Why One Website Gets Remembered and Ten Don't

Someone looks at your website and drops one word: AI slop. And it's not the first time you've heard it — in a comment somewhere, from a designer friend, from a client who can't quite explain what's wrong but feels it anyway. First reaction: offense. Second: the urge to rebuild everything immediately. Third, if you manage to get there while still calm, is a better question — what would actually have to be true about a site for that feeling to go away?

Then the renovation starts. Fonts get swapped, blocks get rearranged, the "unique value proposition" gets bumped up a size — and somewhere in the middle of all that, if you're lucky enough to stop and think, a more interesting question shows up: does any of this actually work? Does making a site a little strange do anything at all, or is it just another way to spend a weekend repainting a fence that was fine to begin with?

The instinctive answer is to add character. A bit of strangeness, a voice, something a competitor wouldn't have. And this is exactly where it's worth slowing down — because the instinct is both right and wrong at once, and sorting out the details is cheaper than slapping on another coat of paint and landing right back on the same feeling.

What "AI slop" actually means

First thing worth clearing up: the label doesn't stick to the fact that a neural network was involved — it sticks to the result. You can hand-code a site by hand and get the exact same effect, just by following the same defaults everyone else follows. The real issue is indistinguishability. A fintech landing page that would become indistinguishable from a CRM's landing page the moment you swapped the logo — that's the formula. Same hero layout, same faceless font, same shadow under every card, same button with exactly 8 pixels of border-radius.

The feeling isn't unique to websites. It's a conference room in a business hotel — could be Dubai, could be Dayton — same patterned carpet, same dim lighting, same vase of fake flowers at the front desk. The room does its job perfectly and leaves absolutely nothing in memory, because it was built for the average guest of the average conference. And the average, by its very nature, isn't about anyone in particular.

Same story with websites. AI slop does its job just fine — buttons click, forms submit. It's just the output of averaging across a thousand similar sites, and an average doesn't argue for anything and isn't remembered by anyone.

Funny enough, the name isn't an accident — tools trained on millions of existing interfaces are, quite literally, computing an average and handing it back as an answer. So it's not that the AI is bad at this. It's doing exactly what it was built to do — offering the most probable continuation, and the most probable thing is, almost by definition, never the distinctive one.

Why standing out isn't about taste — it's about memory

Here the instinct is right. There's a surprisingly concrete mechanism behind it, and it's been studied for a while.

There's an old psychology experiment: show someone a list of ten similar items, one of which stands out somehow, and later they'll remember that one item while the other nine blur into a vague smear. The effect is named after the researcher who found it, von Restorff, and it's been confirmed dozens of times since, in contexts ranging from shopping lists to ad campaigns. The brain doesn't spend resources on anything indistinguishable from the background. It economizes.

It plays out in more mundane ways too. A single conference could leave you with a stack of business cards — and without prompting, you'd recall maybe one, two at most, and it's rarely the tidiest one. It's the one with something off about it. The rest did their job as business cards and never surfaced in memory again.

Marketing has its own version of this, formalized into a whole field around what's called distinctive brand assets: not "what makes us better than the competition," but "what makes people recognize us in a crowd." It could be a shape, a tone, a detail — anything consistent enough that it belongs to you and no one else. And it isn't just theory: brands with strong distinctive assets get recalled at the moment of choice far more often than brands without them.

Applied to a personal site, here's what that means: look like one more business card in a stack, and you won't be remembered — not even by the people who liked you. Being liked and being remembered turn out to be two different jobs.

The trap: forced quirkiness backfires

This is where instinct starts to lead you wrong, and it's the most uncomfortable part of the whole story.

There's a classic experiment from social psychology: people listen to a recording of someone answering quiz questions. In one version, the person nails every question. In the other, they do almost as well — but spill coffee on themselves at the end. Listeners consistently rate the coffee-spiller as more likable — but only if he'd already shown himself competent. Play the same recording with someone who answered badly, and the spilled coffee doesn't save him. It just confirms he's no good.

Translated to websites: strangeness, irony, an authored voice — these only work as a layer on top of demonstrated skill. Not instead of it. And there's a sharper wrinkle to it — the moment character starts to feel deliberate, rehearsed, it stops working entirely. You know the type at a networking event who tells the same self-deprecating story about themselves for the third time in one evening, clearly polished in advance? It was funny the first time. By the third, it reads exactly for what it is: a bit. People are pretty good at telling a genuine rough edge from one performed on cue, and the performed version works against you, not for you.

It's hard to fake this even on purpose — you can commission "a bit of personality" from a designer, or have a model generate a "bold headline," but the result gets spotted just as easily as a rehearsed joke. The difference between real and manufactured quirk isn't how bold it is. It's whether it grows out of some specific fact about a specific person, or exists on its own, purely for effect.

Which means bolting "character" onto a site as a separate layer over nothing underneath isn't a fix. It's the same slop, just handwritten.

And distinctiveness has a ceiling

Third piece of the puzzle, and the one that almost never gets said out loud: being too different is a problem too.

The brain prefers whatever's easiest to process. This has been tested more than once — people consistently like the typical better than the atypical, simply because typical things get read without effort, and that ease of reading is itself pleasant. When researchers statistically stripped out the effect of that ease from the well-known "beauty makes it feel usable" effect, the correlation nearly collapsed. Meaning most of "if I like it, it must be easy to use" comes down to plain cognitive ease, not beauty itself.

Which is where the old industrial-design rule comes from — the name most associated with it is Raymond Loewy, the man behind the shape of the Coca-Cola bottle: people are torn between curiosity about the new and fear of anything too new, and what sells best is bold but instantly legible. Novelty, in his own phrase, has a shock zone — a point past which the appetite for strangeness flips into resistance.

On a website, that looks like this: so much decoration that the "get in touch" button drowns in a pattern, and a visitor has to stop and figure out where to even look — that's not standing out, that's just added friction. Being different needs to be legible. Not chaotic.

What this looks like from the other side

Worth standing on the other side of the screen for a second — where the person sits who's scrolling through other people's sites, deciding whether to trust one. Doesn't matter if they're hiring a developer, choosing a contractor for a renovation, or comparing two wedding photographers — the mechanics are the same.

There was an open survey once, run across several hundred hiring managers and recruiters, asking directly: what makes a person's site memorable. The answers varied in detail but agreed on one thing — there are too many templates out there, they all look "the same," and whatever looks purpose-built for the problem beats whatever looks copied from somewhere else. And the thing that actually decides whether someone keeps reading or closes the tab isn't the list of services or the years of experience — it's whether that very first screen feels human or feels generic. Generic, and they close it before they even find out who you are.

Same von Restorff story, just with money riding on it now. Out of ten similar candidates, contractors, specialists, the one who gets remembered is the one who snagged onto someone's memory somehow. One condition applies: real work has to sit behind that snag — not a designer's decision to round the headline off a little more for effect. The trap and the ceiling from earlier sections apply here just as bluntly: forced originality reads as unprofessional in this context, and an overloaded one reads as disrespect for someone else's time. But an empty, generic template — one with nothing to remember it by — loses to both, just more quietly.

So does it work or not

Put it all together and the answer doesn't fit into one word, which is, honestly, the most boring news in this whole piece.

Distinctiveness works — but it's not about bolting on strangeness. It's about surfacing what's already there and just hasn't been shown yet. A phrasing you actually worked out yourself, not one a model handed you in five seconds. A detail in a case study you could only reach by doing the work. A decision that would be strange to copy, because it comes out of one specific problem rather than a component library. That's a distinctive asset in its purest form — not a designer's call to round a headline a little more, but the fact that a specific person with specific experience is standing behind the site, and it shows.

Forced difference — the kind bolted on as a final touch over a stock template — works against its author the same way spilled coffee works against someone who couldn't answer a single quiz question.

And too much difference works against the visitor instead — it eats up attention that should have gone toward whatever they actually came for.

So next time someone drops the phrase "AI slop," it's worth checking not how many quirks are on the page, but whether there's anything behind that page worth showing in the first place. Design can do a lot. Giving you a personality you don't have isn't one of them.