In the 28 days to June 6, 2026, duo.ca ranked position 5.9 for the query "what ai tools produce b2b content without sounding generic," and position 2.3 for "role of content engineer in automating refresh workflows." The site is months old. It has almost no backlinks. Nobody has heard of it.
Those are conversational, question-shaped queries, the kind a person types into an AI assistant before they type it into Google. A domain with no authority is showing up near the top of them. That is the whole argument of this piece, so let me say what it means and what it doesn't.
Most GEO advice is brand-strength advice wearing a disguise
Read ten guides on showing up in AI search and you'll find the same checklist: build topical authority, earn mentions across G2 and Reddit and Capterra, get cited on high-authority domains, grow your branded search volume. All true. All useless to a company that doesn't have any of it yet.
That advice describes the position you reach after you've already won. It assumes the four-platform footprint, the existing search volume, the brand a model already recognizes. If you're a B2B company doing one to ten million in revenue with eleven to two hundred people, you have none of that. You have a small site and a real point of view.
So the honest question isn't "how do I do everything HubSpot does." It's what can a small domain actually control today, before any of the authority signals exist.
The answer is narrower than the guides admit, and it's the one thing the big domains underuse.
The buyers are already there, which is the only reason this matters
Before the how, the why. If your buyers weren't researching in chatbots, none of this would be worth your afternoon.
They are. G2's March 2026 survey of 1,076 B2B software buyers found that 51% now begin their research with an AI chatbot more often than with Google, up from 29% in April 2025 (G2, The Answer Economy). The number that should move you isn't that one, though. It's the related finding that one-third of buyers reported choosing a vendor they had never heard of before, surfaced by the chatbot.
That is the small company's opening. Traditional search rewards the brand the buyer already knows to look for. An AI assistant surfacing an answer to a specific question can hand you a buyer who didn't know you existed thirty seconds earlier. The discovery isn't gated on prior fame.
It's worth being precise about what the model is doing here, because it changes what you optimize for. When a buyer asks a chatbot "what's the best way to handle X," the model isn't ranking a list of ten blue links and letting the buyer pick. It's composing one answer and deciding which sources to pull from to build it. The target isn't a click. The target is being one of the few passages that makes it into the answer the buyer reads. That's a smaller, sharper target, and it rewards a different kind of page than the one that wins a crowded search results page.
The one move a small domain can make: structural citability
A big domain gets cited because the model already trusts it. A small domain gets cited because a specific page is the cleanest available answer to a specific question. That second path is the one you control, and it's structural, not reputational.
When a model assembles an answer, it's looking for a passage it can lift and stand behind. It prefers the page that states the answer plainly over the page that makes it dig. Authority breaks ties. But on a narrow, specific question where no authoritative page has bothered to answer well, structure wins outright.
That's the gap. The big domains write broad, hedged, SEO-padded pages because they're optimizing for head terms and ad-adjacent traffic. They rarely write the tight, claim-dense answer to a narrow question, because that question doesn't have the search volume to justify their time. You don't have that constraint. The narrow question is exactly where you can be the best answer on the open web.
The three structural moves that don't need a big brand
None of these require authority, budget, or a backlink. They require writing differently.
- Answer the question in the first two sentences. Put a self-contained answer at the top of the page, before any context. A model scanning for a quotable passage finds it immediately instead of giving up two paragraphs in. This is why every section on this site opens with the answer, not the setup.
- Make every claim quotable, which means making it specific. "Many buyers use AI" is unliftable; a model won't cite it because it can't stand behind a vague claim. "51% of B2B buyers begin research with an AI chatbot, per G2's March 2026 survey of 1,076 buyers" is liftable, because it carries its own proof. Real names, real numbers, real dates. One per few sentences.
- Structure the page as question and answer. Write headings as the actual question a buyer asks ("How do small B2B companies show up in ChatGPT?"), then answer it directly underneath. Add FAQ markup. You're not gaming the model, you're matching the shape of what it's looking for so it doesn't have to reconstruct your meaning.
These are the same habits that make a page good for a human in a hurry. That's not a coincidence. AI assistants are optimizing to answer a person fast, so the page that serves the impatient reader is the page the model wants to quote.
One more thing the guides skip: you don't need a hundred of these pages. You need a handful that are genuinely the best answer to a question your buyer actually asks. A small site spreading itself across fifty thin pages loses to the same site putting that effort into five claim-dense ones. Pick the questions where you have a real, specific, experience-backed answer, and where the existing answers on the open web are vague. That intersection is where a small domain wins, and it's usually smaller than you'd guess and easier to dominate than you'd fear.
The honest caveat, because the data deserves one
Two things I won't overclaim.
First, the positions above are Google Search Console data, not ChatGPT citation data. They measure where duo.ca lands in Google's results on AI-shaped, conversational queries. That's a strong proxy (the same structural traits that earn a high Google position on a question-shaped query earn citations in AI answers, and the page-level signals overlap heavily) but it is a proxy. Clean, direct citation tracking across ChatGPT, Perplexity, and the rest is still an immature instrument for everyone.
Second, structural citability gets you cited. It doesn't get you the search volume of an enterprise domain. You will win narrow, specific questions, not head terms. For a small B2B company, narrow and specific is where the buyer with real intent actually is, so that's the right trade. But it's a trade, not a free lunch.
If you want the underlying mechanics, how to structure content AI will quote goes deeper on the page-level moves, and GEO vs AEO vs SEO for B2B sorts out the acronyms that the rest of the internet keeps blurring together.
The Upshot
You can't outspend the big domains for AI citations. You can out-structure them.
The enterprise playbook tells you to go win the authority signals. That's the right long game, and it's the wrong place to start when you don't have any of them yet. The move you can make this week is to write pages that answer one specific question, lead with the answer, and back every claim with a real specific a model can lift without hedging.
duo.ca is a months-old site with no brand and no backlinks, ranking position 2 to 6 on the exact conversational queries an AI assistant is built to answer. The reason isn't who it is. The reason is how its pages are built, which is the one thing a small company gets to control, and it happens to be the thing that matters most.
The full system this sits inside is in the AI content systems for B2B guide.
Common questions.
How do small B2B companies show up in ChatGPT?
By making individual pages easy for a model to lift and cite. Answer one specific question per page in the first two sentences, back every claim with a real number or named example, and structure the page so a model can find the answer without parsing your whole site. Small domains can't fake brand authority, but they can out-structure bigger sites on narrow, claim-dense pages.
Why doesn't standard GEO advice work for small companies?
Most generative engine optimization advice quietly assumes you already have search volume, a presence across multiple review platforms, and a brand the model already recognizes. A small B2B company has none of that yet. The advice that survives the brand-strength filter is structural: write answers a model can extract, not assertions it has to take on faith.
Are B2B buyers actually using ChatGPT to research vendors?
Yes. G2's March 2026 survey of 1,076 B2B software buyers found 51% now begin their research with an AI chatbot more often than with Google, up from 29% in April 2025. One-third reported buying from a vendor they had never heard of before the chatbot surfaced it. That last number is why a small company should care.
What makes a page citable by an AI model?
A self-contained answer near the top, specifics a model can quote without hedging (real names, real numbers, real dates), and clean structure (a clear question as the heading, a direct answer under it). Models cite the page that states the answer plainly, not the page that buries it under three paragraphs of throat-clearing.
