An AEO agency that drafts the answer, then publishes it on your domain.

Answer engine optimization is the work of becoming the answer an assistant gives, rather than a link in a list. An AEO agency finds the questions buyers actually ask, checks which sites get named when an assistant answers them, and then makes an existing page the better source: the question stated plainly, a direct answer near the top, and the evidence that supports it close by. The half most vendors skip is drafting that change and checking again afterward.

Research and supported drafts are included within your site’s scope. Related findings become coordinated page improvements, and you choose what to approve. Essential refreshes the research cycle monthly; Pro refreshes it weekly. Publishing and verification follow your approval.

Foliora writes the approved change through the connected Shopify, WordPress, or Webflow integration, or opens a GitHub pull request, after a remote re-read. If you skip a publisher, you still get the exact approved diff to apply yourself or through a customer-owned agent. There is no PDF to hand to a developer: the output is a live, rendered URL or a pull request.

Shopify, WordPress, and Webflow through native integrations, or a GitHub pull request on a code-owned site. If you skip a publisher, you still get the exact approved diff. The free snapshot still reads any public site.

The literal before-and-after: the current heading and passage next to the proposed rewrite, plus the buyer question and evidence that motivated it. Schema and canonical changes stay behind the same approval. Nothing writes from a description of intent. If the remote page has been edited since the proposal was built, Foliora re-reads it before writing and flags the proposal as stale rather than overwriting whatever changed.

After publish, Foliora re-reads the rendered page to confirm the change actually took — not just that the API returned success — then re-runs the exact buyer questions that previously named a competitor against the same surfaces, on the same schedule as before. Each re-ask is dated and stored with the response and citation URLs. Approval already counted against the monthly limit before that write.

Essential is $49.99 per site per month for monthly research and improvements. Pro is $149.99 per site per month for weekly checks and improvements. Both include supported drafts, customer approval, publishing, and live verification. Annual billing saves 20%.

By asking, on a schedule, and keeping the receipts. Foliora runs a fixed panel of buyer questions against each surface and stores the engine, the prompt, the date, the full response, the citation URLs, and which competitors were named instead. That record is what lets you distinguish a real change from run-to-run variance, which is substantial — the same question can return different sources on consecutive days.

ChatGPT and Gemini. We report the surfaces we actually sample rather than listing every assistant in existence, and every observation is stored with its engine and date so you can reproduce it yourself.

For people, in a shape a machine can also use — and the distinction matters, because the failure mode of AEO is robotic fragments that answer a question and tell a reader nothing. A good answer passage states the conclusion early, then keeps the qualifications, the evidence, and the limits attached to it. Stripping the nuance to make something extractable produces a page that gets quoted once and trusted never.

The shipping work is on pages that already exist and already rank. The question is buried in a heading that does not state it, the answer is three paragraphs down, and the supporting evidence sits on a different page. Restructuring an existing page is faster and lower risk than inventing a new one, so that is where Foliora starts. New pages stay behind your approval.

No. Nobody controls model output, not us and not anyone charging you for the promise. What you get instead is a specific, checkable claim: here are the questions a competitor currently answers, here is a restructured passage drafted to answer them better, and — after publish — here is the same question re-asked so you can see the before and after yourself.