Read any five LinkedIn posts today and tell us honestly whether you can remember which one was written by which company.
You can’t. Neither can we. And that’s the problem.
We’ve hit a quiet crisis in marketing that nobody’s naming clearly enough. The democratization of AI-assisted writing has created a massive wave of competent, professional, on-brief content that is also completely indistinguishable from everyone else’s competent, professional, on-brief content.
Everyone sounds the same. Everyone sounds fine. Nobody sounds like anyone in particular.
If you’re a growing business trying to stand out, this isn’t just an aesthetic issue. It’s a business problem.
What Brand Voice Actually Is
Let’s define the term carefully, because “brand voice” gets used to mean about fourteen different things.
Brand voice is the specific, recognizable, repeatable way your company sounds in writing and speech. It’s the combination of what you say, how you say it, what you refuse to say, and what you’re willing to be honest about when everyone else is hedging.
It’s the reason a Patagonia email doesn’t sound like a Marriott email doesn’t sound like a Liquid Death can. Each one is unmistakable. Each one could be stripped of its logo and you’d still know who was talking.
That’s brand voice. It is an enormous asset. It’s also the single thing generic AI output destroys faster than anything else.
How AI Averages You Into Invisibility
Here’s the problem with asking AI to “write in our brand voice.” Even with guidelines, even with examples, even with tone documents, a language model is trained to produce the average of what professional marketing content looks like.
Average is the enemy of distinctive. Average is what every competitor is also producing. Average is why your email open rates are going down even as your sending volume goes up.
When you ask AI to write in your voice, what it actually does is produce “professional marketing copy, lightly tilted in the direction of the examples you provided.” It’s a regression to the mean of professional copy with a small amount of flavor on top. Sometimes the flavor is enough to fool a casual reader. It’s never enough to fool someone paying attention.
And in a world where your customers see hundreds of marketing messages a week, “not fooling someone paying attention” is the whole ballgame.
What Distinctive Voice Requires
Real brand voice has properties that are unnatural for a language model to produce.
It takes positions. Distinctive voices say things other people won’t say. They have a point of view. They’re willing to be wrong in public. AI, by its training, defaults to neutrality and hedging.
It has specific references. Distinctive voices mention the weird, the local, the insider, the particular. They reference the name of the small-town diner, the specific year something happened, the obscure quote, the in-joke. AI defaults to the generic.
It has rhythm. Distinctive writing has a cadence. Short sentences. Then long ones, doing the work of carrying you from one idea to the next. A mix of registers, a deliberate repetition, the occasional one-word paragraph for emphasis. AI produces competent, regular prose that reads fluently and forgets fast.
It breaks its own rules. Every great brand voice does something technically wrong on purpose — starts sentences with “And,” uses sentence fragments, opens with a question, writes “Y’all” in a corporate email. AI, absent very specific instructions, defaults to the technically correct, the grammatically safe, the professionally bland.
It is consistent across years, not months. The biggest thing AI can’t do is hold a voice across a multi-year span. It can mimic an example for a post. It can’t develop, protect, and evolve a voice over a decade of content, events, hires, and pivots. That’s a craft. It requires humans who know the brand, the history, and each other.
The Agency’s Role in Voice
A good agency’s job is to help you have a voice that is distinctive — not a voice that is “professional.” Those aren’t the same thing, and a lot of in-house teams haven’t caught up to the distinction.
Here’s what that actually looks like in practice.
We spend time with founders to find the way they talk when they aren’t performing — the phrases they use with a trusted friend when they’ve had two beers, the beliefs they hold that they haven’t been saying in public because they thought they couldn’t, the contrarian takes they mutter in internal meetings and then soften for the outside world.
That’s where brand voice lives. Not in a tone document. In the honest, unpolished way your best people actually think.
Our job is to find it, protect it, sharpen it, and keep it consistent across everything you put into the world — even the stuff we generate with AI.
The Test
If you want to know whether your brand has a distinctive voice right now, here’s the test.
Take your last ten pieces of marketing content — emails, posts, web pages, press releases. Strip the logos. Strip the company name. Hand them to someone who doesn’t know your business.
Can they tell it’s all from the same company?
Can they describe that company’s personality in three specific words?
Would they be able to pick your next piece of content out of a lineup that included content from three of your competitors?
If any of those answers is “probably not,” your voice isn’t distinctive yet. And if you’re relying on AI to produce your content without senior human editing, it’s about to get less distinctive, not more.
Where This Is Going
Our prediction: over the next three years, the brands that win will be the ones that double down on sounding like themselves — aggressively, specifically, even at the cost of being polarizing.
The brands that lose will be the ones that let AI quietly average their voice into mush, one “just run it through the tool” at a time, until their content is technically fine and practically forgettable.
This is one of the places a seasoned agency partner is genuinely hard to replace. Voice is a craft. It takes human ears, human judgment, and — most of all — a willingness to resist the pull toward average that AI tools make so tempting.
The Bottom Line
AI averages the internet. Distinctive brands do the opposite.
If your voice is one of your genuine competitive advantages — and for most growing businesses, it is — then letting AI write your marketing without a senior human in the loop is actively destroying equity you’ve spent years building.
That’s not a tooling problem. That’s a strategic one. And it deserves a strategic answer.
Up next in the series: the hidden cost of a bad AI-generated post. In the meantime, if you want a partner who will protect your voice instead of flattening it, let’s talk.
Frequently Asked Questions
Why can't AI reliably write in my brand voice?
A language model is trained to produce the average of professional marketing content. Even with guidelines, it regresses toward that mean — competent, professional, and indistinguishable from everyone else. Average is the enemy of distinctive.
What makes a brand voice actually distinctive?
It takes positions, uses specific references, has rhythm, breaks its own rules on purpose, and stays consistent over years. Those traits are unnatural for a model that defaults to neutral, generic, grammatically safe prose.
How does an agency help build a distinctive voice?
By finding the honest, unpolished way your best people actually talk — the beliefs and contrarian takes they usually soften — then protecting and sharpening that voice across everything you publish, including AI-assisted work.
How do I test whether my voice is distinctive?
Strip the logos from your last ten pieces and hand them to a stranger. If they can't tell it's one company or describe its personality in three specific words, your voice isn't distinctive yet.
Last updated August 27, 2026