Answer · Process
How long does it take for a firm to appear in AI answers?
The answer
A page that is readable and quotable can appear in AI answers within days of being fetched, but earning a place on the questions that matter usually takes months of consistent pages and corroborating mentions. The fast part is mechanical. Once the deciding sentences sit in plain HTML on a stable address and the crawlers are admitted, an assistant can lift them the next time it fetches. The slow part is trust. Assistants lean on repeated, consistent descriptions of a firm across its own site, directories and third-party lists, and on pages that keep answering the question when the facts change. Nobody can promise a citation, and any vendor who quotes a date is guessing. What a firm can control is whether it is quotable at all, and how quickly the server log shows the crawlers arriving.
What happens in the first few weeks?
The pages become readable, the crawlers are let in, and the server log starts to show them arriving.
The first job is to put the deciding sentences into the first HTML response: who the firm acts for, what it will not take on, what happens after contact, and how fees start. The second is to check the robots file, because a surprising number of firms block the assistant crawlers by default without knowing it. The third is to confirm that the important addresses are stable and that old ones redirect.
Once that is done, the fetches begin to appear in the server log within days to weeks, depending on how often the crawlers already visited. A fetch is not a citation. It is the necessary first step, and it is the first thing worth reporting.
What takes longer, and why?
Being the source an assistant chooses for a question that several firms could answer.
An assistant composing an answer about who handles a matter type in a region has many pages to draw on. The ones it leans on tend to be described consistently elsewhere, appear on the lists buyers already consult, and keep answering the question when the facts change. Building that consistency across the firm's own site, its directory listings, LinkedIn and the sector lists is off-site work that takes months, not days.
Pages also need time to be re-fetched after each change. The work that makes a firm readable is quick. The work that makes it the obvious choice is slow, and the honest programme sets expectations that way from the first month.
How do you know it is working before enquiries move?
By reading the server log, watching branded search, and asking new enquiries how they found the firm.
Client-side analytics cannot see a crawler and often cannot tell that a visitor arrived from an assistant. The log can: it records which crawler fetched which page, and how often. Branded searches tend to rise when a firm starts being named in answers, because people check the name before making contact. And a single question on the enquiry form, asking how the person found the firm, catches the assistant referrals that analytics label as direct traffic.
The measurement work that puts those three sources beside the dashboard is Analytics and Measurement. The report should also carry a plain note on what still cannot be seen.
What slows it down?
Content that only appears after a click, addresses that change, descriptions that disagree, and a robots file that blocks the crawlers.
Each of those is common, and each undoes the rest of the work. A rebuild that moves every address without redirects resets whatever trust the old pages had earned. A directory profile that describes the firm differently from its own site teaches the market two firms instead of one. A tab that loads its content on a click hides the best paragraph on the page.
The programme that removes those blockers and then builds the off-site corroboration is Search and AI Visibility. Where the platform itself is the blocker, the fix is a rebuild first, and we say so before selling a retainer into a wall.