Is AI visibility just SEO with a new name?

Largely, but not entirely. The foundation is shared: authority, relevance, technical health and accurate entity data drive both. What is genuinely new is optimising at the passage level rather than the page level, and measuring presence in answers rather than position in a list. Treating it as entirely separate wastes money; treating it as identical leaves the passage-level work undone.

Do I need to rank on page one to be cited by an AI assistant?

Usually for the assistants that search the live web, because retrieval draws from the top results of whatever query the system formulates. It is not absolute, since different assistants formulate different queries and some draw on their own indexes, but a firm ranking nowhere is rarely retrieved anywhere.

If we stop SEO work, will our AI citations decay?

Over time, yes, through the same mechanism that erodes rankings. Competitors publish, links age, and retrieval reflects the current index rather than a past one. Citations also decay faster on topics where recency is weighted, which in legal means anything tied to active litigation or a changing rule.

Part of: AI Visibility

Can SEO improve a law firm's AI visibility?

By Mohammad Kashif, Chief Technology OfficerLast updated September 15, 2026

Yes, and for most firms it is the largest single lever. Assistants that cite live sources retrieve from a search index, so pages that do not rank are rarely retrieved at all. SEO does not guarantee a citation, but its absence usually guarantees no citation.

The dependency is mechanical rather than correlational. Retrieval issues a search, pulls back a candidate set of roughly the top results, then selects passages from within it. A page outside that set is not competing on quality, it is not in the room.

Where SEO carries directly across

Most of what earns rankings is the same thing that earns citations, which is why treating these as two budgets usually wastes one of them.

Authority. Referring domains decide what gets retrieved, and they are the slowest thing to build in either discipline.
Relevance and topical depth. A model picks among retrieved candidates, and the candidates are chosen on the same signals Google uses.
Technical health. Crawlability, render-blocking and page speed affect whether your content is read at all, by either kind of system.
Entity consistency. Accurate, matching business data helps local rankings and corroboration in equal measure.

Where SEO alone is not enough

Three things matter for citation that a conventional SEO programme does not usually optimise for, and they are the reason this is not simply a rebranding.

Passage independence. Ranking rewards the page; citation rewards a paragraph that still makes sense after being lifted out of it.
Answer position. A page can rank on thoroughness while burying its answer, and burying the answer is what loses the citation.
Checkable specifics. Rankings tolerate vague claims. Corroboration does not, which is why superlatives never get quoted and dates, jurisdictions and figures do.

What this means for a firm deciding where to spend

If the site has little authority, spend on authority. No amount of answer-first rewriting rescues a page that is never retrieved, and this is the most common mistake: buying content because it is purchasable while the actual constraint goes untouched.

If the site already ranks but is never quoted, the constraint has moved. At that point rewriting the pages that answer client questions, and making the firm's details agree everywhere, is the cheaper and faster work.

The practical answer for most firms is that this is one programme with two measurements, not two programmes. Run the SEO work, add answer-first structure to the pages that matter, and track rankings and citations separately so you can tell which constraint you are actually hitting.

Which constraint are you actually hitting?

Symptom	Likely constraint	Where to spend
Not ranking, not cited	Authority	Referring domains, digital PR
Ranking well, never cited	Passage structure	Answer-first rewrites of key pages
Cited by one assistant only	Coverage	Crawler access, indexing per system
Named inconsistently or wrongly	Entity data	Business Profile, directories, markup

Common questions

Is AI visibility just SEO with a new name?: Largely, but not entirely. The foundation is shared: authority, relevance, technical health and accurate entity data drive both. What is genuinely new is optimising at the passage level rather than the page level, and measuring presence in answers rather than position in a list. Treating it as entirely separate wastes money; treating it as identical leaves the passage-level work undone.
Do I need to rank on page one to be cited by an AI assistant?: Usually for the assistants that search the live web, because retrieval draws from the top results of whatever query the system formulates. It is not absolute, since different assistants formulate different queries and some draw on their own indexes, but a firm ranking nowhere is rarely retrieved anywhere.
If we stop SEO work, will our AI citations decay?: Over time, yes, through the same mechanism that erodes rankings. Competitors publish, links age, and retrieval reflects the current index rather than a past one. Citations also decay faster on topics where recency is weighted, which in legal means anything tied to active litigation or a changing rule.

Find out whether assistants name your firm

We run a fixed prompt set across the major assistants and report which sources they quoted instead of you.

Talk to us