This is The Agentic Commercial Model, a newsletter about the operating model crisis hiding inside the AI gold rush.
Let me describe a deal you're losing right now, without ever knowing it existed.
A buyer at your ideal-customer company needs the exact thing you sell. They open ChatGPT, or Perplexity, or their company's private enterprise AI, and they type a version of: "What are the best platforms for [your category] for a company like mine?" The model returns four to seven named vendors, with a sentence of context on each. The buyer asks two follow-ups to narrow it. Within a few minutes they have a shortlist of two or three, an evaluation framework, and a rough sense of pricing.
Your name was not in the answer.
No form was filled. No demo was booked. No ad was served. Nothing fired in your CRM, nothing showed up in your attribution dashboard, no SDR got an alert. From your side, this deal is invisible, because the single most decisive moment in it, the construction of the shortlist, happened in a chat window you don't control, reading sources you didn't write, before any human at your company knew the buyer existed.
This is not a forecast. This is the median B2B software purchase in 2026. And it means the most expensive parts of your go-to-market machine, the parts built to win a buyer's attention and shape their consideration, are increasingly aimed at a moment that no longer involves a human looking at anything you made.
Your sales motion was designed for a buyer who shows up. The buyer no longer shows up until the decision is largely made. That is the GTM collapse, and it's happening in three stages.
Stage one: the shortlist now forms before sales exists
Start with the single most important fact in B2B commerce right now, because everything else follows from it.
The shortlist is now assembled before you know the buyer exists. Forrester's 2026 Buyers' Journey Survey, covering nearly 18,000 business buyers, found that generative AI and conversational search are now the most meaningful source of vendor research, outranking vendor websites, product experts, and sales reps, with twice as many buyers naming AI as their top source as any other. G2's March 2026 survey of over 1,000 software buyers found 51% now begin the entire purchase process inside an AI chatbot rather than a search engine, up from 29% a year earlier. The proportion of buyers using AI somewhere in the journey has reached 94%.
Now sit with what that does to the funnel. The old model was a sequence: you generate awareness, the buyer discovers you, your content nurtures them, and eventually they raise a hand and talk to sales. Every dollar of that machine assumed the buyer travels through your touchpoints on the way to a decision. That sequence has inverted. The research that used to happen on your website, your blog, your comparison pages now happens inside an AI conversation, and the shortlist comes out the other side already formed. By the time a buyer fills in a contact form, the decision to engage has, in most cases, already been made.
The stakes in that pre-contact moment are brutal, because the AI shortlist is short. A first-page Google result gave you ten slots. An AI answer typically names four to seven vendors, and in many categories a handful of brands capture the overwhelming majority of AI-generated mentions. The field didn't narrow. It went close to binary: you're in the answer, or you're invisible. And invisibility here isn't a soft disadvantage. G2 found 69% of buyers chose a different vendor than they'd originally planned based on an AI chatbot's guidance, and one in three bought from a vendor they'd never previously heard of. The shortlist is now the game, and it's being played somewhere your GTM machine cannot see, score, or enter at the moment it matters.
Stage two: your GTM is optimised for a human who never arrives
Here's where the collapse becomes an operating-model problem rather than a marketing inconvenience.
Almost every element of a modern go-to-market function is built to influence a human in the act of looking. The SEO strategy assumes a person scanning a results page. The website assumes a person arriving to be persuaded. The retargeting assumes a person who visited and can be followed. The SDR motion assumes a person who raised a hand early enough to be nurtured. The eleven-step lead-nurture sequence, the multi-stage pipeline, the carefully choreographed journey we drag buyers through, all of it assumes a human paying attention to your owned properties during the consideration phase.
That human is leaving the building. B2B vendor sites are reporting traffic declines of 10 to 40% over the past year as research migrates into AI engines, and discovery-stage search terms, the top of the funnel, are eroding fastest. The center of gravity of the buying journey has moved into a space you cannot pay your way into with an ad pixel and cannot track with your attribution stack. Your most expensive demand-generation infrastructure is increasingly performing for an empty room.
And this is the question that should now sit above every line of the GTM budget, asked continuously and without flinching: who are we doing this for? What purpose is this serving? Does it actually work? The curated nurture track, the gated whitepaper, the eleven touches to a marketing-qualified lead, these were answers to a question the buyer has stopped asking. A motion that once converted is now, in a growing share of cases, a beautifully instrumented performance for an audience that has already left.
Stage three: you're now being evaluated by a machine, and it isn't reading your website
This is the deepest stage, and the one that should reframe how you think about everything from content to brand.
When a buyer asks an AI to compare vendors, the AI does not answer from your homepage. It answers from what it has indexed as authoritative across the web, and the evidence is stark. One analysis found 88% of AI-mode citations do not appear in the organic top ten search results. An academic study found 37% of AI-cited domains do not appear in traditional search results at all. The bulk of non-paid AI citations originate from earned media, third-party publications, analyst coverage, review sites, the sources the model has learned to trust, not the ones you own and control.
Read what that means for your commercial position. The entity now performing the first and most decisive evaluation of your product is an algorithm, and it is forming its view of you from material you didn't write and can't edit. Your positioning, your homepage hero, the assets your GTM team labours over, are largely not what's being read at the moment of shortlisting. What's being read is the third-party record of you: who cites you, what analysts say, how specifically and credibly your point of view appears in places you don't own.
The machine rewards specificity and authority. Generic, SEO-shaped content, written to capture search volume, doesn't earn citation because it doesn't answer a real buyer's specific question better than anyone else. A genuine, defensible point of view, anchored in specifics, published where it carries authority, does. The discipline that now drives the top of your funnel is not campaign management. It's the slow, compounding work of building real authority in the sources a machine considers credible, a fundamentally different muscle from the one most GTM teams have spent a decade building.
The two futures: the same shift is a gift and a guillotine
Here's what almost nobody is saying, because it requires holding two opposite ideas at once: this collapse is simultaneously the best thing that could happen to your sales motion and the thing that quietly kills it. Which one you get is determined entirely by whether your commercial model is built for it. Same shift. Opposite outcomes.
The future if you're built for it. The agentic buying journey is not autonomous, a human still makes the call, but that human now arrives at your door more educated, more informed, and more precisely matched to their own requirements than any buyer in the history of B2B selling. They have evaluated you, compared you, modelled your pricing, and built their own business case before you knew they existed. Think about what that does to the sales cycle. Gone are the wasted days: the eleven steps to a qualified lead, the multi-stage crawl through a methodology designed to slowly educate someone who is now already educated. The right customer, with the right problem, arriving pre-informed, can be sold to in days, not quarters, if you have done the work to be on the shortlist and your offer is built to meet them.
Think about that properly, because it's extraordinary. Deals closing in days. A buyer who already knows what they want, what it's worth, and what they're willing to pay, arriving with the demand and the expectation already formed. In that world, your sales team stops being a lead-generation and nurture machine and becomes something closer to an OpenAI or an Anthropic sales exec: not sourcing, not generating, not dragging prospects through legacy stages, but managing genuine inbound demand and closing buyers who have already decided they want what you have. The only question left on the table is: can you meet them? Can you match the willingness to pay, the integration expectation, the outcome they've already convinced themselves you deliver? That is a phenomenal place to sell from. It is the dream this shift makes possible.
The future if you're not. Now invert every variable. You're not optimised to show up in the models. You're not present, let alone over-indexed, in the third-party sources the AI reads. Your value proposition is over-complicated, your monetisation model is a maze, your messaging was written to resonate with "the market" in the abstract rather than with a specific buyer and a specific problem, so it resonates with no one in particular. In that world, no amount of lead generation or buyer education saves you, because there are no prospects in the room to educate. Your carefully curated message is being broadcast to an empty channel. You are running a sophisticated process, flawlessly, for a buyer who never arrives, serving no one, and the dashboard won't tell you, because the journey it measures still appears to move.
The same force that lets the prepared company close in days lets the unprepared company fail without ever seeing a prospect. There is no neutral outcome here. The shift rewards clarity and authority and punishes complexity and invisibility, and it does both at a scale and speed B2B has never seen.
The fix: stop running a campaign, start running an operating model
If you've read the earlier issues, you know I'm not about to hand you an "answer engine optimisation" checklist. AEO tactics matter, but treating this as a new channel to bolt onto the existing machine is exactly the surface-level move this newsletter exists to argue against. The collapse is structural, so the response has to be structural. The principle is the one that has run through every issue: stop optimising the legacy motion at the margins, and rebuild the operating model around what the buyer is actually doing. Three shifts, and they rhyme deliberately with the rest of the series.
Move the budget to where the decision now happens. The shortlist forms in AI answers built from authoritative third-party sources, so the investment that used to capture human attention on your owned properties has to move toward earning genuine authority in the sources a machine cites: a specific, defensible point of view, published and earned in credible venues. This is slow, compounding, and hard to attribute, which is exactly why it's defensible and exactly why most competitors won't do it. The mechanism that gets a human expert cited is the mechanism that gets a vendor shortlisted.
Measure the funnel that actually exists, not the one your dashboard shows. This is the GTM cousin of Issue 04. Just as NRR measures billing while the truth lives in value realised, your funnel metrics measure the visible journey while the decisive one happens off-dashboard. You need a new top-of-funnel instrument: a standing audit of whether you appear in the AI answers your buyers actually generate, in what position, described how. Treat "are we in the answer" as a board-level commercial metric, because it now determines pipeline before pipeline exists.
Rebuild the sales motion for a buyer who arrives late, pre-decided, and fast. When the human finally shows up, they're not at the start of a journey you get to shape, they've done several hours of research for every hour they'll spend with you, and many engage vendors specifically to validate how the AI capabilities actually work, the questions a chatbot can't answer. The motion flips from educating and nurturing toward validation, proof, and speed: meeting a pre-decided buyer who can close in days, not one waiting to be introduced to the category. Strip out the wasted stages built for a buyer who no longer needs them, and simplify the value proposition and the monetisation model until they land cleanly in a single AI-generated answer, because complexity is now a disqualifier at the shortlist.
The threat nobody is naming: your GTM is shrinking your own market
Now let me escalate to the part that should genuinely frighten any revenue leader, because it goes beyond losing a deal.
When you lose a buyer at the shortlist, you are not just losing this purchase. Watch where that buyer goes. They land with a competitor whose product is agentic, and, as the moat special argued, every day that customer uses it, the system calibrates more precisely to how they work, growing more integrated, more bespoke, more impossible to leave. Their switching cost compounds daily. Within a year or two that customer isn't merely lost, they are structurally unreachable. They will never be a viable prospect again, because the gap a competitor's accumulated learning has built is one no future campaign of yours can close.
Sit with the full consequence. A go-to-market function that fails to win the shortlist isn't running at reduced efficiency. It is actively, permanently shrinking your total obtainable market, one invisible loss at a time, handing each lost buyer to a competitor who then locks them in for good. Every quarter you stay invisible, your reachable market gets smaller, and it does not grow back. Your own GTM machine, the thing built to expand your market, has quietly inverted into the thing contracting it.
That is not a marketing problem or an efficiency problem. That is an existential threat, and it is the precise one a green dashboard is least equipped to show you, because there is no lost-deal report for the buyer who never contacted you, and no line item for the market you are forfeiting in slow motion.
So the question for every CRO and GTM leader still confidently funding the old motion is no longer "how do we generate more pipeline?" It is this: when our buyers ask a machine who they should consider, are we in the answer? Because if we are not, we are not running a go-to-market function at all. We are running a market-contraction function, expensively, flawlessly, for a buyer who has already left the room and a market that is shrinking every day we fail to notice.
Next issue, the series finale: Value that can't be measured won't be paid for, outcome-based pricing in practice. We've spent five issues dismantling the models built for the old world, pricing, comp, retention, moat, and motion. The last issue is about where this all lands: a commercial model that prices the one thing that survives the agentic transition, the value the customer can actually measure.
First published in the Agentic Commercial Model newsletter on LinkedIn, June 30, 2026. Read the original on LinkedIn.