We publish the pages
AI wants to cite
Measurement only tells you where the gaps are.
Fill those gaps with real pages and AI learns your brand — and cites it more often from then on.
Pages built on the evidence of what gets cited
From the answer data collected each day we derive the page shapes and content that earn citations, then publish the same way on your own domain and on outside channels.
Why does your existing site never get cited?
We audit the whole brand website and lay out the technical and structural faults blocking citations today, with fixes — so the pages you already have start getting cited.
On-site pages, published to your own domain
Content pages designed to be citable go live on the brand's existing domain. Traffic and authority accumulate there instead of being split across hosts.
- 01
Pick the candidates
Page candidates come out of recent prompt records and citation data. The questions you have an answer to, but are missing from, come first.
- 02
Research and gather evidence
Each candidate gets a source pack. Every figure and fact on the page comes from it, and nothing goes in without a source.
- 03
Generate in code
Your existing site's header, type, and colors are carried into code so the page comes out in the same design. It ships as static HTML, so it reads without waiting on a render.
- 04
Connect the domain
Attached under your existing domain by CNAME or path routing. Vercel, Webflow, Shopify, self-hosted — nothing about your setup has to change.
What is the difference between an AI search optimization platform and a rank tracker?03
A rank tracker looks at where your page sits in the search results. An AI search optimization platform looks at whether your brand is mentioned and cited inside the answer AI wrote. Because the thing being measured moved from the link to the paragraph, the two tools give different answers to the same question.
| AI search optimization platform | Rank tracker | |
|---|---|---|
| What it measures | Mentions and citations in the answer | Position in search results |
| How it collects | Real responses, per engine | SERP crawling |
| Unit of action | Page and paragraph | Keyword |
{ "@type": "FAQPage",
"mainEntity": [
{ "@type": "Question",
"name": "If we already do SEO, do we need AI search optimization?" }
],
"dateModified": "2026-08-14" }Your domain, not a subdomain
It attaches as a path under your brand domain. Whatever trust this page earns accumulates there.
Your site's design, unchanged
The existing header, type, and colors are carried straight into code. To a visitor it is a page that was always there.
The question as the title, the answer in the first paragraph
What the extraction stage pulls comes from the top of the page. The conclusion does not wait until the end.
Tables, FAQs, and sources
A table with aligned rows and question-shaped FAQs are the formats that get cited a paragraph at a time. Figures carry their source.
JSON-LD that matches the body
Nothing goes into the schema that is not on the page. A schema that disagrees with the text is noise, not signal.
Off-site channels run alongside
AI does not cite only your own site. Community, blogs, video, editorial, PR — we widen the surface that can be cited.
Approve it and we post it
Connect the accounts once and drafting, posting, and index verification all finish inside the dashboard.
We tell you where to post, and what
For channels we do not take over, you get the target, the reasoning, and a manuscript you can use as-is. Your team does the posting.
- Tistory
- Naver Blog
- YouTube
- Brunch
- Medium
- Forbes
- Amazon
- Editorial
Publishing is not the end of it
Whether what you published actually landed in the answers is measured differently per surface, and the result feeds the next batch. Every turn of the cycle adds to the brand evidence AI has to work from.
Measured by traffic and dwell
It is your domain, so everything is visible. Visits and dwell time per page, and how much of that traffic came through an AI answer, all broken out.
Measured by citations and indexing
You cannot attach analytics to a post on someone else's server. Instead, whether it got cited in an answer and indexed in search is the performance measure.
The numbers set the next batch
Which topics and formats actually landed in answers feeds straight into next month's candidate selection and page structure. What did not land gets dropped or rewritten.
- 01The structure of cited pages gets read again
- 02Formats that worked become the default for the next batch
- 03Empty intents become new candidates
What is AI saying about
your brand right now?
Get a product demo and a report on where your AI visibility stands today.