FAQ Schema Done Right: AEO Wins for Singapore SMEs

First published: 6 June 2026 · Last updated: 6 June 2026

1

Google AI Overviews

FAQ-marked-up pages are 3.2x more likely to be cited as sources inside the AI Overview panel.

2

ChatGPT, Perplexity, Claude citations

AI engines preferentially extract from machine-readable Q&A pairs. FAQ schema is one of the cleanest extraction signals.

3

Voice assistant answers

Google Assistant, Siri, Alexa preferentially read FAQ-marked answers when responding to spoken queries.

4

People Also Ask boxes

FAQ-marked questions still appear in PAA expansions even after the dedicated FAQ rich result was reduced.

There is a confusion floating around Singapore SEO conversations in 2026 that goes something like this: "Google deprecated FAQ rich results in 2023, so FAQ schema is dead." Not quite. Google reduced the visible rich-result format in classical SERPs (the dropdown blue questions you used to see under listings). Google did not deprecate the schema itself, and the ranking and citation benefits have actually grown, not shrunk, with the rise of AI Overviews and answer engines. The 2026 reality: FAQ schema is more valuable than ever, just for different reasons than 2020. In classical SERPs, it became one of the strongest signals for People Also Ask inclusion. In AI Overviews, it became one of the strongest citation signals. In ChatGPT, Perplexity and Claude, structured Q&A is one of the cleanest extraction formats. For Singapore SMEs serious about [AEO and GEO](/services/geo-aeo/), FAQ schema is the highest-ROI structured data project on the board. This piece is the working implementation guide. JSON-LD code samples, Google policy clarifications, where to put it, what mistakes to avoid, and how to verify it is doing what you expect.

What FAQ Schema Actually Is

FAQ schema is a Schema.org structured data type that explicitly marks question-and-answer pairs on a page. The format Google recommends is JSON-LD, a script block embedded in the page head or body that describes the page's content in a machine-readable way without changing the visible HTML. When you implement FAQ schema correctly, search engines and AI engines no longer need to infer which text on your page is a question and which is an answer. They read it directly from the schema. This eliminates extraction ambiguity, raises the engine's confidence that your content directly answers user queries, and increases the probability that you are chosen as a cited source. The schema lives invisibly in the page source. Users do not see it. Crawlers and AI engines see it as the cleanest possible signal of "here is a question, here is the answer".

A Working FAQ Schema JSON-LD Example

Here is a real, validated FAQPage schema block for an SG service business. Drop this into the head of any page that has an FAQ section, customise the questions and answers, and you have working FAQ schema. ```html ``` That is the entire structure. `@type: FAQPage` declares the schema type. `mainEntity` is an array of questions. Each question has a `name` (the question text) and an `acceptedAnswer` containing the answer text. You can include as many questions as you have on the page. The questions in the schema must match questions visible on the page; do not add questions to the schema that are not in the visible content. That is a Google policy violation and will disqualify the schema.

Where to Put FAQ Schema on Your Site

A practical placement guide for SG SMEs. **Service pages.** Every service page should have a 4 to 8 question FAQ section answering the most common pre-purchase questions. Mark up with FAQ schema. This is high-leverage because service pages already have commercial intent traffic and AI Overview pull on commercial-investigation queries. **Pricing pages.** Pricing FAQs answer the questions prospects have right before they convert. Mark up with FAQ schema. The pricing FAQ block is one of the highest converting elements on most SG SME sites and the schema unlock for AI engines is a free bonus. **Blog articles.** Every blog article (we recommend 4 to 6 question FAQ block per article, which is what this article does). The FAQ block doubles as a featured snippet target and a citation extraction surface for AI engines. **Product pages.** Every product or product category page should have product-specific FAQs (sizing, materials, returns, shipping, care). Mark up with FAQ schema or Product schema with embedded questions. **Homepage.** A 5 to 7 question FAQ block on the homepage covering business-level questions ("how do you charge", "what industries do you serve", "where are you based", "how do I get started"). Mark up with FAQ schema. For SG SMEs running [SEO services](/services/seo/), the rule we apply is: every page that should rank for a commercial keyword has an FAQ section with FAQ schema. No exceptions.

How to Verify FAQ Schema Is Working

Three checks before you ship and weekly thereafter. **Check 1: Google Rich Results Test.** Paste the URL into Google's Rich Results Test (search.google.com/test/rich-results). Confirm the FAQPage schema is detected and reported as eligible. Any errors block the schema entirely; warnings are usually safe but worth fixing. **Check 2: Schema.org Validator.** Use validator.schema.org for a stricter validation. Some issues that pass Google's tool fail strict Schema.org validation; these are usually safe but indicate sloppy markup that may cause issues in less tolerant AI engine crawlers. **Check 3: GSC enhancement reports.** Google Search Console under "Enhancements > FAQ" shows pages where the schema was detected, with errors and warnings flagged. Monitor weekly. New errors usually indicate a CMS or template change broke the markup. **Check 4 (manual, for AEO): test the actual citation pull.** Query the question text on Google, ChatGPT and Perplexity. Are you being cited? If your schema is correct and your answer is high-quality but you are still not being cited, the issue is likely upstream (page authority, brand mentions, freshness signals) and our [framework for ranking in Google AI Overviews](/blog/aeo-vs-seo-singapore/) covers the broader signal set.

What Google's Policy Change in 2023 Actually Meant

Worth clarifying because the misinterpretation is common. In August 2023, Google reduced the visibility of FAQ rich results in classical SERPs. The rich result that used to display as a dropdown of questions under the listing became limited to "well-known authoritative websites" (government sites, major health and education sources). For most SG SMEs, the visible FAQ rich result no longer appears in classical SERPs. What did NOT change:

• The FAQ schema specification itself remains valid.
• Google still crawls and indexes FAQ schema.
• FAQ schema continues to help with People Also Ask inclusion.
• AI Overviews preferentially cite FAQ-marked content (this benefit grew massively after the change).
• ChatGPT, Perplexity, Claude and other AI engines extract from FAQ schema.
• Voice assistants preferentially read FAQ-marked answers.

So the 2023 change reduced one specific visible benefit (the SERP dropdown) but did not reduce the underlying value. In 2026, with AI Overviews and answer engines mainstream, the total value of FAQ schema is higher than it was before the change.

Outcome	Without FAQ schema	With FAQ schema
Google AI Overview citation likelihood	Baseline	3.2x higher
ChatGPT / Perplexity citation rate	Lower (extraction ambiguous)	Higher (extraction explicit)
People Also Ask appearance	Possible	Materially more likely
Voice assistant pull	Variable	Strong preference
Classical SERP rich result dropdown	None	None for most sites (post-2023)
Implementation cost	$0	~8 hours dev time once

The 5-Step FAQ Schema Implementation Plan

A working sequence for an SG SME.

Step 1: question audit per page

For each priority page, list 4 to 8 real questions prospects ask. Sources: Google "People Also Ask" boxes for the page's target keyword, support ticket history, sales call recordings, the questions your team answers in WhatsApp daily. Real questions, not "what is [keyword]" placeholders.

Step 2: write direct, complete answers

Each answer should be a complete, self-contained 30 to 80 word response. Direct first sentence, supporting detail second. No "see our pricing page for details" deflections. The answer needs to answer the question fully or AI engines will not extract it.

Step 3: add the visible FAQ section to the page

Render the questions as H3 headings (or use a styled accordion). Render the answers as paragraphs underneath. The visible content must match the schema content exactly.

Step 4: add the JSON-LD schema block

Drop the JSON-LD into the page head or body. Most modern CMS platforms (WordPress, Webflow, Shopify, Next.js) have a way to inject JSON-LD per page; use that rather than hardcoding.

Step 5: validate, deploy, monitor

Run Google's Rich Results Test on the staging URL. Fix errors. Deploy. Add the URL to GSC's URL Inspection tool to request reindex. Monitor GSC enhancement reports weekly for the first month, monthly thereafter.

For SG SMEs with 50+ pages to mark up, the work is repetitive but mechanical. We typically deliver a site-wide FAQ schema rollout (audit, content writing, schema, validation) in 3 to 4 weeks.

1

Audit real questions per page

People Also Ask, support tickets, sales calls, WhatsApp questions. Real questions only, no placeholders.

2

Write direct, complete answers

30 to 80 words each. Direct first sentence, supporting detail second. No deflections.

3

Add visible FAQ section

H3 questions, paragraph answers. Must match schema content exactly. No schema-only ghost questions.

4

Inject JSON-LD schema block

Per-page injection via CMS. Avoid hardcoding. Test in staging first.

5

Validate, deploy, monitor

Rich Results Test, GSC reindex request, weekly enhancement report monitoring.

Common FAQ Schema Mistakes Singapore Businesses Make

Five we see in nearly every audit.

Mistake 1: schema content does not match visible content. Adding questions to the schema that are not on the page violates Google's policy. The schema gets ignored or the page gets penalised. Always: visible content first, schema mirroring second.

Mistake 2: thin or evasive answers. "Contact us for more information" is not an answer. AI engines need a complete answer to extract. Vague answers do not get cited. Write like the user reading it has no patience.

Mistake 3: same FAQ schema duplicated across pages. The questions you mark up should be specific to the page they live on. The "how does pricing work" FAQ on the pricing page is different from the "how does pricing work" FAQ on a service page. Duplicate schema across the site looks like spam to crawlers.

Mistake 4: stuffing 20+ questions per page. Quality over quantity. 4 to 8 high-quality, direct questions and answers per page. 20 padded questions dilutes the signal and looks manipulative.

Mistake 5: never updating. FAQ content goes stale (pricing changes, services evolve, support ticket patterns shift). The schema needs to be refreshed in lockstep with the visible content. Quarterly review minimum.

Frequently Asked Questions

What is FAQ schema?

FAQ schema is a Schema.org structured data type, implemented in JSON-LD, that explicitly marks question-and-answer pairs on a web page. It tells search engines and AI engines exactly which text is a question and which is the answer, eliminating extraction ambiguity. Pages with FAQ schema are 3.2x more likely to be cited in Google AI Overviews and appear far more often in ChatGPT, Perplexity and voice assistant responses.

Is FAQ schema still useful in 2026?

Yes, more than ever. Google reduced the visible FAQ rich result in classical SERPs in August 2023, but the schema itself remains valid and the AEO and AI search benefits have grown materially with the rise of AI Overviews and answer engines. The total value of FAQ schema in 2026 is higher than it was before the 2023 change because AI engines preferentially cite structured Q&A content.

How do I add FAQ schema to my website?

Three steps: write a 4 to 8 question FAQ section visible on the page with H3 questions and paragraph answers; create a JSON-LD script block with FAQPage schema mirroring the visible content exactly; inject the script in the page head or body, ideally via your CMS rather than hardcoded. Validate using Google's Rich Results Test before deploy. The schema content must match the visible content exactly or Google will ignore it.

Does FAQ schema help with AI Overviews?

Yes, significantly. Pages with FAQ schema are 3.2x more likely to be cited inside Google AI Overviews compared to pages without. The reason is mechanical: AI engines extract from machine-readable Q&A pairs more reliably than from unstructured prose. Implementing FAQ schema is one of the highest-ROI technical projects available for AEO and Generative Engine Optimisation in 2026.

Did Google deprecate FAQ schema in 2023?

No. Google reduced the visibility of FAQ rich results in classical SERPs in August 2023, limiting the visible dropdown to "well-known authoritative websites". The FAQ schema specification itself was not deprecated and is still crawled, indexed and used by Google. The 2023 change reduced one specific visible benefit but did not affect the AI Overview citation, People Also Ask, voice assistant or external AI engine extraction benefits, all of which have grown since.

How many FAQs should I include in FAQ schema?

Aim for 4 to 8 questions per page. This is enough to provide meaningful Q&A coverage without diluting signal. More than 12 questions on a single page typically looks padded to crawlers and reduces extraction quality. The questions should be real ones your prospects ask, sourced from People Also Ask, support tickets, sales call recordings and direct customer questions, not invented to hit a count.

Related reading

• AEO vs SEO: how the optimisation rules differ in Singapore
• People Also Ask optimization for Singapore businesses
• Voice search optimisation in Singapore: an AEO approach