On Product Design

The hardest thing to defend is work nobody can see.

I’ve been a Product Designer for about a decade.

Long enough to have heard the same conversation cycle through three or four times, each time with a new villain—first it was stakeholders who thought design was common sense, then it was no-code tools, now it’s AI.

The argument is always the same underneath: I can see what good design looks like, I can replicate it, why do I need you?

It’s not an unreasonable conclusion to draw. The visible part of design—the layouts, the colours, the components—is replicable. Always has been. Copy the pattern, ship the screen, move on.

The decoration misconception—designer as someone who picks colours, makes things pretty—has always been the wrong read too. Design is how something works.

What both miss is that the visible part is the smallest part.

The work that actually determines whether a product succeeds or fails mostly happens before anything is drawn, and mostly doesn’t look like design at all.

It’s in the questions nobody thought to ask, the friction that got removed before the user ever felt it, the decision that quietly made three other decisions easier.

That work is invisible by definition. When it’s done well, nobody notices. When it’s missing, people just say the product feels off and can’t tell you why.

I spent time on a project once building out proper keyboard navigation—full interaction patterns, tab order, shortcuts that actually matched how power users moved through the workflow.

Stakeholders didn’t notice.

That was the point.

The people using it every day got incrementally faster, made fewer errors, felt less friction without being able to name why.

Nobody filed a ticket saying the keyboard navigation changed their life. The product just worked better. That’s the invisible work.

And that’s also why, ten years into the industry taking Product Design seriously, there are still so many products with poor experiences—not for lack of trying, but because the work that actually matters is the easiest to cut when nobody can see it. Teams move fast, skip the thinking, ship the screen.

AI is about to make it much more common.

Which sounds like a bad time to be a designer.

The deeper issue isn’t whether AI can do design. It’s that most AI feature development was never driven by a user problem in the first place. It was driven by capital.

Companies took on significant investment in AI and needed to show returns, so features got shipped—not because users asked for them, not because research pointed to a gap, but because the money was already in.

That’s solutionism with a very large budget behind it. And solutionism at scale, without design thinking at the centre of it, produces exactly what we’re seeing—products that are technically more capable and experientially worse.

The invisible work got skipped, not because anyone decided it wasn’t valuable, but because the incentive was speed and the pressure was financial.

Designers weren’t cut from these processes because AI replaced them. They were cut because nobody stopped to ask what the user actually needed.

I’ll admit the bias. A decade in a discipline makes you protective of it. But I’ve spent enough time adjacent to product management and engineering to have some perspective—and I’m not arguing designers are more important than anyone else.

I’m arguing that the work designers do, the invisible kind, is about to become the thing that actually differentiates products from each other. Everything else is getting faster and cheaper. Judgment isn’t.

I oscillate on this. Some days it feels obvious, some days the speculation is loud enough to drown it out. But the underlying logic holds.

The automation pressure is real.

So is the gap it’s going to leave behind.