Brand Narrative in the Age of AI Discovery

In your customer’s mind, you get to be one thing. Your company may do a hundred things, but expecting someone to remember all of them is a fairly ambitious marketing strategy.

I wrote about this back in 2012, after a conversation with a prospective client who told me her messaging was finished. What she actually had was a handful of product attributes her competitors could have claimed just as easily. She wanted help driving revenue. I thought we needed to spend a little more time on why someone should choose her company in the first place. Traction Co.

I recently joined an American Marketing Association panel during San Francisco Tech Week to discuss how AI is changing marketing and branding. Our moderator, BlueOcean CEO Grant McDougall, asked a question that got me thinking about something I've believed for a long time.

What does AI actually change about branding?

Plenty, as it turns out. But not necessarily the things people think.

David Ogilvy described a brand as the intangible accumulation of a product's attributes. Al Ries focused on the singular idea a brand owns in a prospect's mind. Ann Handley has described the emotional impression a brand leaves behind.

Different definitions, same essential truth: your brand lives in your customer's mind, not in your PowerPoint presentation.

People don't have room in their heads for your capabilities deck. They remember something simple. Volvo means safety. Apple means design. Palmolive means soft hands while you do dishes.

When a company insists on standing for four different things, it's usually remembered for none of them.

AI makes this principle even more important. Increasingly, your customer isn't starting with your homepage. They're asking ChatGPT or Claude for advice. Instead of seeing your carefully constructed messaging, they're getting a summary assembled from information about you. You can spend a fortune getting your messaging just right, only to have a machine explain it to a potential customer in two sentences.

Your brand is now something a machine paraphrases.

That has serious implications for how we write brand narratives, messaging and identity systems. In the mind of an AI agent, you also get to be one thing. But most of them aren't about chasing algorithms. They're about doing a better job of the things branding was always supposed to accomplish.

Branding is not brandbuilding.

I was on a panel during San Francisco Tech Week for the American Marketing Association. The moderator, BlueOcean CEO Grant McDougall, asked me how AI was impacting branding. My answer started in a place most marketers skip over: there is a difference between branding and brand building, and confusing the two is what gets companies into trouble.

Branding is your foundation. It is the act of trying to place your one thing in your customer's brain: what you say, how you say it, and what you look like. Your essence to them. Branding is the inputs. The brand is the output.

AI makes the distinction unavoidable. The foundation work now has to be sharp enough that when Claude, Gemini, or ChatGPT compresses what the world has to say about you, including what you say about yourself, the output matches what you intended. If your branding is muddy, the brand a model builds for you will be muddy too.

Brand-building is the work of earning a place in someone's mind. It now includes earning a place in an agent's understanding too. The same principles apply: consistency, clarity, focus. Those are not just good branding hygiene. They are what give you greater visibility and relevance when an agent is fielding questions in your domain.

Why AI makes the "one thing" rule brutal.

AI makes this principle brutally literal. When a model is asked to synthesize everything written about you into a sentence or two, it does exactly what a human customer's brand does, but faster and at scale. The implication? If you don't decide what your one thing is and build a brand narrative to support it, the model will decide for you. It will usually pick whatever it reads the most about you in the wild.

I learned this the hard way at my own company. For years Traction described itself as a full-service agency that did strategy, creative, digital, media, and content. It was all true, and it was useless. Nobody could repeat it, so nobody did.

When we reinvented ourselves in 2019, we looked at 25 years of work and found one truth: almost all of it helped brands through some kind of transition, and all of it was high stakes for the people involved.

We recast ourselves as a marketing accelerator built for the moment you can't afford to lose. In three years, our revenue doubled.

What is a brand narrative, and why does it matter more now?

A brand narrative is a story. It tells who your customer is, what problem you solve for them, how you do it, and why you do it.

It is the story that lets everything you say sound like it came from the same place. At Traction, we build that story around the psychology of how people form attitudes, perceptions and behaviors. We look for the intersection of what matters to customers, what we can credibly promise them, and what will motivate them to act.

In an AI-mediated world, those four questions turn out to be exactly the questions a model is trying to answer when a buyer asks it about you. A brand with a strong narrative hands your customer a clear reason to choose and hands an LLM clean answers, consistently, across every source it reads. A brand without one leaves the model to stitch together an answer from a press release, two product pages, and a Reddit thread, and the result sounds like it.

A strong narrative answers four questions:

The part I see teams skip most often is deeply understanding their customer. They jump straight to what they do and how they do it, and the customer becomes a vague player somewhere in the background.

But a narrative that opens with a sharply drawn customer does something a feature list never can. It lets the person reading, or the model summarizing, recognize who the brand is for in one beat. When we do Brand Foundation Design at Traction, that is where we spend the most time, and it is almost always where the client's existing messaging is thinnest.

The why matters more than it used to, too. Models are getting good at surfacing specific, verifiable claims and dropping adjectives. "Innovative" and "customer-centric" do not survive compression. A clear statement of why you exist and what you believe does, because it is distinctive by nature. Nobody else can say it the same way.

How to design messaging for AI discovery.

Write messaging that survives the paraphrase. Brands often fall into one of two traps: emotional messaging that doesn't drive demand, or practical messaging that gives people no reason to care. AI makes that tension more important because you need to be memorable to humans and understandable to machines.

Here's how to approach it.

As my colleague Lauren Evans puts it on our site, the brands that win will have the best story, not the best technology. The trap is over-rotating toward the machine. Humans still make decisions, and they make it on memory and feeling.

Why brand identity still matters.

I don't buy the argument that visual identity becomes less important because people are discovering brands through AI. That's like saying packaging doesn't matter because someone recommended a product to you over the phone.

AI may be responsible for introducing someone to your brand. But when that person visits your website, sees your advertising or uses your product, the experience still has to deliver. And with AI making it easier than ever to create competent-looking design, being distinctive is arguably more important than ever.

Good design is a minimum standard. Tools like Lovable, Claude, and Figma make competent design easy. Great design is what sets you apart.

You also need to build a system. AI-generated assets drift quickly without a tight, well-documented design system in a tool like Figma. You can import that to Claude Design. Your visual ID is not a brand book. A PDF Style Guide collects dust. You need a system.

Traction is now making skills part of your brand kit. Put your brand guidelines, voice rules, and approved claims into reusable skills that AI tools load automatically, so every draft and generated asset follows the same standards. Custom prompts for GenAI tools should become part of the visual design system. For example, we included MidJourney style reference codes in the brand guidelines for Recorded Future so their team could consistently create assets

Where TractionOS fits.

TractionOS is our AI Strategy platform we use for to accelerate strategy work. It takes the principles above and puts them to work: positioning, brand narratives, audience research, and answer-engine presence, connected in one system. Four ideas are built into its foundation, so the brand strategies we create with it are inherently discoverable:

Can AI replace talking to customers?

Absolutely not. AI does not replace talking to customers. If anyone tells you otherwise, kick ‘em in the shins. AI can identify patterns across conversation and it can help you make sense of what you heard. It cannot sit across from a buyer and watch them hesitate before they give an answer the way a human being who is a great listener can.

I say this as someone who runs an AI-powered accelerator and uses these tools all day. We use AI to synthesize research, pressure-test personas, and draft discussion guides in a fraction of the time it used to take. It is a real advantage.

But every time we have gotten a brand narrative right, the breakthrough came from something a customer said out loud that was not in any dataset; a phrase they used that we would never have written; a problem they described that was two levels deeper than the one on the RFP.

You cannot prompt your way to that. You have to ask, and then shut up and listen.

There is a second reason this matters more now, not less. If you build your narrative from the same public online presence the models are trained on, you end up describing yourself the way everyone else in your category already does. The model has no reason to prefer you.

So when a client asks whether they can skip the interviews because we have AI now, my answer is that AI is exactly why they cannot. The machines are going to summarize whatever you give them. Make sure what you give them came from a human being who actually buys your product.

Drop us a line to start a conversation about your brand →