The short version: AI hasn't made small business marketing worse, it's made the laziness that was already there faster and more visible. The fix isn't a better prompt or a different tool, it's fifteen minutes of unglamorous work most business owners skip because it feels like homework rather than marketing.
Worth reading next: AI Meeting Notes Are Quietly Making Your Small Business Worse at Liste.
The email that made me look at my own content
Last spring a client, a bookkeeper in Kent who'd been running her own practice for eleven years, sent me back a client welcome email I'd drafted with her using ChatGPT. Her note said: "This reads like every accountant in Tunbridge Wells wrote it together."
She was right, and it stung a bit because I'd written it. Not AI on its own, me, with AI, in about four minutes. It said things like "we're passionate about helping your business thrive" and "in today's fast-paced world of financial management." None of that was a lie exactly. It just wasn't her. She once told me she'd rather eat glass than use the word "passionate" in a sentence about tax returns.
That was the moment I stopped treating AI copy as a shortcut and started treating it as a first draft that needs a second pass built entirely from things the actual human would say.
I pulled 40 client websites to check how bad this really is
Because I didn't want to write a blog post based on one awkward email, I went back through content audits I'd done for clients over the past year, roughly 40 small business websites and email sequences across bookkeeping, coaching, trades, and local retail. I counted specific phrases.
- 34 out of 40 used at least one of: "in today's world," "take your business to the next level," or "we're passionate about."
- 28 used some version of a three-item list that started with an adjective plus a noun, like "trusted, reliable, results-driven."
- Almost none, four out of 40, mentioned a specific client, a specific number, or a specific mistake the business had learned from.
None of these businesses had copied each other. Most had never met. But most had run a version of "write me a bio for my landscaping business" through the same tool with roughly the same instructions, and the tool did exactly what it was trained to do: produce the statistically most likely, most inoffensive version of that sentence. That's not a flaw in the AI. That's the AI working correctly on an under-specified prompt.
The uncomfortable bit nobody wants to hear
Here's the part that annoys people when I say it in workshops: the sameness problem existed long before AI. Go back and read ten small business "About Us" pages from 2015. Same rot, different decade, it just took longer to write because a human had to type out the cliches by hand. AI didn't create bland marketing. It just made bland marketing available at scale, for free, in nine seconds, to everyone who used to be too busy or too broke to hire a copywriter.
The businesses that stand out now aren't the ones avoiding AI. They're the ones who worked out that AI is only as specific as what you feed it, and most people feed it almost nothing. If your prompt is "write a LinkedIn post about why customer service matters," you'll get the same 200 words that 4,000 other small businesses got that morning. If your prompt includes an actual customer's actual complaint, the actual thing you said back to them, and the actual result three weeks later, you can't get a generic output even if you tried.
What distinct marketing looks like
I send people to look at how bigger brands solved this, not to copy the size of their budget but the shape of their thinking. The Mailchimp marketing strategy worked for years because their copy sounded like one slightly weird, slightly funny person, never like a template. Away's marketing strategy leaned on real traveller stories instead of "premium quality luggage" language. Airbnb's marketing strategy is built almost entirely on specific hosts and specific guests, not "experience the world" fluff. And if you want the blunt end of this, go read the business lessons from Alex Hormozi, the man built an audience by refusing to write a single sentence that sounds like it came from a template, on purpose, as a strategy.
None of those brands got there by finding a cleverer AI prompt. They got there by deciding what they'd never sound like, then checking every piece of output against that list.
The five-minute fix (do this before you write anything else)
This is the bit I make clients do now, before we touch a single AI tool together.
- Write your banned words list. Ten to fifteen words or phrases you never want to see in your own marketing again. Mine includes "passionate," "," "unlock," and "in today's world." Yours will be different. Put it somewhere you'll check it.
- Record one real conversation. Next time you're on a call with a customer, or chatting to a colleague about work, record it (with permission) or jot down three sentences you said naturally. That's your voice. Not the voice you use when you're "writing marketing," the one you use when you're not thinking about writing marketing at all.
- Build a five-line brief before every prompt. Not "write a blog intro about pricing." Instead: who exactly is reading this, what specific thing happened last week that prompted it, what's the one sentence you'd say to this person's face, what number or example proves your point, and what should this never sound like. Five lines, takes about two minutes.
- Feed the AI your own past work, not the topic. Paste in three emails you've sent to real customers and ask the tool to match that tone, rather than asking it to invent a tone from scratch. This one change alone fixes most of the sameness problem.
- Read it out loud before you post it. If you wouldn't say it to a customer standing in front of you, don't publish it. This step takes ninety seconds and catches nearly everything.
That whole process takes about fifteen minutes the first time and gets faster every time you do it, because your banned words list and your voice notes don't need rebuilding each time, only your five-line brief changes.
Where this pays off, not just reads nicer
The bookkeeper rewrote her welcome sequence using her own actual phrases, including one where she tells new clients "I will absolutely tell you if you're about to do something daft with your VAT." Open rate on that sequence went from around 38% to 61% over the following quarter, and she picked up two referrals directly off the back of clients forwarding that specific email to friends, because it was quotable. Bland copy never gets forwarded. Specific copy does.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
This is also where interactive tools help, not as decoration but as proof you understand a specific problem rather than a generic one. A trades business I worked with added a simple job-cost calculator to their site, similar in spirit to the approach covered in how interactive calculators convert more visitors, and it outperformed every generic "get a quote" form on the site because it asked for the customer's actual numbers instead of talking at them in general terms. Same principle as the banned words list: specificity beats polish every time.
What to check before you publish anything AI helped write
A quick habit worth building, and it takes less time than proofreading for typos:
- Does this contain one number, one name, or one specific detail that couldn't apply to any other business in your postcode?
- Would a competitor's AI tool have produced this exact sentence if fed a similar topic? If yes, rewrite it.
- Does it include anything you'd be slightly embarrassed to say out loud to the customer's face? If it's too smooth, it's probably too generic.
- Is there a sentence a real client said, verbatim, anywhere in this piece? If not, add one.
None of this is about being anti-AI. I use it every day, for research, first drafts, restructuring, headline options, all of it. The point isn't the tool, it's what you put into it. Businesses I've worked with through an AI consultant for small business engagement almost always assume the fix is a better tool or a smarter subscription. It's rarely that. It's usually that nobody's ever sat down and written out, in their own words, what they'd never want their marketing to sound like, and then checked their output against it.
If you want to see how a brand builds this kind of distinctiveness at scale on a search-heavy budget, the Ahrefs marketing strategy is worth a proper look, most of their content works because it's stuffed with real screenshots, real numbers, and real "here's what happened when we tried this" detail rather than advice that could have come from anyone.
Frequently asked questions
Does AI make marketing copy worse for small businesses?
Not on its own. AI produces the most statistically likely output for whatever prompt it's given, which sounds generic if the prompt is generic. Feed it specific numbers, real customer quotes, and your own past writing, and the sameness problem mostly disappears.
How can I tell if my AI-written content sounds like everyone else's?
Check for stock phrases like "in today's fast-paced world," "passionate about," or "take it to the next level," and count how many sentences include a specific name, number, or real event. If you have none, it will read as generic even if the grammar is perfect.
Should small businesses avoid using AI for marketing entirely?
No, avoiding AI just means writing slower with the same generic instincts. The fix is a better brief, not less AI, feed it your real customer conversations and a list of words you never want to use, rather than asking it to invent a tone from nothing.
How much does it cost to get proper help fixing this rather than doing it alone?
Costs vary widely by scope, but a focused brand voice and AI workflow review for a small business typically runs from a few hundred to a couple of thousand pounds depending on how much content needs rebuilding. It's worth comparing that against how much a single forwarded, quotable email or referral is worth to your business before deciding it's not worth the spend.
Related reading: AI Tool Costs Explained for Business Owners in 2026 and Instagram Marketing Strategy: How They Built a Brand That Wins.
This builds on my main AI marketing guide, my main guide on the topic.
