Frederick Winslow Taylor timed workers with a stopwatch in 1911 and concluded that efficiency was a matter of eliminating variance. He was spectacularly right about factory floors and spectacularly wrong about everything that came after. The psychologist George Kelly spent the 1950s developing Personal Construct Theory — the idea that every person operates like a scientist, running private experiments and updating their model of the world through what he called 'construing': the act of interpreting events through a lattice of personal contrasts. Kelly's insight that cuts directly into product work is this: people don't resist change because they're irrational — they resist it because the change invalidates constructs they've built their competence on. Your users aren't slow adopters. They're scientists whose experiments have already returned results, and your product is asking them to throw out their data. The practical move isn't to explain features more clearly or reduce onboarding friction. It's to design transitions that let users carry their existing constructs forward — that honor what they already know how to do — before asking them to build new ones. A roadmap that ignores construct continuity is just Taylor with a Figma license.

In the last product decision you made about onboarding or a feature transition — what did you assume users would be willing to abandon, and did you ever actually ask whether that thing was load-bearing for them?

Drawing from Constructivist Psychology / Kellyan Personal Construct Theory — George Kelly (The Psychology of Personal Constructs, 1955)