"AI-powered" is not a positioning strategy.
AI is not a positioning strategy. It's a mechanism. The companies winning on AI aren't the ones leading with the technology. They're the ones who found the buyer problem first and used AI to solve it better than anyone else.
Twenty years ago, I was selling the benefits of “cloud computing” to large enterprises with Citrix Systems and VMware (do you remember those guys?). Then "mobile-first" became the differentiator every agency put in their pitch deck. Then "digital marketing" became a specialisation worth calling out explicitly, a signal that you understood the internet as a channel, not just a medium.
Each one became invisible. Nobody says "our software is cloud-based" anymore because everything is. Nobody says "we do digital marketing" because that's just marketing now. The label that used to differentiate became the thing everyone has, and the thing everyone has stops being a differentiator.
AI is following the same arc.
Right now "AI-powered" looks like a differentiator. In three to five years (I think, but I’m no wizard with a magic eight ball), every product will have an AI layer. The companies that built their positioning around being AI-native will have a problem: the thing they led with will be the thing everyone else has too.
Three types of AI companies, three different positioning problems
Not every AI company has the same challenge. There are three distinct groups I’m noticing, and the right approach is different for each.
The first is AI infrastructure and foundational model companies. The ones building the layer that other products run on. Their buyers are developers and technical architects who are evaluating a product they'll build on top of. Speed, model performance, ease of deployment, security posture, API design — all of it matters, and all of it belongs on the page. Leading with the technology is correct here; it makes sense. The buyer needs the specs. That's the purchase decision.
The problem is that founders of AI point solutions and established companies watch the infrastructure players lead with technology and assume that's how AI companies are supposed to position. It isn't. It's how infrastructure companies position, for a very specific technical buyer who is actively shopping for those specs.
The second group is AI-first point solutions, mostly built in the last 12 to 24 months. Products born in the AI wave, leading with "AI-powered X" or "the AI agent for Y." Their buyer is not a developer evaluating infrastructure. It's a business buyer with a specific problem. That buyer doesn't care about the model. They care about whether their problem gets solved. Most of these companies haven't answered that yet. They're still describing what they built rather than who needs it and when.
The third is established companies adding an AI layer to an existing product. Real customers, real positioning, real proof. Then the pressure arrives from investors wanting an AI story, from competitors announcing AI features, from boards asking "what's our AI strategy?" And so the website gets updated to follow suit. A new badge. A new hero section. "Now powered by AI."
But the update happens before anyone has asked what the AI layer actually means for the buyer. What changes in their world. What problem it solves that wasn't solved before. So the homepage ends up holding two stories that weren't designed to sit together. The core product story is still there. The AI announcement is layered on top. Neither one is doing the job of making a buyer feel like they're in the right place.
The identity crisis goes deeper than the homepage. Are we a platform that uses AI or an AI platform that does what we used to do? Most try to be both. The homepage reflects the internal confusion rather than resolving it. And the buyer, who came looking for a solution to a specific problem, leaves with a vague sense that something changed but not why it matters to them.
All three groups, except the first, are making the same mistake: leading with the technology before establishing who the technology is for.
The companies that got it right
Two AI companies that have figured this out, in my opinion, are Heidi Health and Legora. Neither leads with AI. Both lead with the specific pain of the specific buyer they serve.
Heidi Health is an AI medical scribe. Their mission is doubling the world's healthcare capacity. Their entry point is the doctor burning out. 13 hours a week going to indirect care, another 7 to admin, with patient demand still rising. The problem is the message. The AI is invisible in the positioning. Their tagline is "By your side." Completely human. Zero technology language. A clinician landing on that page knows within seconds this was built for their world.
It works. Now, I’m not a trained medical practitioner nor a clinician in their target audience, but it’s so clear to me why they matter.
Legora positions around specific legal practice lanes: M&A, Litigation, Banking, Tax, Insurance. Not "AI for lawyers." The work that lawyers do, and the specific workflows where the time goes. The AI is the mechanism. The positioning is about the work getting done.
And have you seen the latest brand campaign featuring Jude Law? It's brilliant — and not just because it's entertaining.
Lawyers are known to be a conservative, serious, dry bunch (side note - I’m married to one). The brightest minds in the room, but rarely the ones making bold creative moves. Every legal tech company looks and sounds exactly like the industry it serves — measured, formal, cautious. Legora did the opposite. They built an entire global campaign around a pun. Jude Law. Law just got more attractive. The creative director at the agency said it out loud: "Surprisingly often, a dumb idea turns out to be the smartest strategy."
It's not dumb at all. It's strategically precise. The pun only works if the audience is specific enough to get it. You can only be that playful externally if you're that clear internally about who you are and who you're for. The campaign is brand confidence. The positioning clarity underneath it is what made it possible.
Both companies made the same call: lead with the buyer's world, not the technology. The AI reveals itself once the buyer feels found.
The trigger moment
Some AI companies understand the problem deeply.
The copy is sharp. The pain is real. But the problem is written for nobody in particular. "AI is changing everything and most companies aren't ready." True. Compelling. But who specifically? Which company? Which team? Which moment? The urgency is there. The buyer isn't. So it dissipates. The reader feels the category pain but can't tell if it's their pain. They keep looking.
The trigger moment is what fixes this. It's the specific situation that sends a buyer looking. Not the problem in general. The moment. "When your doctor finishes a consult and still has 40 minutes of notes ahead of them." "When a partner asks for a first draft of a due diligence memo by morning and the associates are already at capacity."
Those lines name a role, a situation, a specific pain. And you can only write them if you've already decided who you're talking to. That's what makes the trigger moment such a useful discipline for AI companies specifically. It forces both decisions at once — the buyer and the moment they need you. Most AI companies can't write one yet. Not because they don't understand the problem, but because they haven't committed to the buyer the problem belongs to.
If you can't write your trigger moment, you haven't made the positioning decision yet. The technology is ready. The positioning isn't.
What to do with this
Alright, so here’s the “so what."
For established companies adding AI: the question is not "how do we talk about our AI features?" It's "what specifically changes for the buyer when AI is in the product?" Not features. Not capabilities. What changes in their world on the day they start using it. Start there. Build the AI story backwards from that.
For AI-first companies: strip "AI" out of your hero. Read what's left. If nothing coherent remains - no named buyer, no named problem, no named moment - the positioning was always about the technology rather than the person who needs it. Start with the problem. Start with the moment the problem becomes urgent. The AI story follows naturally once the buyer feels found.
The companies that survive the moment AI becomes invisible are the ones who never needed it as a headline in the first place.
