Search engines and chat assistants now answer questions directly and cite a few sources. Being one of those sources is the new version of ranking, and it's earned mostly the same way.
For a growing share of questions, the answer appears before the links do. It's assembled from a handful of sources, and only those sources get named.
There's no separate technique here. What gets quoted is content that answers a question cleanly, from a source there's a reason to trust.
A clear answer in a couple of sentences, near the top, before the detail. If the answer is buried in paragraph nine it won't be extracted.
In practiceQuestion as the heading, answer immediately underneath, detail after.
Specific figures, ranges, and dates get cited far more than vague claims. "It depends on your setup" is useless to a machine and to a person.
In practicePublish the range and what moves it, rather than asking people to enquire.
Your own data, your own testing, your own experience. Pages that restate what everyone else says have no reason to be picked over the original.
In practiceNumbers from your own jobs, cases, or customers. Nobody else has those.
Content loaded by scripts, hidden in tabs, or locked behind forms is often invisible. This is old advice and it matters more now than it did.
In practiceCovered under technical SEO. It's the same work.
Named authors with real credentials, a clear about page, and a business that's identifiably real. It affects whether a source is trusted enough to name.
In practiceAuthor bios that state actual qualifications, not "the team".
These systems weigh what other sources say about you. Being referenced across the web makes you a more likely pick.
In practiceThe same work as off-page SEO, with a second reason to do it.
You'll notice none of this is new. That's the honest position: the pages that get cited are the pages that were already good, made easier to quote. Anyone selling a distinct AI optimisation method separate from ordinary SEO is selling the same work with a newer name on the invoice.
This is a genuinely new area and it's moving quickly. Here's the line between things that are established and things being guessed at confidently.
Reasonably established
Not actually known
Both columns matter. We work from the first one and tell you when something is in the second, rather than presenting a guess as a method. If that sounds less impressive than what you've been pitched elsewhere, it's because most of what's being pitched is in the right-hand column.
Imperfectly, and worth being upfront about. There's no clean report for any of this yet.
Visits arriving from AI assistants show up in analytics as referrals. Usually small numbers, and they can be tracked.
Asking the questions your buyers ask and recording whether you appear. Crude, repeatable, and currently the most direct method.
A page holding impressions while losing clicks often means an answer is appearing above it. A useful early signal.
"How did you find us" now sometimes gets answered with a chat assistant's name. Unscientific and genuinely informative.
If someone offers you a dashboard showing your exact AI visibility score, treat it carefully. Some of these tools are useful for spotting trends. None of them can see inside the systems they're reporting on, and a precise-looking number from an unmeasurable process is worse than an honest estimate.
No, and the sources being cited in those answers are largely the pages that already rank. The work that gets you cited is the work that gets you ranked.
What has genuinely changed is that simple informational queries send fewer clicks. If your traffic depended on answering easy factual questions, that's a real problem worth planning around.
Honestly, no. It's a lens on the work rather than a separate service, which is why it's part of our SEO rather than an extra line on an invoice.
This page exists because people search for it and deserve a straight answer. If you're paying separately for AI optimisation on top of SEO, ask what's actually being done differently.
It's your call and there's a real argument each way. Blocking protects your content from being used without credit. It also removes you from answers where a competitor will appear instead.
For most businesses trying to be found, being included is worth more than the content is worth protecting. Publishers with paid content often decide the opposite, reasonably.
Not reliably, and anyone promising it is overstating what's possible. These systems draw on a large amount of web content and nobody controls what any single answer says.
What we can do is make sure the information about you across the web is accurate, consistent, and present in the places these tools draw from. That improves the odds without guaranteeing anything.
Sometimes. Usually it's repeating something outdated or incorrect that exists somewhere on the web, so the fix is finding and correcting the source rather than the assistant.
Old listings, stale directory entries, and out of date pages on your own site are the usual causes. It takes time to filter through and there's no button that forces it.
Standard structured data is worth having regardless, and we'd set it up properly. Whether it specifically improves your chances of being cited isn't established.
Be wary of anything marketed as a new AI-specific format that will get you featured. Those claims are running well ahead of the evidence.
Not a separate retainer. It's included in how we approach content and technical work. Pricing sits on the main SEO page.
Other SEO services
Thirty minutes checking what appears when your buyers ask, and whether your traffic pattern has already shifted.
Or email hello@culmen.digital