Any competent engineer will tell you that good engineering depends on understanding materials. If you misunderstand your materials the system will fail.

Materials must be understood with highly nuanced specificity. Two metals that look and feel the same to an untrained eye and hand might behave drastically differently when subjected to friction, heat or strain.

Design has a similar respect for materials. Design, however, also includes categories of “material” that are not physical — ones that engineers typically factor out, namely subjectivities.

Where an engineer sees a car as a system, a designer is trained to see the car plus the driver as the system — a system with both physical and psychological components. Same with a retail space: a system of objects, surfaces, light, sound, merchandise, customers, sales personnel. Or a hammer — a system that includes an arm, a hand, a nail, a board and a metal form. And with the insight of Design Thinking — (the realization that all systems that include and depend on people are actually best approached by methods used to solve design problems) — things that aren’t normally thought of as “design” are viewed through a design lens — like organizational structures, or business processes, or political policies.

How do you recognize a design problem? If a system being developed will succeed only if people cooperate or participate or behave in some particular way, your problem is a design problem. Approaching such problems as a design problems, using people-centered design methods, will increase your odds of success.

But doing the usual, and treating design problems as if they were engineering problems by failing to factor people in at all, or making uninformed generalizations of “how users are” (or “how women are” or “how female general managers are”, etc.) will get you exactly the kinds of results as you’ll get if you specify “metal” for your engine block. You might get high-carbon steel. But you are just as likely to get tin.

Of course, you can reinforce a tin engine block with a different metal and maybe get it to function, but you’ll never get the level of quality as you would if you considered the material from the start.

Likewise, trying to add design after something has been fundamentally developed for nobody in particular will lead to a mediocre solution is unlikely to be embraced by anyone with much enthusiasm.

Or to put in another way: engineering your way through a design problem is bad engineering.

Know your materials!