FAQs: the Collection and the FAQ Section

Last Updated: August 12, 2026 | Audience: Everyone

Why FAQs Are a Collection Now

FAQ content used to be typed by hand into accordion components on individual pages. That produced the same question answered three different ways on three pages, no way to audit whether an answer was still true, and schema markup that was valid on some pages and broken on others.

FAQ answers now live in the FAQs collection (slug faq). One question is written once and surfaced wherever it applies through filtering. There are 25 questions in the collection today.

Important: FAQs Do Not Have Their Own Pages

The FAQs template page exists at /faq but it is unpublished and marked do-not-link. FAQ answers are meant to appear inside solution pages, hub pages, and persona pages, not on standalone URLs. Never link to a /faq/{slug} path from navigation, an email, or a campaign.

Fields

Content

• Question (the Name field) — The question exactly as a customer would ask it, in their words rather than internal product language. This is the string AI answer engines match against, so "How much does CAS SciFinder cost?" beats "SciFinder commercial licensing overview."
• Answer (Rich Text) — Lead with a direct answer in the first two or three sentences, then elaborate. Put related internal links inside this field; there is deliberately no separate link field. This one rich text field is the single source for both the on-page answer and its schema.org markup, so anything you put here appears in both.

Tagging

• Solutions (MultiReference to CAS Solutions) — Which solutions this question belongs to. This is what solution pages filter on, and it is the field that matters most. One FAQ can serve several solutions without being duplicated.
• Personas (MultiReference to Personas) — Which audiences ask this. Powers "contains current persona" filtering on persona landing pages. The Personas collection is not populated yet, so leave this empty for now.
• Topics (MultiReference to Topics) — The same taxonomy that drives CAS Insights, so an FAQ can sit alongside related articles on a topic page.
• Category (Option) — Groups questions on the FAQ hub and drives filter chips. Fixed list: Product & capabilities, Access & accounts, Pricing & licensing, Data & security, Training & support, About CAS. It is an option list rather than a reference so it cannot be edited into fragments.

Display and Ops

• Sort Order (Number) — Lower numbers appear first within a filtered block. Alphabetical order is almost never right for FAQs; the broad "what is it" question should lead.
• Featured (Switch) — Include in condensed "top questions" blocks where only a handful fit.
• Last Reviewed (Date) — When a human last confirmed the answer is still true. Set this on every edit. Answers about AI behaviour, security, and pricing go stale fastest, and this field is the only thing that makes a staleness audit possible.

Writing a Good Answer

• Answer first, context second. Someone scanning the page and an AI answer engine both take the first sentences. Don't open with background.
• Use the customer's words in the question. Match how people actually search.
• Keep one question per item. If the answer covers two questions, split it into two items.
• Link internally inside the answer. Send readers to the solution page or article that goes deeper.
• Don't hedge into uselessness. "It depends" with no follow-through is worse than no FAQ.

The Legacy Accordion Component

The component named Sections - Accordion - FAQ is a single question and answer accordion item, not a whole section, despite the name. There are still 107 hand-typed instances of it across the site.

Two things to know:

• It carries schema.org Question and Answer microdata, so it must sit inside a parent element carrying itemscope and itemtype="https://schema.org/FAQPage". Without that parent the markup is invalid and Google will not use it.
• Never unlink an instance. Unlinking breaks the microdata relationship and cannot be undone.

For new work, use the CMS-driven FAQ section instead. It supplies the FAQPage parent automatically and pulls content from the collection, which means the schema is correct by construction rather than by memory.

Migrating a Page Off Hand-Typed FAQs

1. Copy each existing question and answer into a new FAQs collection item.
2. Tag it with the relevant Solutions, and set Category and Sort Order.
3. Set Last Reviewed to today, after confirming the answer is still accurate. Migration is the right moment to catch answers that went stale years ago.
4. Ask a Web Developer to replace the hand-typed accordion block with the CMS-driven FAQ section, filtered to the right solution.
5. Verify on the published page that the questions render and that Google's Rich Results Test still reports valid FAQPage markup.

Do not delete the old accordion instances yourself. Removal is a Web Developer step and happens after the CMS version is verified on the live page.

Common Mistakes

• Linking to /faq/ URLs. That template is unpublished. The link will not work.
• Leaving Solutions empty. An untagged FAQ appears nowhere. Solutions is what solution pages filter on.
• Leaving Last Reviewed empty. The field exists so the collection can be audited. An empty value reads as never verified.
• Writing the question in product language. Nobody searches for "commercial licensing overview."
• Duplicating an FAQ per solution. Use the MultiReference. One item, several solutions.