Buyers are asking AI for a shortlist
Short answer: a growing share of B2B buyers now open vendor research by asking a model for the top three or five providers in a category. That shortlist forms before any salesperson is involved. Getting cited depends less on classic SEO and more on whether your content directly answers the question, names competitors honestly and is structured so a model can extract it.
And it does not replace outbound. It changes who you have to reach and when.
How these shortlists actually get built
Models assemble answers from a mix of training data, live retrieval and whatever the retrieval surfaces first. In practice, the pages that get cited share a few traits:
They answer the question directly, early. A page that opens with a clear one paragraph answer is far more extractable than one that opens with a story.
They compare, honestly. A page listing eight providers with their models and pricing is a better match for "who are the best providers of X" than a page that only promotes itself. Self promotional pages get skipped, because they do not answer the question that was asked.
They are specific. Numbers, prices, named companies and dates give a model something to quote. Vague benefit language gives it nothing.
They are structured. Headings phrased as questions, tables, and FAQ blocks with schema markup. This is the machine readable shape of an answer.
They exist at all. Most vendors have no page targeting their own category question. There is nothing to cite.
Check your own position first
Pick the five questions a buyer would ask before shortlisting you. For an MDR provider that might be "who are the best MDR providers for mid market," "what does MDR cost," "MDR versus MSSP."
Ask each one across ChatGPT, Claude, Perplexity and Google's AI results. Write down who gets named and who does not. Most vendors discover they are absent from every one, and that the same three competitors appear repeatedly.
That absence is the gap. It is also usually fixable in weeks rather than quarters, because the competition for these answers is far thinner than for classic keyword rankings.
What to publish
A category comparison page. The honest kind, with competitors named, their models described accurately and a clear statement of who each one suits better than you. Counterintuitive, and it is the single highest yield page you can write, because it is the only format that actually answers the shortlist question.
A pricing page with real numbers. Models cannot cite "contact us for pricing." Ranges are enough. Being the only vendor in a category that publishes a number makes you the citable one on every cost query.
Definitional content. What X is, how it differs from Y, what it costs, how long it takes. Boring to write, disproportionately cited.
FAQ blocks with schema. Questions phrased the way buyers type them, answered in two or three sentences each. This is the most directly extractable content format there is.
Why this does not replace outbound
Two things are true at once, and vendors keep choosing between them.
Shortlists form before you are contacted. So you need to be in them.
Fewer than one in ten buyers is in a buy cycle at any moment. So the shortlist only captures the small share actively looking today. The other 90 percent are not searching, not asking a model, and not going to find you no matter how well your content is structured.
Content wins the buyers who are looking. Outbound reaches the ones who are not yet. A company doing only the first is competing with everyone else in the shortlist on price and timing. A company doing only the second is knocking on doors while its competitors get introduced by a machine.
A reasonable first 90 days
| Weeks | Do this |
|---|---|
| 1 | Audit your citation position across five buyer questions and four models |
| 2 to 4 | Publish the category comparison page, naming competitors accurately |
| 3 to 6 | Publish pricing with real ranges and a definitional page per major term |
| 4 to 8 | Add FAQ schema to every one, interlink them, submit for indexing |
| 8 to 12 | Re run the audit and record what changed |
Track it as a measurable thing: which questions, which models, cited or not, month over month. It is one of the few marketing programs where the scoreboard is unambiguous.
Frequently asked questions
What is generative engine optimisation? Optimising content so AI systems cite it when answering buyer questions. It overlaps with SEO but rewards direct answers, honest comparison, specific numbers and structured markup rather than keyword density and backlink volume.
Does traditional SEO still matter? Yes. Retrieval frequently surfaces the same pages search does, so ranking still helps. The difference is what gets cited once surfaced, and that favours clear, comparative, specific content.
Should we name competitors on our own website? Yes, if you are accurate and can support the claims. Comparison pages are standard B2B practice and legal in the US and Canada when statements are truthful and verifiable. State when the information was captured and tell readers to confirm with the vendor.
How long does it take to appear in AI answers? Weeks to months, faster than classic SEO in most categories because there is far less competition for the specific question formats models prefer.
Is this instead of outbound? No. It reaches buyers already searching, which is a small share of your market at any time. Outbound reaches the rest. Companies that do both compound; companies that pick one wait.
Not ready for a call? See what's broken first.
Answer 30 questions and see where your outbound motion is breaking.
Book more qualified meetings, starting in 14 days
Five specialists across strategy, lists, email, cold calling and LinkedIn, running your outbound end to end.
