Your AI salesforce is already selling your brand. The question is who trained it.

This series began with an observation: AI systems don’t always give the same answer to the same question. In the first article, I argued that the inconsistency wasn’t randomness. It was confidence loss across a measurable pipeline we can diagnose and fix.

Working through the AI engine pipeline, gate by gate, eventually led me to the won gate, where three clicks await: the imperfect click of search, the perfect click of recommendations, and the agentic click of agents. 

That’s where I realized the series couldn’t stay within marketing. When an agent makes the purchase, it becomes a client you need to satisfy directly.

The funnel now runs through a machine that plugs into the business itself. SEO has no choice: it becomes assistive agent optimization, part of how the entire business is engineered.

To understand why, it helps to see how the underlying concepts fit together.

The framework is the theory. It explains why AI systems make the decisions they make and what influences those decisions. Apply those principles across the business, and you get AI-era business engineering: a company organized so search engines, AI assistants, agents, and people can all find you, understand you, recommend you, and buy from you.

Everything builds on SEO

The process is the day-to-day methodology. It sits above your existing disciplines — SEO, content, PR, paid media, and digital marketing — helping you prioritize the actions that have the greatest impact on recommendations and visibility.

Here’s the part every SEO should appreciate: assistive agent optimization is built on SEO. It doesn’t replace it.

Picture a Russian doll. SEO sits at the center. It draws from the open web, the same crawled and indexed foundation search has always relied on.

At that core are two parts of the algorithmic trinity: the search engine, which indexes and ranks information, and the knowledge graph, which stores entities and the relationships between them.

The next layer is assistive engine optimization. It adds the third component of the trinity: the large language model. The LLM provides reasoning, grounding, and conversation.

Instead of returning a list of links, it evaluates corroborating evidence and responds directly to the user. This layer builds on traditional SEO with entity corroboration, machine-readable proof, and the signals that help AI systems understand what your content means.

The outer layer is the agent. It introduces something the layers beneath it never had: direct access to business systems through protocols such as MCP. An agent can check inventory, compare prices, and complete transactions without visiting a page or navigating a search result. This is where AI stops recommending and starts acting.

Each layer depends on the one beneath it. The stronger your SEO foundation, the more effectively you can build everything above it. That makes SEO more central to digital marketing — and to the business itself — than it’s ever been.

If you understand how machines read the web, you’re holding the foundation that every other AI-facing initiative depends on.

The acquisition funnel hasn’t changed in 130 years, but the build direction has reversed

The funnel hasn’t fundamentally changed since marketers first drew it in the 1800s: awareness at the top, consideration in the middle, decision at the bottom, the customer moving down while the brand tries to catch them. What’s changed is where you stand to catch. It has shifted in three layers, each adding to the last rather than replacing it.

Traditional marketing stood in front of the audience in the real world, on billboards, shelves, and stages. Digital marketing did the same online, through SEO, paid search, social media, and content. AI-era marketing extends the logic again, and this is the part the industry is struggling with.

You now have to stand where you always stood and, on top of that, inside the AI engines. The engines place you in front of the audience, present you as the best solution, and increasingly make the purchase.

The modern buyer mixes all three inside the same purchase, and you have to be present in all of them. Your client still comes in from the top of the funnel down, and your offline and classic digital marketing can follow that path. But the engines run the other way. Their logic goes from the bottom up, so that’s how you build for them.

Marketers draw the funnel top-down because that’s the customer’s path, but businesses have always had reason to read it the other way. Winning the result for your own name is the cheapest move on the board and the highest-converting. That’s your warmest traffic, the people already at the door. 

I’ve been making this case since 2012, when I started on brand SERPs. Your name is the one search result you completely own, and the whole field walked past it for years.

Comparison and consideration queries are the next move up, down near the purchase, where buyers are most likely to convert. Awareness is the last thing you build. That’s where people don’t yet know what they want or what the solution might be, the furthest point from a sale. Chase it first, and you’re spitting in a vast ocean, hoping to hit a whale.

The engines make that flip unavoidable. With search engines, the user hopped between sites on the way down the funnel, so you could build top-down and get away with it. The assistive engines pull the whole funnel inside themselves. Now you build from the bottom up because that’s how the machine learns about you. 

Agents take it further still: the funnel goes dark, and the choice goes with it. Each step takes more of it out of your hands. And each rewards the same brand: the one built from the bottom up, the way it should have been all along.

Where you sit on the agentic spectrum determines how much of the business has to change

Two ideas determine how much of your business has to change. The first is the delegation boundary from Part 13. The second is the agentic spectrum.

• The delegation boundary is the micro view. It tracks how much of one buyer’s journey — searching, comparing, choosing, and buying — that person hands to a machine instead of doing themselves.
• The agentic spectrum is the macro view of the same thing. It asks what share of your clientele has gone agentic and how quickly that share is growing.

The micro view tells you how to win a buyer in the moment. The macro view tells you how much of the business has to change to keep winning them. This is the number most brands have never thought about, let alone measured. It’s the one I’d start with.

Here’s why it reorganizes the business, not just the marketing.

When the agent makes the purchase, it becomes a client you have to satisfy directly, even as it acts on behalf of the person behind it. It answers to one thing: keeping its own user. So the whole sale turns on confidence. Can the machine trust you to meet the need and keep its client happy?

That confidence has to clear a far higher bar than search or assistive engines ever set. It runs the entire length of the funnel. Earn it across the stack, and you’re the brand the agent buys from.

Preparing the business for that is what AI-era business engineering means: pricing, qualification, product data, checkout, all built so that an agent can transact with you as cleanly as a person ever did.

The agent navigates the whole funnel on its own. You have to convince it every step of the way, from awareness to the final yes. And you get almost no visibility into that journey, even less than assistive engines provide.

What you get instead is granular measurement at the negotiation and transaction stages. The agent tells you what it wants, and you either satisfy it or you don’t.

So build the business to work cleanly with agents and people alike, speaking to each from the top of the funnel to the bottom, and at the moment the deal is struck.

Translating what you do for a human into something a machine can read and act on used to feel optional. Ignoring search engines, then assistive engines, was never wise, but you could usually survive it. In the age of agents, ignore the engines, and you hand an ever-growing share of your clientele to the competition.

Your untrained salesforce is selling around the clock, for you or your competitor

Every business now runs a salesforce it never hired: Google, ChatGPT, Perplexity, Claude, Copilot, Siri, and Alexa. The real number is higher and growing every month. Every major tech platform now answers questions in-app the way an assistive engine does, across social media, video, search, and workflow tools. 

Operating systems are building AI into the platform layer. The apps you already use embed assistants that recommend tools and vendors, and no one has to open a separate engine for it to happen.

Those engines reach your prospects in three ways: explicit, implicit, and ambient. However they appear, the outcome is the same. They work around the clock. They never sleep. They talk to your prospects in rooms you’ll never see, and they decide whether to recommend you or a competitor.

The default state of that salesforce is untrained. Ask it about your category, and it answers with whoever it happens to know, and that probably isn’t you. It hedges on basic facts. It confuses you with a namesake, cites testimonials you never collected, recommends you for problems you don’t solve, and, in the worst case, names your competitor at the exact moment the user was looking for you.

The cost is real, and it never shows up on your dashboards. You can’t watch the AI research you, evaluate you, recommend you, or talk a buyer out of working with you. It all happens inside the machine. Pay attention to the taxes: invisibility, ghost, and doubt.

These engines recommend the solution they’re most confident in, and that’s often not the best one. It’s the one they understand best. The recommendation depends on what they grasp about you and how confident they are in it. 

So if your solution really is the best, train them. Educate them. Brief them. They answer to the user. Your client is their client, and they retain that client by surfacing the strongest solution they can see.

Have you made it abundantly clear that you’re the best answer to the specific problems you solve, for the ICP you solve them for?

3 taxes quietly costing you recommendations

You’ll pay a tax at every stage of the funnel for as long as this salesforce isn’t working explicitly for you.

Someone types your name directly into an engine, and instead of a clean answer, it hedges: “claims to be,” “reportedly serves,” “says on its website.” Worse, it starts offering alternatives: “You asked about this brand, but have you considered that one?” 

Search engines only do that when a competitor pays heavily to claim a spot on your brand SERP. Otherwise, you own your name. 

But AI raises the alternative on its own, out of nothing more than its own uncertainty. That’s why brand SERP and AI résumé protection are no longer optional.

That hedge and that nudge are the doubt tax. You pay it because the engine doesn’t have enough independent corroboration to commit. It sits at the Understandability layer. The cost is every prospect who came looking for you by name and left with doubt in their head, about you or a competitor.

A prospect asks the engine to weigh your category and name the best in it. It reels off three brands, and yours isn’t among them. You’re in the game. The engine knows you exist. You just don’t surface at the moment a buyer is choosing because its confidence in your credibility is too low to put you forward. That’s the ghost tax, and it lives at the credibility layer.

At the top of the funnel, you don’t even get the chance to lose. Someone asks a question you’re well qualified to answer, and you’re nowhere in the reply because the engine never flagged you as someone who belongs in that conversation. That’s the invisibility tax, a deliverability failure. You never see it because the conversation finishes without you.

You need to track this across every engine and every layer, and you shouldn’t use your own account. It’s biased toward you. It knows who you are, and no response it gives you is neutral. So you’ll need proper tracking tools and to ask the right questions. 

I covered the micro and macro views of measuring this earlier in the series, and the funnel query pathway is the best way to read it over time and across the web. What you’re measuring is how much you leak at each layer, and because the system is opaque, you read it at the macro level rather than from any single query.

Then you build from the bottom up, clearing the taxes in revenue order: 

• The doubt tax first, on your warmest traffic.
• The ghost tax next, against the competition for buyers who are almost decided.
• The invisibility tax last, at the top, where the buyer is furthest from the purchase.

That’s the funnel flip again, one more reason the engines have turned top-down upside down.

The algorithmic trinity is where your work lands, and there are only a handful at the root

You train the salesforce in three places, and you have to be present in all three for that training to hold over the long term.

• Large language models do the reasoning at the moment of the query, the intelligence layer: ChatGPT, Claude, and Gemini.
• Search engines index and rank fresh content, the information layer: Google and Bing.
• Knowledge graphs store entities and the verified relationships between them, the verification layer: Google’s Knowledge Graph, Wikidata, and Bing’s entity graph.

Those three are the algorithmic trinity.

Here’s something you might not have considered. You’re aiming at dozens of platforms and surfaces where the salesforce shows up, but there aren’t many machines at the root.

A handful of LLMs matter, and at mass-market scale, where the sales actually happen, that list narrows to ChatGPT and Gemini. There are two web indexes at that scale, Google and Bing, and two knowledge graph owners, Google and Bing again.

Everything you train reaches back to the same small set of underlying systems. So the corroboration work you do for one engine is, structurally, the work for all of them.

That’s why the effort compounds. The knowledge graph confirms the entities the LLM reasons about. The search engine surfaces the fresh content the LLM grounds on. The salesforce reaches full training when all three converge on the same answer about you. 

That convergence is where you win: independent systems reaching the same conclusion about who you are, what you do, who you serve, and why you’re credible. Give them that picture in detail, and they’ll hold it with real confidence.

At that point, the trinity surfaces you. It converts prospects at the bottom of the funnel, recommends you over the competition in the middle, and advocates for you at the top across search engines, assistive engines, and agents. 

The results vary a little from one to the next because each mixes the technologies in its own way, but they land in your favor.

Google owns all three and remains the dominant force across search and assistive engines, so it remains your main target.

I’m not suggesting you ignore smaller players such as Claude and DuckDuckGo. They matter to the audiences that use them. But for most brands, users, and SEOs, the major public engines are the key to commercial success.

A tight digital footprint, cleaned up and optimized on your site and off it, already covers them because it feeds the Trinity: at mass-market scale, six technologies — Gemini and ChatGPT, Google’s and Bing’s knowledge graphs, and Google’s and Bing’s search indexes.

Happily, that strategy covers the smaller players, too.

Third-party proof is what the salesforce actually believes, and without it, nothing stands

Knowing where your work is ingested is only half of it. The other half is knowing which evidence the salesforce believes because not all evidence carries the same weight, and the gap between the weakest and strongest is the differentiator.

The weakest evidence is what you publish yourself, on your own properties, in your own voice: your homepage, your about page, your product copy. Call it first-party. It’s a claim, the baseline, and it proves nothing on its own because you wrote it and published it, and the engines know that. 

Surface a client outcome, a case study, or a real customer review on your own off-site channel, and you’ve moved up a step to second-party, where the substance is no longer entirely your assertion, even though your hand is still on the publish button.

Then there’s the evidence you had no hand in at all. Clients and partners describing their own experiences in their own words, an independent journalist’s article, an analyst’s report, coverage gatekept entirely outside your reach. That’s third-party, and it’s the heaviest evidence the salesforce can read precisely because you couldn’t touch it.

It’s also the category most brands have nothing in because filling it takes real-world activity rather than publishing. First-party claims, second-party corroborates, third-party proves, and without proof, nothing stands.

3 levels of effort lead to completely different outcomes 

Barely trained, partially trained, and fully trained

Most brands sit at the bottom without ever choosing to. The minimum-effort brand keeps a website, runs some content marketing, responds to the occasional mention, and otherwise lets the ecosystem do what it does. It shows up in machine-readable form but never shapes that form.

Minimum effort is what the industry treats as normal, so most companies land here and never recognize it as a decision. The salesforce is barely trained.

A step up is when a brand starts noticing specific problems and fixing them: an incorrect fact in an AI Overview, a competitor outranking it for a query, a gap in the structured data. The fixes work, and this brand is genuinely better positioned. 

But the work is triggered by symptoms, not strategy. It patches what breaks loudly and never builds the discipline that would prevent the next break. The salesforce is partially trained, but fixing problems is driving the strategy, which is clearly the wrong approach.

The systematic brand runs a full operational discipline against the pipeline, week after week: entity home maintenance, evidence harvested from service teams and codified into machine-readable proof, then distributed across the publication tiers the engines weight most heavily, with the brand SERP and AI résumé monitored continuously. 

Most companies aren’t organized in a way that makes this happen naturally. If you can pull that off — harvest, codify, and distribute your business operations — you’ll win. You’ll be the brand that has trained the salesforce to work in your favor, 24/7/365.

Start from the entity home. Organize the brand SERP and the AI résumé, and optimize the brand’s digital footprint wherever it appears. That’s understandability done, and it’s the single most important thing you can do. With the core entity locked, you can build credibility on top of it: engagement, reviews, client feedback, PR, and evidence that you’re good at what you do.

Deliverability follows naturally because the moment you focus on the brand SERP and the AI Résumé, you’ve already done part of the credibility and deliverability work for free. Then spread the same discipline across every entity you own: the products, the services, and the people who work for you. 

For each, build the right content, stand where the audience is looking, invite them down the funnel, and connect every dot back to the entity home on your own site. Walk the walk, and apply the mirror principle.

The salesforce is already working. The only question is who trained it.

In 2026 and beyond, the salesforce operates inside your supply chain as well as your sales funnel. AI sits at the gates that decide whether to include you in what it knows, whether to deploy you in an answer, and whether to reselect you after every transaction. 

Every outcome your customers experience feeds back into the system for the next prospect who’s never heard of you. That’s the convergence this entire series has pointed toward. Your salesforce is selling 24 hours a day, right now, for you or your competitor. Whether it sells for you comes down to how well you trained it.

This is why the discipline is AI-era business engineering, not AI-era marketing. It isn’t a content strategy or a tactical tweak. It’s a reorganization of how the business operates, so pricing, qualification, product presentation, sales, retention, and customer success each throw off machine-readable evidence as a byproduct of doing the job. That’s how you train the salesforce.

You’re sitting in the best seat in the room, and you might not have noticed yet

This is what I present to entrepreneurs and CEOs in my talks. Nine questions, and the quality of the answers tells them exactly where the company stands.

Tech, bottom to top

• Is our Entity Home locked down so the engines have a single source of truth about who we are?
• Is our structured data complete enough for them to verify what we claim?
• Are we discoverable across every engine so they can reach for us when topical questions come up?

Marketing, bottom to top

• What does our Brand SERP look like today, and what does the AI Résumé say when you ask the engines about us directly?
• Where is our third-party corroboration weakest, and what are we doing about it this quarter?
• Which topical territory do we own in the engines, and which do we want that we don’t yet hold?

Branding, bottom to top

• Does our brand story match what AI is currently saying about us, and where’s the gap?
• Are our client outcomes being engineered into machine-readable evidence, or are they dying in CRMs and quarterly retrospectives?
• Are we placing proof now for the categories we want to own in three years?

They all run from the bottom up, which is ironic given how marketers work the funnel from the top down. The customer is the only one moving from top to bottom, looking for a solution.

So take a step back and read the funnel from the bottom up. You and everyone you work with are building the same thing — understandability, credibility, and deliverability — just from different ends.

The business builds from the foundation up: know who you are, know who you serve, be credible, then reach the right people.

The marketer wants the maximum audience and starts at the top, with reach, then works down to who you are and why you’re trusted.

AI starts at the bottom: Who are you, and are you credible? Only then will it put you in front of the maximum number of people.

You’re the one who sees that it’s all the same. You understand that you must work from the foundation up, the way the machine does. Then you meet the customer coming the other way, down from the top, looking for a solution.

Build for the customer, but work up toward them. That’s always been the best approach, but the engines have only now made it obvious.

The business you work for has two kinds of clients now: the human and the agent, and you’re able to speak to both. You know how the machines work, and you know an agent is only ever emulating a person, reflecting back the world’s view of the brand. Pleasing the agent and pleasing the human are essentially the same thing, and you’re the person who sees that.

That’s what makes you, the SEO, impossible to sideline. You’re best positioned to tell the business and the marketers what to change to satisfy the agent without losing the human.

Whether agents are 5% of the business today or close to all of it, the agentic share will increase year over year. So step out of your corner and look at the wider world. You’re in a unique position to see business, marketing, and the machines all at once.

The audience used to be only human. Now it’s machines, too, and you’re the one who speaks to both.

This is the 19th and final piece in my AI authority series, and it’s been a long journey. My thanks to Danny Goodwin, Angel Niñofranco, and the Search Engine Land team for their immense support throughout.

When I started, the framework was a complete idea, but I hadn’t fully worked through all the details. Week by week, I worked through each of the 15 gates, and every one turned out to be more intricate, more in-depth, and more thought-provoking than I’d imagined when I set out. It’s been a huge job.

What I’ve finished is an important and genuinely useful framework for SEO, marketing, and business in the AI age, one that search professionals, marketers, and business leaders can apply to real business problems.

Series index (Parts 1-18)

• Part 1, “Rand Fishkin proved AI recommendations are inconsistent, here’s why and how to fix it,” introduced cascading confidence.
• Part 2, “AAO: Why assistive agent optimization is the next evolution of SEO,” named the discipline.
• Part 3, “The AI engine pipeline: 10 gates that decide whether you win the recommendation,” mapped the full pipeline.
• Part 4, “The five infrastructure gates behind crawl, render, and index,” walked through the infrastructure phase.
• Part 5, “5 competitive gates hidden inside ‘rank and display’,” covered the competitive phase.
• Part 6, “The entity home: The page that shapes how search, AI, and users see your brand,” mapped the raw material.
• Part 7, “The push layer returns: Why ‘publish and wait’ is half a strategy,” extended the entry model.
• Part 8, “How AI decides what your content means and why it gets you wrong,” covered annotation.
• Part 9, “Why topical authority isn’t enough for AI search,” opened the competitive phase with topical ownership.
• Part 10, “The funnel flip: Why AI forces a bottom-up acquisition strategy,” named the process.
• Part 11, “The framing gap: Why AI can’t position your brand,” exposed the gap between evidence and recommendation.
• Part 12, “The 10-gate AI search pipeline: Find where your content fails,” showed how to find and repair your weakest gates.
• Part 13, “The delegation boundary: How AI decides which brands win,” mapped how delegation moves across Search, Assistive, and Agent modes.
• Part 14, “The funnel query pathway: A framework for measuring AI visibility,” built the measurement instrument.
• Part 15, “The micro-macro shift: How to measure AI visibility now that precision is gone,” moved measurement from micro precision to macro trend.
• Part 16, “How SEO turns customer success into AI-readable proof,” put SEO inside post-sale operations.
• Part 17, “How AI forms opinions about your brand,” named the discipline that turns evidence into machine-legible signal.
• Part 18, “Why Google isn’t dying: paid and organic just collapsed onto every AI surface,” showed why the ads business is bigger in AI than it ever was on the SERP.