role did its job, and one question had no owner

The flattering version says designers care about users and other roles don’t. That is false. Engineers care about users. Analysts care about users. Nobody sets out to build the wrong thing.

The useful version: roles differ by the question each one owns.

Boeing’s MCAS software on the 737 MAX worked exactly as specified. It read its sensor, judged the aircraft to be stalling and pushed the nose down. It read one sensor, though the aircraft carried two, and nothing checked one against the other. The safety case assumed a crew would diagnose a misfire in about four seconds. 346 people died in two crashes.

The engineering was not defective. The specification was. Somebody needed to ask whether one sensor was enough, and what a crew would face meeting this for the first time with three other alarms going. That question appears to have belonged to nobody in particular.

Neither column contains a mistake. The design question is whether the measure was right in the first place, and you cannot answer that from inside the measurement.

Name the question, and give it a deadline

On a good team it barely matters who asks the design question, only that somebody does. Plenty of engineers ask it. Hiring a designer is one way to make sure it gets asked. It isn’t the only way, and it doesn’t work on its own: plenty of teams have designers and still never ask it, because the designers arrive after the specification is locked.

The reliable version is plainer. Name the question, attach it to a decision point, and make somebody answer it in writing before the money is committed.