A client called me a year ago, panicked, because a competitor had just launched a campaign that looked almost exactly like one of ours. Same colors, similar tagline, even a similar joke in the headline. She wanted to know what we were going to do about it, possibly involving lawyers, possibly involving a strongly worded letter.
Nothing, I told her. Because they can copy the campaign. They can’t copy her. And frankly, if they’re spending their afternoon reverse engineering our headlines instead of running their own business, that’s a them problem.
That stuck with me, through every new platform, every new tool, and now, every new AI model that promises to write your marketing for you. Tools change. Tactics change. The actual thing that makes a business worth choosing has never once changed, and it’s not something you can prompt your way into, no matter how politely you ask.
Your voice isn’t a prompt. It’s the one thing AI can’t write for you.
Here’s what I mean by voice, since it gets thrown around loosely, usually right before someone hands it a meaningless meaning. It’s not your font choice or your color palette, though those matter too. It’s the specific way you’d explain your business to a friend at a barbecue, three beers in, no filter. The thing you’d say without thinking about it, that a competitor genuinely wouldn’t say the same way, even with the exact same information sitting in front of them.
AI can write in a voice. It cannot invent one for you from nothing. If you ask it to “sound professional but friendly,” you’ll get professional but friendly, and so will the next twelve businesses who typed the exact same instruction, probably at the exact same time zone, probably all thinking they’d just discovered something clever. That’s not a flaw in the technology, it’s just what happens when you skip the step where you actually decide who you are before you ask something else to say it for you.
Here’s the mechanical reason this happens, because it helps to understand why. These tools work by predicting the most likely next word based on everything they’ve ever read. “Most likely” is the whole game. When you give it nothing specific to work with, it reaches for the statistical center of gravity, the most average, most expected, most already-written version of whatever you asked for. That’s not the tool being lazy. That’s the tool doing exactly what it was built to do, which is find the middle. Your job is to drag it off the middle, and it can’t do that part for you, because the thing that makes you not-average is the one input it doesn’t have unless you hand it over.
This is the mistake I see most often with small business owners who are excited about AI, and I get the excitement, I use these tools constantly myself. They open a chat window, type “write me an Instagram caption for my bakery,” and wonder why it reads like every other bakery’s Instagram caption, right down to the suspiciously enthusiastic emoji. The tool didn’t fail. It did exactly what was asked. Nobody told it what makes this bakery different from the one two blocks over, the one that also, coincidentally, just started using AI too.
Voice comes from specifics. Not “we care about quality,” which is the marketing equivalent of a shrug, but the actual story of why you started, the exact phrase your regulars use to describe you, the thing you refuse to do even though it would probably make you more money. Once you know that, AI becomes genuinely useful, because now it has something real to work with instead of a blank slate it has to fill in with whatever’s statistically average.
And here’s the part that surprises people: the more specific and particular your input, the less your output sounds like AI at all. The generic smell everyone’s learning to detect comes from generic input. Feed it something only your business could have said, and what comes back carries that fingerprint. The tool didn’t get more creative. You did, and it just carried your creativity further and faster than you could alone.
So before your next AI-assisted anything, spend ten minutes writing down three things: how you’d describe your business to a friend, the phrase your best customers actually use about you, and one thing you do that your competitors don’t. Feed that in every single time. That’s the difference between AI writing for your brand and AI writing for anyone’s brand, and it’s honestly less work than it sounds like.
Takeaway: AI can write in a voice, but it can’t invent yours for you, that part is still entirely your job.
Try this: Write down your three specifics (the friend explanation, the regular’s phrase, the thing you won’t do) somewhere you’ll actually see them, and paste them into your AI tool as context before your next piece of content, every time, not just once. Then do one small experiment: run the same prompt twice, once with those specifics and once without, and read the two side by side. The gap between them is your voice, made visible.
More on this in the next issue, since knowing your voice and protecting it from getting quietly automated away are two different skills. Talk soon, Tiffany