What your company knows, published where it can be quoted.
AnswerLink reads your company, reads your market, and writes the answers to the questions your buyers are asking — on a hub you own, shaped so that search engines and AI assistants can lift them whole. This is the whole mechanism, in the order it runs.
It reads your company first.
Point it at your website. Within minutes your services, cases, regions, people and numbers are records in a Company Brain, each with the page it came from. You confirm or correct them. Nothing is written from what the model assumes about your industry; everything is written from what you have confirmed.
- Website, documents, mail and a spoken description
- Every record keeps its source
- Conflicts are surfaced, not resolved for you
Then it reads your market.
Your subjects sit on a canvas, sized by what you can say about them. Under each one the product gathers the questions your buyers actually ask — measured where there is a market index to measure against, and flagged when a question is rising. You decide what is worth answering.
- Questions per theme, with volume where measured
- Rising questions marked as they move
- What assistants are already answering about you
You answer. It writes.
The workshop asks the handful of things only you know and writes them into the page while you answer. Every paragraph is built from confirmed records and says which. A named person from your company signs it, and the page says when it was written and when it was last checked.
- Written from confirmed records, with the record named
- A person behind every page
- Dated, and re-dated when checked
Published where you own it.
The hub lives on your domain — answers.yourcompany.com, or under a path of your main site. Your colours, your type, your tone. Every page carries the markup an engine reads: the question and its answer, the organisation behind it, the author, the dates. Nothing is indexed at an address that is ours.
- Canonical to your domain, always
- FAQPage or Article, Organization and Person markup
- A sitemap, answers.json and llms.txt alongside the pages
Tied to your main domain.
A hub on a subdomain starts with none of the authority your main site has earned, so the two are joined. Every hub page names your main site as the company's address and your profiles as its own. The link kit puts static links on your main site back to the hub — in the navigation, in the footer, and one line on each service page to the answer about that service. Under a path of the main domain, through a proxy, the pages inherit the authority outright.
- One Organization node on every page, pointing at your site
- Static links from your site to the hub, where engines weigh them
- Subdomain to start; a subfolder through a proxy where you can
Told to the engines.
Publishing a page tells the engines that take IndexNow within seconds. The sitemap and the structured data are what Google reads. The llms.txt and answers.json feeds are what assistant crawlers read before they decide what on a site is worth fetching. None of it is a trick; it is the page saying plainly what it is.
- IndexNow on every publish and edit
- Sitemap, robots and structured data for search
- llms.txt, llms-full.txt and answers.json for assistants
Measured apart from the rest.
The hub counts its own readers, the crawlers that fetched it, the assistants that sent someone, and the leads it produced — per page. Search Console rows are counted only when they belong to a hub page, so a property that also covers your main site never flatters the hub. What the hub earns is reported as the hub's.
Every one of these is measured, not estimated, and the report says where the number comes from. None is promised: a company that has published nothing has a zero on each, and the product shows the zero.
- Search impressions and clicks
- Search Console, hub pages only
- Average position per question
- Search Console, per page
- Visits to the hub
- First-party, crawlers counted apart
- Visitors sent by an assistant
- Referrer, per engine
- Assistant crawls
- Server logs, per page and engine
- Citations observed
- The visibility scan, per question
- Presence in AI answers
- The AI visibility reading
- Clicks through to your site
- First-party outclicks
- Leads and value items taken
- Forms on the hub
- Questions answered, per theme
- The canvas
- Publishing cadence
- Pages published per week
- Site readings: SEO, AEO, GEO
- The scan, re-run on request
And it learns back.
The hub's own agent answers visitors from the published pages, and says so when it cannot. The questions it could not answer land back on the canvas as questions to write. The more that is answered, the more precisely the next question is asked.
- Visitor questions become the next pages
- The Brain grows with every confirmed answer
See where you stand first.
A free scan of your website reads the eight things this page is about and says what each becomes once your answers are live.
How AnswerBrain works — the same knowledge, pointed inward