CoreWeb
4 engines
Google AI, ChatGPT, Gemini, Perplexity
New metric
share of answers
Zero-click
citation without a visit

The technical layer is LLM Optimization.

AI SEO

Your buyers increasingly ask an assistant instead of running a search. The assistant names three companies. If you are not one of them, you were eliminated before anyone visited a website — and no analytics report will tell you it happened.

Search Engine Optimization

Being found on Google, in AI answers, on maps and in marketplaces.

The shift
From ten links to three named sources.

What changed

A traditional results page offers ten links and lets the user choose. An AI answer synthesises a response and cites perhaps three sources. The funnel narrows dramatically at the top, and the selection happens somewhere you cannot see.

This breaks the measurement model everyone relies on. You can be cited by an assistant, influence a purchase decision, and receive no session, no referrer and no conversion attribution. Traffic can fall while influence rises, or fall while influence collapses, and standard analytics cannot distinguish between the two.

So the metric has to change. We measure share of answers — across a defined set of queries that matter to your business, how often is your brand named, and by which engines. That is a number you can track and move.

Uncertainty
This discipline is roughly two years old.

An honest caveat

Anyone claiming a settled methodology here is overselling. The engines change their retrieval behaviour frequently, they disclose little about source selection, and practitioner consensus is thin.

What we do have is reasonable evidence about what helps: content that is genuinely retrievable, clearly structured at passage level, unambiguous about entities, and corroborated across sources the models already trust. Much of this overlaps with good technical SEO, which is convenient but not accidental.

We will tell you what is established, what is inferred, and what is a hypothesis we are testing. If someone offers you guaranteed AI visibility, they are describing a system nobody outside those companies fully understands.

7 areas
One body of work, not a menu.

All services

What’s included

Baseline visibility measurement

Establishing where you currently stand across the major engines for a defined query set, so there is a number to improve against.

Retrievability

Ensuring your content can actually be fetched and parsed by the crawlers these systems depend on. Content behind rendering barriers is invisible to most of them.

Passage-level structure

Organising content so a specific answer can be extracted cleanly, rather than requiring a model to infer it from a long unstructured page.

Entity clarity

Making unambiguous what your company is, what it does and how it relates to the things it discusses, through structured data and consistent external corroboration.

Corroboration across sources

These systems weight agreement between independent sources. Being described consistently across the web matters more than any single page.

Citation monitoring

Ongoing tracking of where you are named, for which queries, on which engines, and how that changes.

Crawler access policy

Deliberate decisions about which AI crawlers to permit. Blocking them protects content and guarantees invisibility, and that trade-off should be a choice rather than an oversight.

Common questions
Asked on most first calls.

Questions

Is this genuinely different from SEO?

Partly. The technical foundation overlaps heavily — crawlable, well-structured, credible content helps in both. What differs is the measurement, the emphasis on entity clarity and corroboration, and the fact that success may produce no click at all.

Can you guarantee we appear in AI answers?

No. We can measure where you stand, improve the factors evidence suggests matter, and track the change. Anyone guaranteeing placement in systems whose selection logic is undisclosed is not being straight with you.

Should we block AI crawlers to protect our content?

A real trade-off with no universal answer. Blocking prevents your content training or grounding these models, and also removes you from their answers. For most businesses selling something, visibility is worth more. For publishers monetising content directly, often not.

How do we measure ROI with no clicks?

Share of answers across your priority queries, plus brand search volume, which tends to rise when assistants name you. Neither is as clean as last-click attribution. Both beat measuring nothing.

Is this worth doing yet, or should we wait?

It depends on your category. If your buyers research before purchasing, they are already asking assistants. Waiting means your competitors establish the corroboration and citation history that these systems reward.

Next step
We look at your site before the call.

Start with a free audit

An automated crawl plus a 30-minute walkthrough of what we can see from outside your site. If what you need isn’t something we do well, we’ll tell you.