In a world flooded with AI-generated content, being unmistakably human is your greatest advantage. The hard part is actually pulling that off without ignoring AI or leaning on it too hard. That’s what this series is about: exploring real ways to use AI to sharpen your voice instead of flattening it. We’ll walk through strategy, video and podcasts, blog posts and case studies, and social and email — one post at a time.
You won’t find this in my bio, but foundation inspection is a key component of my job. Nobody hired me for it. Nobody trained me for it either. I just kept noticing the same thing over and over — foundationless ideas, everywhere — and at some point, I couldn’t keep pretending I didn’t see it.
Here’s how it usually goes. Someone invites me into a meeting to show me their big idea. A campaign, a content plan, a creative swing they’re excited about. And there it is, floating in the room with us. Not sketched on paper, not sitting on a table — actually hovering, fully formed, mid-air. It’s gorgeous, too. They walk me through it like a proud architect giving a tour, pointing out details I never would’ve noticed myself.
I let them finish. Then I ask a few questions.
“Who’s this for?”
“Is anyone else already doing this?”
“How does your audience actually use this?”
That’s usually when the tour guide goes quiet. No amount of clever design can save something with a structural problem. And we both just stand there and watch this beautiful thing come apart, piece by piece, and drop to the floor. It was never built on anything. It was just floating — and floating’s fine, right up until someone asks it not to.
I hate this part. Genuinely. I’ve watched good, well-meaning ideas fall apart over one basic question, and I know exactly what that feels like from the other side of the table. But the idea was never the problem. Nobody had poured anything underneath it strong enough to hold the weight.
This is exactly what’s happening with AI right now. It’s incredible at helping you build fast — a whole campaign, a month of content, done before lunch. But fast isn’t the same as grounded. AI will help you construct something beautiful in midair and never once mention there’s nothing holding it up. That part is still on you.
Understanding Your Audience with an AI-Assisted Analysis
Let’s start with the question that crumbles most well-intentioned floating ideas: who is this actually for?
Many marketers think they’ve already answered that. “B2B decision-makers.” “The VP of Marketing at a mid-size SaaS company.” Sounds great on paper, but knowing a job title isn’t the same as understanding a person. A real audience has good days and bad days. They’ve got their own jargon, their own shorthand, a specific way they complain about a problem to a coworker. They’re out there right now looking for solutions to what’s been holding them back. And you can’t market to someone you don’t actually understand.
So how do you get from a vague job title to that kind of understanding? You run audience research prompts. It sounds like a big, dry exercise — the kind of thing that ends up as a slide nobody reads twice. But it doesn’t have to be. And the good news is, AI can do most of the heavy lifting for you.
Step 1: Gather real inputs.
AI can’t run a compelling analysis without something real to chew on. So here’s the temptation: you’ll want to hand it your own take on your audience. Your personas, your gut instinct, the mental sketch you’ve been carrying around for years. Don’t. If your picture of your audience is off — too general, too shallow, built on assumptions instead of evidence — the AI just runs with your mistake and hands it back to you dressed up nicer. Garbage in, polished garbage out.
What you actually need is accurate, raw material. Words straight from your audience’s mouth, or their keyboard. Try:
- A LinkedIn post from someone who fits your ideal customer
- A customer review — good or bad, doesn’t matter
- A Reddit thread where people are venting about the exact problem you solve
- A customer email or Slack message that felt a little too honest
That’s your data. AI can’t invent this for you, no matter how good the prompt is. You have to walk in with something true.
Step 2: Run the audience research prompts.
Now paste those inputs into AI and put it to work. Ask it to pull out the exact words your audience uses when they talk about their problems. Not the polished version — the raw one. The stuff they’d say to a coworker, not a vendor. What are they afraid of? What are they sick of hearing from people trying to sell them something? What do they wish someone would just say plainly, for once?
A few things make this step actually work. Be specific about what you want back — ask for direct quotes and recurring phrases, not a summary of themes. Tell it to ignore anything that sounds like marketing language; if a sentence could’ve come from a press release, it’s not useful here. And push it to flag the emotional undercurrent too, not just the words themselves — frustration reads differently than confusion, even when the vocabulary overlaps.
What you get back is your real content vocabulary. Not the one from a strategy meeting last quarter. The one your audience is actually using, right now, to describe the problem you solve.
Step 3: Build your frustration map with the prompt.
Now take all that raw language and turn it into something you’ll actually use. Not another report that sits in a shared drive collecting dust. Something simple: three fears, three things they’re flat-out tired of reading, three questions they genuinely want answered. That’s it. That’s the whole map.
Keep it to one page. Pin it next to your monitor, tape it to your laptop, whatever gets it in your line of sight. Because here’s the real test — before you hit publish on anything, glance over at that page. Does this piece answer at least one of those questions? If it doesn’t, that’s worth catching now, not after it’s already live and doing nothing for anyone.
Step 4: Check it against real life.
Last step, and don’t skip it just because you’re eager to move on. Pull up actual conversations — real ones, with real customers or prospects. Not the polished case study version. The messy Slack thread or the call recap you jotted down five minutes after the meeting ended.
Compare it to what the AI handed you. Where it matches up, good. You’ve found solid ground, and you can build on it with some confidence. Where it doesn’t match? Trust the conversation. Every time. AI’s a mirror here, not an oracle — it can only reflect what you fed it, and if what you fed it was slightly off, the reflection will be too. A real conversation with a real person will always outrank it.
That’s it. That’s the whole process. Four audience research prompts, one page at the end, and suddenly you’re not guessing at who you’re writing for anymore. You actually know. And that knowing is going to hold up a lot better than the version you had floating around in your head before we started digging.
Mapping the Competitive Landscape with AI
Here’s the second question worth asking before you build anything: is anyone already saying this?
Your competitors are cranking out campaigns constantly. You know this. You’ve probably got a browser tab open right now to prove it. And here’s the uncomfortable part: most of it is commodity content. Interchangeable. Swap the logo, and nobody would notice the difference. Same tone, same topics, same three-tips-and-a-CTA structure wearing a different color scheme.
That’s not actually bad news, though. It’s an opening.
Commodity content is what happens when everyone’s afraid to say anything real. The angles nobody’s taking. The opinions nobody’s willing to put their name on. The questions everyone raises, and nobody actually answers. That gap is where non-commodity content lives — the stuff that’s genuinely yours, that couldn’t have come from anyone else’s brand voice or strategy deck. And finding exactly where that gap is? That’s a job AI can help you do fast, if you know how to ask.
Step 1: Collect competitor content.
Pull the last ten blog posts from your top three competitors. Don’t bother copying full posts. Just grab the headlines and opening paragraphs and drop them into a document. That’s genuinely enough. A headline and an opener will tell you almost everything you need to know about what someone is trying to say, and you don’t need three thousand words to confirm it.
Step 2: Run the content gap prompt.
Feed AI those headlines/opening paragraphs and ask it to break down the landscape. What topics is everyone hammering? What tone have they all seemingly agreed on without ever actually agreeing on it? Then ask the two questions that actually matter: what position is nobody taking, and what questions does this content raise but never answer? Those last two are your opening. Everything else is just noise everyone’s already making.
A few things make this step land better. Be explicit that you want gaps, not summaries. If you just ask “what’s this content about,” you’ll get a book report. Ask instead for what’s missing, what’s unclaimed, what’s being dodged. And push AI to separate “nobody’s saying this” from “nobody’s saying this well” — those are two very different opportunities, and conflating them will send you chasing the wrong angle.
Step 3: Map the contrarian angles.
Here’s a unique way to turn your competitors’ “truth” into your advantage. Take the single most common claim you found floating around your competitors’ content — the one everyone repeats like it’s gospel. Drop that claim into AI and ask it to generate three different ways to push back on it. I don’t mean hot takes or contrarian clickbait. Real, defensible challenges: a way to complicate it, a way to flip it, a way to argue the opposite and actually mean it.
Then read through what it gives you and ask yourself the only question that matters: do I actually believe any of this? AI can hand you three sharp angles in ten seconds. It can’t tell you which one is true for you. That part is still yours. If none of the three angles ring true — if you read them and think “eh, I don’t really disagree with the competition here” — that’s not a failure. That’s information. Keep looking.
Step 4: Build your differentiation brief.
Here is the culmination of all your hard work, with a little help from AI. Feed it everything you’ve gathered so far — the competitor gaps from Step 2, the angle you landed on in Step 3, and a few notes on your own experience or opinion — and ask it to help you structure a single-page brief. Three sections: here’s what everyone else says, here’s what I believe instead, here’s the specific experience or evidence that backs it up.
AI’s good at this part. It’ll take your scattered notes and turn them into something clean and usable. But the belief in the middle section? The actual opinion? That has to come from you. AI can organize conviction. It can’t manufacture it. This brief becomes the starting point for every piece you write from here on, so it’s worth getting right the first time.
Foundation Generation Gets Personal
Here’s the truth about this kind of work: it’s boring to explain and impossible to skip. Nobody’s going to praise your frustration map. Nobody’s mentioning your competitor research in a meeting. It just sits there, quietly doing its job. And its job is making sure everything you build next actually stands.
You know who you’re talking to now, really talking to, not the sanitized version from a slide. You’ve seen what everyone else in your space is already saying, so you’re not wasting a sentence on repeats. That’s the foundation. It might not be flashy, but it is load-bearing.
In Part 2, we’ll take a more introspective approach to foundation work. We’ll define your brand’s voice and establish what your brand actually stands for. Of course, AI can help you get there faster, same as it did here. It’ll spot patterns you might have missed on your own and save you hours you didn’t have to spend. But it can’t tell you what your brand believes. It can’t decide what makes you different. That part’s still yours. It always was.
If you enjoyed this content and want marketing know-how delivered straight to your inbox, you’ll love our Marketing Minds newsletter. Sign up here>>
