There’s a question I keep hearing in every search conversation: “How do we rank in ChatGPT?”
I get why people ask it. It’s most people’s muscle memory of traditional SEO. But I think it’s a question that stops one step too early.
Two decades of chasing the same answer
For most of my career, a big part of the awareness job came down to one thing: help a search engine find the right page (aka our page) and put it in front of the right person. Keywords, backlinks, technical SEO, content calendars, all of it, in service of a single test:
“Which page best answers this query?”
That question built an entire industry, and it’s not going away. Despite all the noise around it, I’ve come to understand that AI assistants and AI search experiences still lean heavily on the same infrastructure traditional SEO created that’s crawlable, well-structured, authoritative content. Feed an AI system a thin, chaotic, inconsistent web presence, and it has nothing solid to work with. So no, SEO isn’t dead. I’d argue it’s more foundational than ever. And that’s exactly what I always tell my team @datawrkz.
But something underneath it has shifted, and it’s easy to miss if you’re still measuring success the old way. SEO optimized retrieval. Now, AI optimizes reconciliation.
A different question is now being asked
When someone asks an AI assistant about your company, it’s not retrieving a page. It’s doing something closer to an evaluation, sometimes even forming a judgment:
“Based on everything I can find, what is this company, and how should I describe it?”
That answer doesn’t come from your homepage. It’s stitched together from your site, press coverage, analyst notes, G2 reviews, a Reddit thread from years back, your LinkedIn posts, a partner’s case study, maybe a disgruntled former employee’s blog.
AI doesn’t know which of those sources you’d consider “official.” It just sees signals, and it reconciles.
Bottom line: The raw material your brand is understood through is no longer entirely yours to control.
Your website was never the whole story and it’s more obvious now
This isn’t entirely new, to be fair. Word of mouth, analyst reports, and review sites have shaped brand perception for years, long before generative AI showed up. What’s changed is the speed and the visibility of the reconciliation.
A human reading five different sources over a week might not notice the contradictions. An AI model reconciling those signals into a few paragraphs in three seconds makes every inconsistency immediately legible.
If your website calls you “the enterprise leader in X,” your latest funding announcement calls you “an emerging challenger,” your support docs read like a startup, and your reviews complain about the exact thing your homepage claims you’ve solved, the model doesn’t get to ask you which one is true. It averages, hedges, or picks whichever source seems most confident.
In other words, when faced with ambiguity, AI will resolve it using whatever signals it considers most reliable. That may not be the story you intended to tell.
So maybe the sharper question for a marketing leader is:
“If I asked an AI to describe my company right now, with no context from me, would it get it right?”
I now do this as a monthly exercise. I ask different AI tools what our company does, who it’s for, and how it compares to competitors. I ask them to recommend companies as a customer looking for our solutions. I ask why it wouldn’t recommend our company. And I try to do this from an AI chat interface that doesn’t know me already (I used my daughter’s laptop).
The gaps between their answers and your positioning deck are usually the most honest audit you’ll get all year.
What this actually asks of marketing teams
What most viral content refers to as a call to abandon content or SEO for some vague new discipline called “AI optimization” is better defined as an expansion of scope. A few of the harder, less comfortable questions this raises:
- Does your executive team’s public messaging match what your website says, or has it drifted?
- Are your website and content crystal clear about your category and what you do, and who you do it for?
- Do industry analysts and journalists describe you the way you’d describe yourself? If not, why not?
- Are you still producing keyword-driven content, or are you answering the underlying questions your customers are actually trying to solve?
- Are your customers’ reviews reinforcing your value proposition, or contradicting it?
- Is your content easily machine-readable, in that it provides straightforward answers? Or are you still hoarding content behind sign-up walls and downloads?
- Is your competition shaping what AI knows about your category?
None of these are new questions, really. Brand consistency has always mattered. What’s new is that inconsistency used to be diffuse, spread across a hundred scattered impressions nobody quite added up. Now it gets compressed into a single, confident-sounding paragraph that someone might trust more than they should.
But consistency doesn’t require uniformity
It’s worth naming the tension here: consistency and authenticity can pull in opposite directions. , The goal isn’t to scrub every source until they’re identical. It’s closer to making sure the substance holds together even when the tone and format differ. Your marketing team, your CEO, and your happiest customer can sound different and still be describing the same real thing.
There’s also a genuine open question about how much any of this is measurable or actionable today. AI answers change from model to model, session to session, and prompt to prompt. Anyone selling a tidy framework for “AI brand optimization” right now is getting ahead of the evidence.
The AI search discipline is still being figured out in real time, including by people writing posts like this one.
SEO isn’t becoming obsolete
I’ve come to think of SEO as the floor instead of the ceiling, i.e., necessary infrastructure that AI systems still depend on, but no longer sufficient on its own.
What sits above it is something marketing has technically always been responsible for, just without this much pressure: making sure that everywhere someone encounters your company, they encounter roughly the same truth. Not identical words. The same truth.
Visibility used to be enough. But now, it’s the entry fee. Increasingly, consistency determines whether AI understands who you are. What actually earns trust, whether from a human or a model synthesizing multiple sources, is whether your story holds together when nobody controls which piece gets read first.
