Now that machines can pretty much make anything, one principle remains unassailable: make the right thing, and make the thing right. Getting to that outcome involves uniquely human input. Some days that means a board conversation about what change we are actually trying to make and why. Other days it means sitting next to a designer on a detail almost nobody will consciously notice. Here, I lead up as hard as I lead down, because a great idea dies in the gap between.

The thing I’ve seen repeatedly is that the ask and the actual need are rarely the same thing. Someone asks for a page redesign, but the real question is, “What might be influencing our traffic that we can’t see?” The ask is to minimise loss on a shop floor, but the real question is, “What solution won’t make well-meaning customers feel like suspects?” A brand requires a new look but, “What else could this organisation mean to people?” might be a better question to ask. The work is finding out whether something different is happening underneath what looks obvious.

The days I count as the best are the ones where the people around the work got better at it.