Get cited by ChatGPT, Perplexity and AI Overviews.

It depends entirely on your category. If your buyers research before they buy — software, professional services, healthcare, education — they are already asking assistants and you are already either cited or invisible. If you sell on price and proximity to walk-in customers, local SEO will return more, and we would rather point you there.

The audit runs a fixed set of prompts a real buyer in your category would type, across the major assistants, and records who gets cited and what is said about you. That baseline is the whole point: without it, every later claim of improvement is unverifiable.

A file at the root of your site that tells AI systems what they may read and how to understand your content — the emerging convention for AI crawlers, in the way robots.txt became the convention for search crawlers. It is not universally honoured yet, which is why it is paired with explicit robots rules per user-agent.

No. These systems are non-deterministic — the same prompt can return different sources on different days, and no one outside the model providers controls what gets retrieved. What is controllable is whether your content is structured, consistent and accessible enough to be retrievable at all, and that is measurably improvable.

Faster than classic SEO in most cases, because you are competing on structure rather than on domain authority. Technical work lands within weeks; changes in citation behaviour usually appear over two to four months as engines recrawl and reindex.

The same prompt set, run monthly, showing citation share across each assistant, alongside snippet capture and the traditional rankings the same work lifts. It is a trend line, not a single reading — one spot-check on one day proves nothing about a non-deterministic system.

Almost always. Clean structure, valid schema, consistent entities and fast pages are what classic SEO wants as well. The two reinforce each other, which is why running GEO in isolation from SEO rarely makes sense.

That is a real strategic choice, not an obvious one. Blocking protects content from being used as training data, and it also removes you from the answers your buyers are reading. Publishers with licensable archives often block; businesses trying to be found rarely should. The decision gets made explicitly, with the trade-off written down.

No. Most of the gain comes from structure, schema and entity consistency applied to content you already have. Rewriting is targeted at the pages that answer real buyer questions badly, not at the whole site.

Often, though not instantly and not with certainty. Wrong facts usually trace back to stale pages, inconsistent listings or third-party profiles nobody has updated in years. Correcting the sources and reinforcing the right facts with schema is what shifts it — and it takes a recrawl cycle to show.

Yes, and local questions are among the most commonly asked of assistants. The work overlaps heavily with local SEO — consistent listings, LocalBusiness schema, genuinely distinct location pages — so for most local businesses it runs as part of the SEO engagement rather than separately.