🦑
Designing beyond beige
November, 2025
Designers can’t afford to settle for “average”.
This has been coming up a lot in my coaching recently, especially when talking about designing in a competitive landscape alongside AI. It's a confronting message, but it also comes with a silver lining about more butts and weirdness… which I'll get to.
In 2024 (a decade ago in AI years) Emmet Connolly from Intercom gave this brilliant talk: Why Old UI Designs Won't Work with AI. Around the 16-minute mark he explains how AI raises the acceptance criteria for design — the “Age of Average”, as he calls it. The idea is simple: if AI can commoditize an acceptable design outcome, then the opportunity for designers lies in the non-average — in producing work that's bespoke, thoughtful, and opinionated. That stuck with me.
Then there's this gem by Ben Strak of Design Lobster — who, somewhere between tangents about Japanese culture, makes a compelling point about the need to go beyond what's considered “normal”:
“Once you know what normal is for your category you can intentionally flex your design along different dimensions to tell a story.”
This is exactly what's showing up in my conversations. Today's AI tools are trained to produce an averaged approximation of everything they've consumed. They generate an average landing page, an average signup flow, average code, an average article. It's not AI's job — or capability — to produce the bespoke or unexpected. And honestly, as Emmet points out, that's fine. For most people, “average” is already miles better than what they used to have access to.
But now AI can spit out a pretty decent idea and stitch together a sensible end-to-end UX. So where does that leave us? I can tell you for sure that if you're spending an entire sprint perfecting a forgotten-password flow, or a month thrashing on a landing page that Marketing needed yesterday… you're competing with a machine that can do that same job in 40 seconds, for free.
1's aren't enoughDirect link
At some point in my career, I started using what I call the “creativity scale” — there's probably a catchier name, but it's basically a mental model for grading ideas. It's a scale from 1 to 10, where 1 is the most obvious execution of a problem, and 10 is the most absurd, unfeasible, rule-breaking interpretation. Each point is an order of magnitude more inventive than the last… 2's make 1's look like garbage, and so-on. In the real world, nobody ships a 10. That's not the point. The point is that whatever number you start with — after research, feasibility, deadlines, scope, politics, entropy, whatever — you'll always ship something a few points lower. Start at a 6, you'll ship a 3.
1 is what AI will give you, 10 is a strange hole you owe it to yourself to explore.
And a 3 is game-changing in a world of 1's.
AI produces a solid 1 — maybe a 1.5 if you're really pushing it. And sometimes that's fine. Not every brief needs to be transcendent; the marketer who needs a quick ad or the team who needs a password reset flow by Friday would agree. But there are moments that call for magic, delight, and differentiation. That's where the bar has suddenly been raised. Thanks to AI, 1's aren't enough anymore. We've got to think bigger.
Take a simple example: you're a mobile designer briefed to come up with a way share data between two nearby phones. A creative 1 is whatever AI gives you: a button labelled “Share”, then some kind of “Sending…” modal. It's fine. It works.
On the other end of the scale you have touching two devices together and getting a full-screen ripple effect, haptics, acoustics — a full sensory assault. Apple absolutely didn't need to go that hard… but they did. And it's fucking cool. It makes me smile, and it instantly puts every “Share” button into the realm of boring sameness. This applies to marketing sites, brand moments, onboarding flows — basically anything someone can spin up using an off-the-shelf AI tool.
And I can already hear the pushback — there's no ROI on 10's! You say. Leadership isn't exactly throwing money at uniqueness over fast and cheap. Fair.
But indulge me in another example from my time at Campaign Monitor — an email marketing tool like Mailchimp and a thousand others. The UI to support “build a list, send an email” is pretty dry across the board. So we made a feature called Worldview — a feature that appears when you send a campaign, and plots all activity like opens, shares, forwards and unsubscribes onto a Google map in real-time.
Utterly pointless! Provided zero utility that couldn't be shown in a table. But holy shit, it was cool. Seeing Sarah from the Falkland Islands open your email? Pure dopamine.
Thousands of customers flocked to Campaign Monitor because of Worldview. People shared their maps everywhere. And customers who churned eventually came back because everything else felt flat and joyless in comparison.
The future of softwareDirect link
In the past, customers chose between either your product or your competitors. “Good enough” meant parity; “differentiation” meant going a little bit further than the next guy. No need to push until someone else did. Occasionally someone would leap ahead and reset expectations — the kind of thing Bezos referred to as turning “yesterday's wow into today's ordinary” in his 2017 shareholder letter (citing Amazon, ew, so sorry).
Remember when every client wanted their version of Apple.com, or “Uber for…” or when suddenly every landing page looked like Notion? Those players were cooking, and those moments rare. The commercial incentive to invest in creative excellence wasn't strong when keeping up with competitors was enough.
AI enters the chat.
Now we're back to the threat of normal — except more existential. Your customers aren't just choosing between you and your competitors; they can now spend a weekend vibe-coding their own bespoke solution with Lovable. Suddenly your product isn't just competing with the market — it's competing with an infinite incoming tide of hand-rolled, AI-assembled, “good enough” alternatives. In a perverse way, this scenario is what excites me the most, because it's opportunity to change gears.
Because you and I both know those vibe-coded tools — and your closest enterprise competitor — are only ever going to produce average, predictable experiences. Which means the value of anything delightful, unique, cleverly crafted or downright weird just shot up. There’s your ROI.
We're in a moment where anyone can produce something passable in minutes. That invites creative 10's. It encourages designers to reach higher: mobile ripples, pointless campaign maps, brands centered around agressive owls and mascot monkeys with butts, bank accounts where you swipe to save coins, and landing pages that go unnecessarily hard (touché, Microsoft). Delight isn't new — but suddenly it matters more than ever.
Reality checkDirect link
Okay back here in reality, I concede that not every brief calls for weirdness or surprise. Most designers are already overworked, undervalued, or both — and most companies will always prefer “done” over “special.” I get it. In that regard, my advice to design teams has remained the same for years — we cannot be the bottleneck.
But the secret to success (and professional fulfillment) is knowing where to put your effort. When good enough is truly good enough, ship it and move on. But when there's an opportunity, appetite, or even a sliver of space for something more, consider what your work might look like articulated higher on the creative scale. Even a 3 or a 4. In thinking beyond ‘normal’, you might find yourself fighting for something exceptional.
To help the designers I coach think this way, I've shamelessly taken Jony Ive's rules for designing products that people love and reframed them for a post-AI world. In my experience, these questions can nudge work from “good enough” toward something that actually leaves a mark. I'd love to know if it helps you too.
