Pretty isn’t a KPI

We’ve never met a client who asked for an ugly website. We’ve met plenty who got a beautiful one and quietly wondered why nothing changed.

So here’s the uncomfortable thing, coming from a shop full of designers: pretty isn’t a KPI. No board deck has a slide for gorgeous. Revenue has a slide. Signups, retention, qualified leads, time-to-close: those get slides. “The site looks incredible” is a compliment, and compliments don’t compound.

A gorgeous site that doesn’t move a number is a poster.

A very expensive poster that took four months, two workshops, and a naming exercise for the design system.

Easy to approve

How does this keep happening? Because beauty is easy to approve. It demos well in a conference room. Everyone can see it, nobody has to wait for it, and signing off on it feels like progress. Outcomes are slower and ruder. They require instrumentation, patience, and the possibility of being wrong in public. Given the choice between a launch that looks like a win today and a metric that might embarrass us in six months, most rooms pick the launch. Quietly. Every time.

So design reviews drift toward the room instead of the user. Does it pop. Is it premium. Can we see it bigger. Meanwhile the only review that matters is happening later, alone, on a phone, in a checkout flow, where nobody says pop and everybody says nothing.

What beauty is for

Now the other half, because this is not an argument for ugly.

We love beautiful work. We make it on purpose and we’d be lying if we said craft doesn’t matter for its own sake around here. But beauty in commercial work isn’t decoration. It’s labor. It builds trust in the first few hundred milliseconds, before a single word gets read. It makes a complicated product feel comprehensible, which is another way of saying it lowers the cost of understanding you. It’s a big part of why two companies can sell the same thing and one of them gets to charge more.

That’s beauty doing a job. Beauty with receipts.

The distinction isn’t pretty versus effective. It’s accountable versus ornamental. The same visual decision can be either one, and the difference is whether anyone bothered to define what it was hired to do.

The number question

Which is why, before we design a pixel, we ask one question: what number is this supposed to move? Not in a cynical way. In a load-bearing way. Sometimes the answer is conversion. Other times it’s qualified pipeline, or support tickets that stop coming, or the caliber of candidate who applies. And occasionally the honest answer is pricing power, which is hard to measure and still very real.

But if nobody in the room can answer at all, that’s not a design brief. That’s a strategy problem wearing a design brief’s clothes, and no amount of typography will fix it.

The follow-through matters just as much. If work is going to be held accountable, it deserves a fair trial: a baseline before launch, a real measurement window after, and the humility to change the work when the number disagrees with the room. Holding design accountable cuts both ways: it also means letting the wins count, so the next brave decision is easier to approve than the last one.

We’ll keep making beautiful things. That part isn’t negotiable; it’s most of why we exist. We’d just rather make the kind of beautiful that shows up in a spreadsheet eventually.

Make it gorgeous. Then make it prove it.

Posters belong on walls.