Use AI for the Work. Keep People on the Why.
I'm an advocate for AI on every lean marketing team. I'm also convinced that the more of it you use, the more the human parts matter. Here's where I draw the line, and why.
There are two loud camps right now.
One says AI will replace your marketing team. The other says AI content is slop and you shouldn't touch it.
I think they're both wrong.
For a lean team, AI may be the most valuable addition you make this year. It's also the worst strategist you'll ever put in charge of anything.
I use AI every day in my own work. I'm also the person who'll tell you when to put it down.
Both of those come from the same place.
Why lean teams can't sit this one out
Most of the teams I work with are one or two people doing the work of ten. Sometimes it's a founder doing marketing between everything else. Their constraint isn't ideas or effort.
It's hours.
That's exactly where AI earns its keep. In an MIT study, 453 professionals, marketers among them, were given ordinary business writing tasks. The ones with access to ChatGPT finished 40% faster, and their work was rated 18% higher in quality. The people who'd been struggling most gained the most.1
Small businesses have noticed. In Constant Contact's early-2026 survey of more than 1,500 small business owners, 54% were already using AI marketing tools, and another 27% planned to start this year. Forty-four percent use it to write content.2
So the question for a lean team isn't whether to use AI anymore.
It's where.
The line is real, and it's jagged
The best research I've seen on this comes from Harvard Business School and Boston Consulting Group. They ran an experiment with 758 consultants doing realistic work, some with AI and some without.3
On tasks AI handles well, the consultants finished 12.2% more of them, worked 25.1% faster, and produced work rated more than 40% higher in quality.
On a task just outside what AI does well, the consultants using it were 19 percentage points less likely to get the right answer than the ones working without it.
Same people. Same tool. Opposite results.
The researchers called it a "jagged technological frontier." On one side of the line AI makes good people better. On the other side it makes them confidently wrong.
The tool didn't get worse. People trusted it with the wrong job.
So the real skill isn't using AI.
It's knowing which side of the line a task is on.
How I sort the work
Every marketing task lands in one of three places.
Hand it off. AI does the heavy lifting, and a person checks it before it goes anywhere. Research and summarizing. First drafts of routine copy, like product descriptions, event recaps and FAQ answers. Turning one good piece into ten formats. Cleaning up data, building reports, organizing spend. Brainstorming volume ("give me thirty subject lines"). Resizing, formatting, scheduling. This is where a lean team gets its hours back.
Work it together. A person leads and AI assists. Emails, posts and articles in your voice, where AI can draft or tighten but you rewrite until it sounds like you. Campaign concepts, where AI widens the options and a person chooses. Analysis, where AI finds the patterns and a person decides what they mean. (A number means nothing until a human puts it next to the right other number. I wrote about that here.)
Keep it human. These are the things that don't work without human eyes and brains. For me, there are four.
The four things I keep human
Strategy and purpose. Why a company exists, what it stands for, and what it won't do are never AI's call. AI can describe a mission. It can't hold one.
There's a deeper reason, too. AI works by predicting what's most likely, so it pulls toward the middle. A study in Science Advances found that AI-generated ideas made individual stories better, but the stories written with AI help also came out more similar to one another.4 For fiction, that's a curiosity. For a brand, it's the whole problem. Sameness is the one thing a values-led company can't afford, and it's why I've argued AI averages your positioning rather than inventing it.
My voice, and the final word. Nothing goes out under my name that I haven't rewritten or approved. Not because AI can't write a decent sentence. It can. But audiences are paying closer attention than most marketers think. In a 2025 study of 3,000 U.S. adults, content that readers suspected was AI-generated was rated 48% less trustworthy, 57% less authentic, and 60% lower in emotional connection, whether or not AI actually wrote it.5 NielsenIQ found people described AI-generated ads as "annoying," "boring," and "confusing," and their brains showed weaker memory activation even for the high-quality ones.6
You don't just lose a little polish when your voice goes generic.
You lose the reason anyone was listening.
Relationships. Replies, comments, DMs, client conversations, the customer who's upset. When someone reaches out to you, they're asking for a person. AI can help me get organized before a conversation. It doesn't get to have the conversation.
Taste and judgment. What fits the brand. What doesn't. What not to say at all. When to stay quiet. That last one may be the most valuable marketing skill there is, and no tool will ever volunteer it.
What I'm actually teaching a lean team
This month I'm running a training with a two-person content team at a family-owned company I've worked with for years.
They're good. One has been with the company a long time and knows its customers cold. The other is a recent marketing grad with plenty of theory and drive. Between templates, design tools and AI, producing content isn't the hard part for them anymore.
So speed isn't what I'm teaching.
Judgment is.
Not long ago one of them came to me wanting to recreate a competitor's video because it was getting a lot of engagement. That's a good instinct. He noticed what was working.
It's just the wrong question.
The question isn't "will this get engagement?" It's "does this sound like us, and does it say what we stand for?" AI can help make that video in an afternoon. It can't tell you whether you should.
That's why the training produces a short brand filter, a few questions every piece runs through before it ships. The tools make the team fast. The filter makes sure they're fast in the right direction.
Tools → speed. Filter → direction. People → meaning.
Use it. Don't overuse it.
Here's the rule I give every lean team I work with:
Every hour AI gives back, spend part of it on something only a person can do.
Call a customer. Answer reviews yourself. Sit in on a sales call. Show up at the event. Write the essay, the real one, in your own words.
Because that's the trap. AI makes it easy to produce more of everything. And a team that uses its new hours to produce more ends up with more of what everyone else has, in a market that's already buried in it.
The teams that win will use AI to clear the busywork off the table, then spend what they saved on the part of marketing that was always the point.
Good marketing is about human connection.
AI can do a lot of the work that gets you there. It can't be the connection.
So use it for the work. Keep people on the why.
And if you're not sure which side of the line your team is on, start with one question: who wrote the last thing you published, and did it sound like you?
Sources
- Shakked Noy and Whitney Zhang, "Experimental evidence on the productivity effects of generative artificial intelligence," Science 381 (2023); summary via MIT News, July 14, 2023.
- Constant Contact, Small Business Now report, Q1 2026, as reported by Roger Dooley in Forbes, February 11, 2026.
- Fabrizio Dell'Acqua, Edward McFowland III, Ethan Mollick and colleagues, "Navigating the Jagged Technological Frontier," Harvard Business School Working Paper 24-013, September 2023.
- Anil R. Doshi and Oliver P. Hauser, "Generative AI enhances individual creativity but reduces the collective diversity of novel content," Science Advances 10 (2024); summary via ScienceDaily, July 12, 2024.
- Raptive, "The 'AI stink' is real, and it's costing brands," August 22, 2025 (3,000 U.S. adults).
- NielsenIQ, "NIQ research uncovers hidden consumer attitudes toward AI-generated ads," December 12, 2024.
Decided what your business is for, and can't get it said clearly yet?
Thirty minutes, no pitch. Tell me what you're working on and I'll tell you honestly whether the problem is marketing at all — and whether I'm the right person for it.