In the spring of 2003, I placed my first paid search bid on a platform called Overture. You picked a keyword, set a price, and if someone typed those words into a search engine, your ad appeared. It was beautifully simple. And for the next quarter century, every major innovation in digital advertising was essentially a more sophisticated version of that same idea: figure out what someone wants, show up at the moment they're looking for it, and make the path to conversion as frictionless as possible.
I spent 25 years helping grow companies using that path. I built the keyword data models, attribution systems, and performance frameworks that turned clicks into revenue. Hundreds of millions in ad spend, optimized down to the basis point.
That principle served us well. It still matters. But in 2026 — the ground has shifted beneath it.
We are no longer optimizing the customer experience. We are now optimizing for the agent experience — the way AI systems perceive, evaluate, and represent your brand before a human ever enters the picture.
This is the shift from CX to AX. And it's the most consequential change I've seen in my career and it's why I'm passionate about helping companies navigate this change.
2026: The Inflection Point — When AI Became the Customer
Here's what most marketers haven't fully internalized yet: the customer journey you've been optimizing no longer starts with a search result. Increasingly, it starts — and sometimes ends — inside an AI conversation.
A customer asks a question. AI recommends a product. The customer buys it. The transaction completes. Your website was never part of the conversation.
Your landing page didn't lose. It was never in the game. The AI made the decision about which brands to recommend before the customer even knew they were shopping.
This is why the CX-to-AX shift is not incremental. It's structural. For 25 years, brands competed on who could deliver the best experience after the click. Now the competition is for who the AI recommends before there is a click.
Defining AX: What Does Your Brand Feel Like to a Machine?
Customer Experience (CX) was always about the human perception of your brand across touchpoints — how your site loads, how tailored is the landing page to their search/need, how easy it is to check out. We measured it with conversion rates, ROAS, time on site, CLV.
Agent Experience (AX) is the machine's perception of your brand across data surfaces. It's how an AI system evaluates whether your claims are credible, your data is structured, and your authority is verifiable. AX determines whether you even make it into the consideration set.
Think of it this way: CX answered the question, "Once a customer finds you, will they choose you?" AX answers a different question entirely: "Will the AI find you, trust you, and recommend you in the first place?"
In my work with clients across retail, B2B, and enterprise, I've identified three elements that determine AX. I call it the Visibility Framework:
1. On-Site Structure
AI doesn't browse your site the way a human does. It parses it. Schema markup, structured data, and properly formatted content allow AI to extract your product details, specifications, and key claims with precision. If your site is optimized for visual appeal but not for machine readability, you're invisible to the systems making recommendations.
2. External Validation
This is the piece most brands underestimate. Answer engines don't just read your content — they verify it. They cross-reference your claims against third-party sources: reviews, industry publications, authoritative mentions. Without external corroboration, even accurate information gets deprioritized. Your PR strategy is now, functionally, an AI visibility strategy.
3. Technical Infrastructure
Structured data feeds, clean product taxonomies, and proper technical foundations allow AI systems to access your inventory, pricing, and offerings without ambiguity. As we move toward Universal Commerce Protocol (UCP) and Agentic Commerce Protocol (ACP), this infrastructure becomes the interface through which AI agents interact with your business directly.
When all three elements are synchronized, your brand is one that machines can find, verify, parse, and recommend with confidence. When any element is weak, you have a visibility gap that no amount of ad spend can fill.
The Measurement Problem: Your Dashboard Has a Blind Spot
One of the most dangerous aspects of this transition is that your existing analytics won't show you it's happening. Google Analytics tracks sessions, clicks, and conversions from channels you know about. It has no mechanism to measure whether ChatGPT recommended your competitor, or whether Gemini cited your brand in a comparison, or whether Perplexity used your content to answer a query without sending any traffic your way.
In the CX era, "if we can't measure it, we can't manage it" was gospel. In the AX era, the most important interactions are happening in places your current tools can't see.
This is why we have adapted our measurement approaches around new metrics: Share of Answer (how often your brand appears in AI-generated responses), citation frequency (how often third-party sources that validate your claims are referenced), and recommendation rate (how consistently AI systems include you in relevant comparisons). These aren't vanity metrics. They're the leading indicators of how much revenue you will generate from LLMs in the quarters to come.
What This Isn't
I'm not saying CX doesn't matter. If a customer lands on your site and the experience is terrible, you'll still lose. I'm not saying paid search is dead tomorrow. There's a meaningful runway left. And I'm not saying AI is going to replace every human interaction with your brand.
What I am saying is that the sequence has changed. CX used to be the first battle — win the experience, win the customer. Now AX is the first battle. Win the AI's trust and recommendation, and then CX takes over for the human experience that follows. Lose at AX, and your customer experience never gets a chance to compete.
The brands that understand this will invest in both. The brands that don't will watch their traffic decline and blame the algorithm.
The Agentic Horizon: Where This Is Headed
If you think the shift from search to AI answers is significant, the next phase will be even more disruptive. We're moving toward agentic commerce — a model where AI agents don't just recommend products, they compare, negotiate, and purchase on behalf of the user.
Imagine a near-future scenario: a customer says, "Buy me the best stainless steel cookware set under $200." An AI agent queries product feeds via UCP, compares specs and verified reviews, checks inventory and pricing in real time, and completes the purchase via ACP. The customer never browsed. Never compared. Never visited a website. The agent did all of it.
And your brand either had the structured data, the verified authority, and the machine-readable infrastructure to be in that agent's consideration set, or it didn't.
This isn't science fiction. The infrastructure is being built right now. Stripe checkout in ChatGPT was the first transaction layer. Affiliate fee models are the first revenue layer. Clean feed standards are the first data layer. By late 2026, the full agentic stack will be operational for early adopters.
What Should You Do About It?
After 25 years of navigating platform shifts, I've learned that the companies who move early don't have to move perfectly. They just have to move before the window closes and the cost of catching up becomes prohibitive. Here's where I'd start:
Audit your AI visibility today.
Query the major LLMs with the questions your customers ask. Are you in the answers? Are your competitors? If you don't know, you're already behind. This is the equivalent of checking your search rankings in 2004 — except the stakes are higher because there's no page two in an AI answer. You're either the recommendation or you're not.
Fix the structural foundation.
Schema markup, structured data, entity relationships in the Knowledge Graph. This is the technical infrastructure that makes your brand parseable by machines. It's not glamorous work, but it's the bedrock of AX.
Build your authority layer.
Invest in the third-party validation sources that AI models use to verify claims. Strategic placements in authoritative publications, verified reviews, industry citations. This is the corroboration layer that separates brands AI trusts from brands AI ignores.
Prepare your data feeds for agentic commerce.
If you sell products, start thinking about how your inventory, pricing, and specifications would be consumed by a machine that has no browser and no eyeballs. Universal Commerce Protocol readiness is coming faster than most retailers expect.
Build new measurement systems.
Share of Answer. Citation frequency. Recommendation rate. If these metrics aren't on your dashboard, you're flying blind into the single biggest channel shift since the launch of AdWords.
I've been doing this long enough to know what a genuine inflection point feels like. I was there when Overture proved that intent-based advertising could work. I was part of the AdWords beta and helped multiple companies scale on the back of keywords. I was there when Facebook monetized their platform with ads, and when automated bidding made human campaign management less important.
Each of those shifts rewarded the practitioners who recognized the new reality early and adapted their frameworks accordingly.
The shift from CX to AX follows the same pattern — but it's deeper. It's not just a new channel or a new platform. It's a new intermediary between your brand and your customer, one that makes autonomous judgments about your credibility, your relevance, and your worthiness of being recommended. Getting that relationship right is the work of the next decade.
This isn't the end of advertising. It's the end of advertising as you've known it. And for those willing to adapt, it's the beginning of something better.