The brief arrives on a Monday, as it often does.

Someone spent the weekend scrolling—an awards show, a competitor's campaign, something that racked up numbers. By the time they walk into the conference room, they haven't decided what to do. They've decided what to look like. This distinction matters more than anyone in that meeting will admit.

"We need a Liquid Death moment." "Something that feels like Spotify Wrapped." "That Duolingo energy—can we do that?"

Within hours, a studio somewhere is reverse-engineering the aesthetics of a canned water company to sell indemnity insurance, supply-chain logistics, or whatever category has no business wearing a punk-rock sneer but will wear one anyway. The tuxedo never fits. It wasn't made for those shoulders. But the brief is signed, the mood board is full of screenshots from someone else's best year, and the room nods.

Brands have always copied each other. That's not new. But there's a difference between copying and stealing, and most marketing departments have quietly forgotten which one actually works.

This isn't a new distinction. Painters, writers, and musicians have argued about it for the better part of a century—the difference between the artist who takes a surface and the one who takes an idea and transforms it until it's unrecognizable. The good ones always steal. The bad ones copy and call it homage. What's new is that marketing has started pretending the distinction doesn't exist. That every lift is research. That every reference deck is a strategy. It isn't, and it never was.

Stealing is slower because it requires you to go past the thing you admired and find the reason it worked—the human observation underneath that the brand was brave enough to build on. Then you set that observation aside, walk back to your own category, and see whether something like it lives there. Most of the time it doesn't, and you let it go. When it does, the work that comes out the other side looks nothing like what you started from. That's how you know it's yours.

Duolingo spent fourteen years earning the right to kill its owl. The TikTok account that grew from fifty thousand to sixteen million followers, the Dua Lipa bit running since 2021, the weird passive-aggressive voice the brand had carefully built into something people genuinely looked forward to — all of it came from the brand wrestling with a single unglamorous question for over a decade: how do you make a teenager care about conjugating Spanish verbs? When the owl "died" in February 2025, the grief was theatrical, but the attachment underneath wasn't. People had spent years getting bullied by that bird into doing their lessons, and a character you've had a relationship with for that long can actually be mourned, even as a joke.

What happened next was predictable and instructive. Vita Coco, Dr. Squatch, Mr. Clean, Foot Locker — all posting condolences the same week, as if the industry had entered collective mourning for an owl that most of their audiences had never even opened the app to meet.

You can call it social media. You can call it being reactive. And maybe that's all it is — brands seeing something work and wanting in on the conversation. That instinct isn't new. It's human. It's the same reason people quote movies they love or reference songs everyone knows. Participation feels good.

But something shifted.

The algorithm didn't invent the copycat instinct. It weaponized it. Once a format gets engagement, it becomes "proven." And proven formats get prioritized, amplified, recommended. Which means the thing that worked once now carries less risk than the thing nobody's tried. So brands aren't just choosing to copy anymore — they're being rational inside a system that makes not copying feel like the dangerous bet.

It's the same logic that keeps Hollywood remaking Twister. Not because anyone believes the world needs another tornado movie, but because sequels and reboots have precedent. They have comps. They have data. An original idea has none of that, which makes it harder to defend in the room where budget gets approved. So studios keep making worse versions of the same film, and everyone involved knows it's worse, but the structure makes it feel mandatory.

That's what's happening here. None of these brands had a mascot worth burying. They had logos, and logos don't die — they just get refreshed in the next brand guidelines PDF. The funeral didn't spread because it meant something. It spread because it was easy to step into, and because stepping in felt safer than staying out.

The shift isn't that brands are participating. It's that they're starting from participation instead of from an idea. And the algorithm has made that shift feel like the only move left.

Spotify Wrapped is the same story at industrial scale. By 2025, Spotify's year-end recap had become the most copied and most failed execution in the business. Amazon Music did one. Apple Music did one. YouTube, Gymshark, Domino's, easyJet, Ryanair, Specsavers — all filed their own versions into the feed, each slightly smoother and less distinguishable than the last. Eventually the Empire State Building posted its top artists of the year, and nobody thought to ask why a skyscraper would have top artists, because by then the format had fully detached from any underlying reason to exist.

The reason Wrapped works, and the reason almost none of its imitators do, is that Spotify spent the better part of a decade noticing something its competitors hadn't bothered to name—that music taste reads as identity rather than information, and that the data they were sitting on wasn't consumption data at all but a high-signal personality marker people were desperate for an excuse to broadcast. Sharing your top five artists tells the world you're cultured, sophisticated, or obsessively deep into some niche scene nobody else in your office has heard of. It's a digital badge, and people wear it gladly, which is why Wrapped doesn't feel like a brand asking you to post for them. It feels like the brand giving you a reason to post for yourself.

Strava understood this in 2017 and did the rarer thing, which was to steal properly. They didn't lift the slides or the color palette or the share mechanic. They went past the surface of Wrapped entirely and found the same underlying instinct living in their own category—runners and cyclists, it turns out, care about their training data with even more intensity than music listeners care about their play counts, because the miles are earned, the splits are receipts of effort, and the whole thing is a form of self-presentation wearing the clothes of fitness tracking. Strava's year-end recap celebrated grit and discipline the way Wrapped celebrated taste, and it worked for exactly the same reason, even though it looked nothing like the thing it had learned from.

Graza asked a different question. It wasn't "how do we become the Liquid Death of olive oil?" It was smaller and far more useful: why does the home cook hesitate before using the nice bottle? Every other brand in the category obsessed over provenance, terroir, single-estate cold-press whatever—all answers to questions about prestige. Graza looked at the actual moment of cooking and noticed something tiny: a pulse of guilt. Too expensive to cook with. Too expensive to finish with. So the bottle sat there, rationed. A squeeze bottle answered a question nobody had bothered asking out loud.

So what did their competitors do instead of taking the note? They saw bright colors, playful caricatures, fun energy—and they copied it. Six olive oil brands launched versions of the same thing. Similar colors. Similar packaging. Different images on the label, but the same move. They became dupes of Graza.

If you see something working, it's not your invitation to do the exact same thing. It's your invitation to ask why it worked. What question was it answering? Does your brand have that same question, or just like the answer? Because you can see the principle—fun branding, personality, energy—but that doesn't mean your execution has to be identical. You actually want to do the opposite. You want to find your own hesitation and solve it differently.

Anthropic asked a different question in a much shorter window. When OpenAI announced ads inside ChatGPT on January 16th, 2025, Anthropic treated it as an opening. Their question wasn't "how do we launch our brand?" It was "what can we say out loud right now that our competitor can't?" Nineteen days later, four Super Bowl spots ran. Each one showed an AI conversation hijacked mid-sentence by an ad. Each ended the same way: ads are coming to AI, but not to Claude. The statement worked because it was already true. Claude jumped from forty-first to seventh in the US App Store that week.

AI didn't make the brief lazy. It just stopped hiding it.

A strategist can prompt their way to a Duolingo-style content strategy in four minutes. The output has rationale, tone of voice, a content calendar—all of it technically correct. What it's actually done is pattern-match against every successful brand ever fed into the model and return the average. The problem? Your best insights never live in the average. They live in the outliers—the anecdotal eccentricity, the strange specific thing one customer said that turned out to be how half of them felt. AI can't see outliers. By design, it smooths them out. And smoothing out outliers is exactly how you lose the question.

Most CMOs will tell you they know algorithm-driven optimization is erasing what makes their brands distinct. Most of those same CMOs will also tell you they can't afford to stop. Everyone sees the trap. Nobody wants to be the first one out. So they stay, watching the thing most likely to kill their brand's distinctiveness become the thing they believe they can't live without.

AI doesn't fix the baseline—it industrializes it. The lazy brief gets executed overnight instead of over a quarter. The borrowed format ships to a hundred brands by morning, each one slightly smoother, slightly more average, slightly more forgettable than the last. They arrive simultaneously, drown each other out, and disappear into feed noise. The irony: by accelerating the execution of lazy briefs, AI accelerates their failure too.

Before the next brief, three questions worth asking:

Most briefs skip all three and start at the answer. And when you start at the answer, you're already copying. You're already one of a hundred brands shipping the same thing overnight, each one slightly smoother and slightly more forgettable than the last. You're already part of the split that's quietly underway.

The next ten years of marketing will be defined by a split that's already quietly underway. On one side: a generation of brands using AI to execute lazy briefs faster, cheaper, and with more polish than ever before—a hundred Wrapped clones shipping overnight, a thousand owl funerals for logos nobody will miss, an infinite feed of work that's technically proficient and spiritually identical. On the other side: a much smaller group of brands doing something that will increasingly feel almost transgressive. Sitting with a question for longer than a quarter. Noticing what nobody else has bothered to name. Refusing to ship the mood board.

The tools aren't the problem. AI is a spectacular instrument for executing ideas. It's a disaster as a substitute for having them. The marketers who figure out the difference—who use AI to move faster through the work they already understand, and refuse to use it for the work of noticing—are going to make almost everything else look like wallpaper.

The question is which side of that you want to be on. There isn't a third option.

The brief isn't the villain. It's the mirror.