Should local businesses still use FAQ schema for AI search?

For years, local SEO teams treated FAQ schema as an easy win. Add a few questions, wrap them in FAQPage markup, test the page, and hope Google showed expandable answers in the search result.

That play is over for normal local businesses.

Google's FAQ structured data documentation now says FAQ rich results are no longer appearing in Google Search as of May 7, 2026. Google is also removing the FAQ search appearance, FAQ rich result report, and Rich Results Test support in 2026. The remaining eligibility language is narrow: well-known, authoritative government-focused or health-focused sites.

For a plumbing group, HVAC rollup, pest control franchise, restoration platform, med spa brand, or hospitality operator, the practical answer is narrower than "delete every FAQ." Visible questions and answers still help customers and AI search systems understand the page. FAQ schema just no longer buys extra Google result space.

What changed in Google Search

Google's current FAQPage documentation says two things local service brands should separate.

First, the rich result benefit changed. FAQ rich results are no longer appearing in Google Search, and Google is removing the FAQ-specific reporting and test support in 2026. That means a normal local business should avoid treating FAQ schema work as a promise of expandable FAQ rows in Google Search.

Second, FAQPage remains a Schema.org type and Google still documents how it works for eligible use cases. The markup stays valid even though the rich result is gone for most sites. It just carries much less value as a search appearance tactic.

That distinction matters. A 75-location HVAC brand may have hundreds of FAQ blocks across service pages and branch pages. The wrong reaction is a panic deletion sprint. The better reaction is an audit. Which questions help customers choose the right service? Which answers are visible? Which answers are duplicated across every location? Which markup exists only because an old plugin generated it?

The answer determines whether the work is cleanup, rewrite, consolidation, or removal.

AI search does not create a special FAQ schema requirement

Google's AI features guidance is explicit: AI Overviews and AI Mode do not require special schema.org structured data. Google says the same Search foundations apply: indexable pages, crawlable content, useful text, internal links, page experience, current Business Profile information, and structured data that matches visible text.

For local service brands, that makes FAQ schema a support layer at best. Markup will never make a thin page useful, turn a generic city page into a real local source, or stand in for service pages, location pages, Business Profile discipline, reviews, citations, photos, and clear booking paths.

Where FAQs still help is the visible answer layer. A customer asking "Do you offer emergency drain cleaning in Mesa?" needs a clear page answer. An AI answer system inspecting that page also needs the service, market, availability, and contact path to be unambiguous. If the answer is present only in JSON-LD and not readable on the page, it violates Google's structured data guidance and does not help the customer.

For the broader baseline, read How Local Businesses Can Show Up in Google AI Search. For the machine-readable layer behind this article, read What Is Schema Markup? and What Is JSON-LD?.

When local businesses should keep FAQ schema

Keep FAQPage markup when the page has a real FAQ section, each question has one accepted answer, and the exact question and answer are visible to the user. That is the clean use case Schema.org describes and Google's documentation still models.

Good examples for multi-location service brands include a central financing FAQ, a warranty FAQ, an emergency-service FAQ, a membership plan FAQ, or a service-page FAQ that answers real buyer objections. A garage door brand can explain whether same-day spring repair is available. A med spa group can explain consultation requirements. A restoration company can explain what happens after an insurance-related emergency call.

The key is that the FAQ should clarify a decision. It should not exist to repeat keyword variants. A page about water heater replacement can answer "Do you install tankless water heaters in Phoenix?" if that service and market are true. It should not include 40 near-identical questions for every nearby suburb unless the answers differ in a way a customer needs.

For franchise and rollup teams, the safest pattern is a governed FAQ inventory. Corporate owns the reusable policy answers. Local teams can request location-specific answers when the branch has a real service, credential, hours, offer, or coverage difference. Marketing and development then decide whether that answer belongs on the branch page, a service page, a central FAQ page, or no page at all.

When to remove or rewrite FAQ schema

Remove FAQPage markup when it no longer describes the page accurately.

The common offenders are easy to spot. A plugin marks up hidden questions that customers cannot find. A template repeats the same five FAQs across 300 location pages. A service page includes promotional "questions" that read like ad copy. A branch page claims a service area the team does not actually serve. The visible page changed, but the JSON-LD still contains old answers.

Google's general structured data guidelines say structured data should be a true representation of page content, should not mark up content hidden from readers, and should not be irrelevant or misleading. For FAQPage specifically, Google says all FAQ content must be visible on the source page and that repetitive FAQ content across a site should be marked up on only one instance.

For AI search, stale FAQ markup creates a source-quality problem. If the visible page, Business Profile, location page, and structured data disagree, the brand has not clarified the entity. It has added another conflicting source.

The fix is usually editorial before technical. Rewrite the answer in plain language. Put it on the page where a customer needs it. Remove repeated template questions. Then decide whether FAQPage markup still matches the final section.

What a better answer section looks like

Local service pages should answer the questions that change the buyer's next action.

For an emergency plumbing page, that may be response windows, after-hours dispatch, service boundaries, minimum visit fees, and what the customer should do before the technician arrives. For a roofing page, it may be inspection timing, storm-damage documentation, financing, licensing, and whether repairs or replacement are available in that market. For a med spa page, it may be consultation requirements, provider credentials, treatment eligibility, recovery expectations, and appointment policies.

Those answers should be specific enough to reduce ambiguity without overpromising. If a franchise location cannot guarantee a same-day appointment in every city, the answer should say how dispatch works. If a service-area business hides its address, the answer should explain coverage and booking without inventing a storefront. If pricing depends on inspection, the answer should explain what factors change the estimate.

This is also where reviews and location pages connect. How to use customer reviews to improve location pages for AI search explains how repeated customer language can reveal questions customers actually ask. What should location pages include for AI search? explains how those answers fit into a branch page without turning it into a generic FAQ dump.

A 30-day cleanup plan

Use this sequence before the next template push or CMS plugin change.

• Inventory pages that currently emit FAQPage, Question, or Answer schema. Group them by page type: homepage, service page, location page, blog article, support page, and central FAQ.
• Check whether each marked-up question and answer is visible on the page and still accurate. Remove schema that does not match the page before rewriting anything else.
• Consolidate repeated corporate FAQs onto the strongest central page or relevant service page. Do not mark up the same reusable answer on every location page.
• Rewrite weak FAQs as buyer answers. Each answer should help someone choose a service, understand coverage, verify eligibility, compare locations, or contact the right team.
• Keep FAQ schema only on pages with real FAQ content, one accepted answer per question, and a clear reason for that answer to live on that page.
• Re-test the page for valid structured data and inspect it the way a customer would. The visible answer should do the work even if no rich result appears.

After that, assign ownership. Marketing should own answer quality. Operations should verify service, hours, coverage, policy, and location truth. Development should own schema generation, validation, and template behavior. A multi-location FAQ system breaks when one team treats the markup as an SEO field instead of a public source of business facts.

The operating rule

Do not optimize FAQ schema for a rich result that normal local businesses no longer get.

Optimize the answer itself. Make it visible, accurate, local when needed, and consistent with the page, Business Profile, reviews, service coverage, and structured data. Then use FAQPage markup as a truthful label for that content when the page genuinely fits the type.

That is less flashy than the old FAQ-rich-result play, and more durable. AI search systems can inspect, summarize, and cite clear page answers only when those answers exist as real content. Schema supports that clarity. The visible answer does the real work. A per-market AI visibility check shows whether those answers are actually reaching the engines that matter.

Sources

• Google Search Central: FAQPage structured data. Supports the May 7, 2026 FAQ rich result deprecation, current eligibility limits, visible-content requirement, and repetitive FAQ guidance.
• Google Search Central: AI features and your website. Supports the article's guidance that AI Overviews and AI Mode use normal Search foundations and do not require special schema.org markup.
• Google Search Central: introduction to structured data. Supports the explanation that structured data gives Google explicit clues about page meaning.
• Google Search Central: general structured data guidelines. Supports the requirement that structured data represent visible, accurate, relevant page content and that rich results are not guaranteed.
• Schema.org: FAQPage. Supports the definition of FAQPage as a web page presenting one or more frequently asked questions.
• Schema.org: LocalBusiness. Supports the contrast between FAQPage and local business entity markup for branches, addresses, opening hours, area served, and related business facts.

Amadeus Peterson is the CTO & Co-Founder of Cheers, the local search platform for multi-location service businesses.