Here's a thought that should keep every CMO up at night.
Your customer just asked an AI what product to buy. The AI considered thirty options, synthesized ten thousand reviews, and handed back a shortlist of three. Your brand wasn't on it. And you never even knew the conversation happened.
This isn't a hypothetical. It's already Tuesday morning for millions of buyers.
Bain & Company found that 80% of consumers now rely on AI-generated results for at least 40% of their searches, and 60% of searches end without the user clicking through to any website at all. Six out of ten searches. No click. No visit. No impression. No chance to tell your story. The buyer got what they needed from the AI summary, formed an opinion about your category, and moved on, possibly to a competitor, without you ever entering the room. Organic traffic is down 15–25% across the board. Meanwhile, traffic from generative AI sources grew 1,200% between mid-2024 and early 2025. The river is moving. Most brands are still fishing in the old spot.
Your new buyer isn't always human.
In an agentic world, your customer may no longer be a human with a browser. It is just as likely to be an autonomous agent acting on that customer's behalf. Discovery, evaluation, and shortlisting all happen inside the model before you ever get a chance to engage. In B2B, click-through rates have already fallen 30% in some categories, and 85% of B2B buyers ultimately purchase from their "day one" list of vendors they had in mind before they even started searching. If you are not in the consideration set of the AI constructs at the first prompt, you may never recover the chance.
Generic positioning is now fatal.
David Ogilvy said that if it doesn't sell, it isn't creative. In 2025, there's a corollary: if an AI can't describe you in one specific, verifiable sentence, you don't exist.
LLMs are ruthless aggregators. They synthesize what's said about your brand through external reviews, third-party editorial, forum discussions, and comparison articles, and they reward specificity and penalise vagueness. One vague brand claim and you're out. The classic positioning template "We are a [category] that delivers [benefit] for [audience]" doesn't survive when an algorithm does the shortlisting. What survives is concrete proof, earned authority, and a distinct point of view that shows up consistently in places the AI actually reads. Broad awareness campaigns, emotional resonance, top-of-funnel affinity all of it depreciates when the gatekeeper doesn't feel anything and can't be charmed.
The L'Oréal lesson.
L'Oréal built Beauty Genius, an AI tool that personalises skincare routines based on facial scans and customer input. But here's the strategic move most people miss: customers' prompts provide L'Oréal with data for future offerings and inform how the company talks about them. L'Oréal isn't just serving customers through AI.
It's learning the language its customers use when talking to AI, and building products and copy around that language. They're not waiting to be discovered by algorithms. They're co-authoring the vocabulary those algorithms use to describe the category. That's what it looks like to build a brand for an AI-mediated world.
The BCG gap nobody wants to talk about.
There's a widening divide playing out right now that has nothing to do with creativity and everything to do with commitment. BCG found that only 5% of companies qualify as genuinely "future-built" for AI, and only 22% have advanced beyond the proof-of-concept stage. The other 78% are still running AI pilots and calling it a transformation.
The future-built companies are achieving 1.7x revenue growth, 3.6x three-year total shareholder return, and 1.6x EBIT margin compared to their slower-moving competitors. That gap is widening every quarter.
So what does winning look like?
Three things.
First, you need to exist in the sources that AI trusts. Not just your own website. Reviews, forums, industry publications, comparison platforms, expert citations. Bain identifies that LLMs prioritise conversational language over static assets, agent-friendly structured content, clean and fully indexed websites, off-site earned authority through external expert citations, and strong brand presence in external reviews and forums.
Second, you need a genuinely specific claim to own. Not "we're customer-centric." Not "innovative solutions." Something concrete enough that when a buyer types a high-intent prompt into ChatGPT, your brand surfaces as the obvious answer to that specific problem.
Third, you need to act on the data signals that AI creates. Buyers are using AI to construct the consideration set, then turning to websites, review platforms, and YouTube demonstrations to validate what the model suggests. If a brand doesn't surface in that first AI-generated list, it may never make it to the validation stage. This is a completely new kind of funnel. You need new metrics to see it.
The honest conclusion.
Most brands will lose this transition slowly, and then suddenly. The clicks will keep declining. The old KPIs will show green, right up until they don't.
The brands that win won't have the biggest budgets. They'll have the sharpest answer to the hardest question of this era: What would an algorithm say about us if it read everything ever written about our brand?
