A useful help-center article can answer a customer's question directly and provide evidence an AI search service may cite. Start with accurate, accessible content: explain the answer, its conditions and the next step in plain text. This guide separates practical publishing checks from claims about AI visibility that you need to test on your own site.
What changed in 2026, precisely
Two things worth knowing before you rewrite anything. First, Google removed its FAQ rich result documentation in June 2026 and the feature no longer appears in search results; in its final months it was already restricted to well-known government and health sites. If your content plan was built around winning FAQ rich snippets, that plan expired. Second, the llms.txt proposal reached a version 2 in August 2026. It is a community proposal rather than an official standard — worth adding because it is cheap, not because it is decisive.
Make the answer useful when quoted on its own
A passage is easier for a reader to check when it names its subject, explains the answer and includes any important exception. Treat that as an editorial test: if someone shares only this paragraph, does it still mean what your business intends? Clear structure helps people find and interpret information; the choice of which sources an assistant cites still varies by query and service.
| Pattern | Quotable version | Why the difference matters |
|---|---|---|
| Opening line | “The return window is 30 days from delivery for unopened items.” instead of “Returns are something we take seriously at Acme.” | The first passage answers the query; the second requires the model to keep reading and risk assembling the answer from elsewhere. |
| Specifics | “Supported on plans from €49/month upward” instead of a checkmark in a feature grid. | Icons and grids carry meaning visually and lose it in extracted text. State the limit in a sentence somewhere on the page. |
| Scope | “This applies to EU orders only; US returns run 14 days.” instead of leaving the exception on a separate page. | An answer quoted without its caveat becomes wrong. Put the qualifier in the same paragraph as the claim. |
| Pronouns | “Acme's return window” instead of “our return window”. | Quoted out of context, “our” has no referent. Naming the subject makes the passage attributable. |
The structural checklist
- Give each page a clear purpose. Separate returns, exchanges and refund timing when they need different instructions, and link the pages so a reader can move between them. Avoid splitting one useful answer into several thin pages.
- Answer in the first hundred words, then expand. Keep the expansion — depth is what makes the page worth citing over a competitor's one-liner — but do not make the model earn the answer.
- Render the answer in HTML, not after a client-side fetch. Crawlers vary in how much JavaScript they execute, and a support article behind a spinner is a support article that does not exist.
- Show accurate publication and revision dates. Update a date when the content materially changes, and identify who owns the policy so readers can check conflicting information.
- Keep one canonical page per fact. When the same policy is restated across five pages that drift apart, you have taught the engines that your site disagrees with itself.
- Publish public guidance without a login when you want it discovered in public search. Keep private account details and internal procedures behind their appropriate access controls.
Check eligibility before changing your writing. Google's guidance for AI features in Search says a supporting page must be indexed and eligible to show a snippet. Check its URL in Search Console, confirm crawling is allowed by robots.txt and your hosting layer, and give it a relevant internal link from a public page. Google also states that special AI text files and special schema are unnecessary. Treat llms.txt as an optional reference for services that use it.
Structured data: what still earns its place
Markup no longer buys a Google FAQ result, but it is still how a machine tells a heading from a navigation label. Article or TechArticle with a genuine dateModified, breadcrumbs that mirror the visible hierarchy, and organisation markup that ties the content to a named entity all remain worth shipping. What has never worked is markup that disagrees with the page — describing a question in JSON-LD that no visible text answers is the fastest route to being ignored.
How to tell whether any of this worked
Use three separate views of performance. Google's AI-feature traffic is included in Search Console's Web search reporting; it is not a complete cross-assistant citation report. Your analytics can measure identifiable referral visits and their outcomes. A repeatable set of questions can show which sources assistants cite and whether their answers are accurate, but that sample is not a market-wide visibility score. Record the service, date and wording for each check, and treat changes in branded searches as context rather than proof of attribution. We wrote separately about why deflection and resolution are not the same number, which matters here because a cited article and a helpful article are not automatically the same article.
The same properties that make an article quotable by an external assistant make it answerable by your own. A support AI reading your help center hits exactly the problems above — buried answers, missing qualifiers, contradictory duplicates — which is why we treat this as one job rather than two. The mechanics of that are in how to write help-center articles an AI can answer from, and the knowledge-base side of the product is on the features page.
Frequently asked questions
Google removed its FAQ rich result documentation in June 2026 and the feature no longer appears in Google Search results, so FAQPage markup no longer earns a rich result there. It remains useful for other search engines and for assistants that parse structured data, and existing markup can be left in place, but it should no longer be treated as a Google ranking tactic.
It is a community proposal, currently at version 2 as of August 2026, for a markdown file at the root of a site that points AI agents to clean summaries of key content. It is not an official standard and no engine guarantees to use it, but it is inexpensive to publish and several documentation platforms now generate one automatically, so it is reasonable to add without expecting it to be decisive on its own.
Publish accurate answers with their conditions in accessible text, use clear headings and internal links, and keep public articles available to crawlers. These steps help make sources discoverable and understandable, but cannot guarantee that ChatGPT will cite them. Test a consistent question set and review both the cited sources and answer accuracy.
Measure identifiable referral visits and conversions in your analytics. Google's AI Overviews and AI Mode traffic is included in Search Console's Web search reporting. Separately, repeat a documented set of questions to check citations and answer accuracy. That sample measures those checks rather than total assistant visibility, and a change in branded searches does not establish which service caused it.
No, and maintaining two versions tends to produce contradictions that damage both. The structural changes that make a page quotable by an assistant, such as answering early, stating specifics plainly and keeping one canonical page per fact, are the same changes that make it useful to a human who landed from a search result and wants an answer immediately.
