June 10, 2026·SEO·12 min read

The 5 schema types that actually move the needle for Houston SMBs in 2026.

Google supports 770+ schema types. For a Houston SMB, only 5 produce measurable lift in 2026 — and three of them are barely covered in the SEO content you'll find on Google. Here's the implementation order and what each is worth in clicks.

An HVAC owner in Cypress sent me his agency's quarterly SEO report in March. Twenty-two pages. Phrases like "comprehensive structured data implementation" and "advanced semantic markup deployment." His paid SEO retainer: $2,800 a month. His actual schema implementation when I ran his URL through Schema.org's validator: zero. Empty. Not a single JSON-LD block on any page on the site.

I see this every week. Houston SMBs paying agencies real money to do "SEO" and the most basic on-page lever — schema markup — is either missing entirely or implemented so generically that Google ignores it.

Schema markup is not a ranking factor. Google has said this clearly and repeatedly. What schema markup does is change what your search result looks like and, increasingly in 2026, whether AI overviews quote your page. Both of those move clicks. Both compound.

// TL;DR

Schema doesn't rank you. It changes how your already-ranked result looks — and whether AI Overviews / ChatGPT Search / Perplexity quote you.
Of 770+ Schema.org types, five matter for Houston SMBs in 2026: LocalBusiness (specific subtype), FAQPage, Service, Review/AggregateRating, and Speakable.
Specific subtypes (Dentist, AutoRepair, HVACBusiness, Restaurant) beat generic LocalBusiness measurably — Google weights vertical-specific subtypes for vertical SERP features.
Speakable is the schema almost nobody implements but is the highest-leverage one for AI-citation eligibility in 2026.
Hand-write JSON-LD. Avoid WordPress schema plugins. ~80–120 lines per page. Validates in 90 seconds.

Why schema mattered in 2018, mattered less in 2023, and matters more than ever in 2026

The schema story moves in three phases.

2018–2022: Schema produced visible rich snippets (FAQ accordions, star ratings, recipe cards, event tiles). These took up real estate in the SERP and lifted CTR 15–30% on the same position. Everyone implemented FAQ schema. Many sites abused it.

2023–2024: Google killed most FAQ rich snippets for non-authoritative sites in August 2023. HowTo rich results died too. A wave of "schema is dead" SEO blog posts followed. Half of agency-built schema implementations were quietly removed.

2025–2026: AI Overviews launched. ChatGPT Search citations. Perplexity citations. All three systems disproportionately pull from pages with clean, comprehensive structured data — because structured data is the cheapest, most reliable signal an LLM has for "what is this page actually about." Schema came roaring back, but the audience changed. The audience is no longer just Google's SERP renderer. The audience is every AI model that crawls your site.

In 2026, schema is back to being the cheapest CTR lever an SMB site has. And almost nobody in the Houston SMB market is doing it well.

The 5 schema types ranked by Houston SMB ROI

Rank	Schema type	What it does	Effort	Approx CTR lift
1	LocalBusiness (specific subtype)	Knowledge panel, map pack, hours, directions	1 hr	15–25%
2	FAQPage	AI citation eligibility, Knowledge Graph	2 hr	10–20%
3	Service	Service-area & service-list rich results	1 hr	8–15%
4	Review / AggregateRating	Stars in SERP (where eligible), AI quote trust	2 hr	10–20%
5	Speakable	AI overview & voice citation lift	30 min	5–15% (rising)

That's it. Five types. Together, well-implemented, they will move CTR on a Houston SMB site by 30–60% from baseline within 6–10 weeks. Nothing else in the Schema.org vocabulary moves the needle enough to be worth your time as an SMB owner.

More schema is not better schema. Five well-implemented types beat thirty half-implemented ones every time.

Schema #1: LocalBusiness — and why the subtype is non-negotiable

Every Houston SMB site needs LocalBusiness schema on the homepage at minimum, and ideally on every page. The mistake almost everyone makes is using the generic "@type": "LocalBusiness" when Schema.org has a specific subtype that applies.

Subtype examples relevant to the Houston SMB market:

Dentist — pediatric, general, ortho, periodontal practices
AutoRepair — independent shops, dealer service bays
AutoDealer — new and used car dealerships
HVACBusiness — heating/cooling contractors
Restaurant, BarOrPub, CafeOrCoffeeShop, FastFoodRestaurant
RealEstateAgent — agents and brokers
Plumber, Electrician, RoofingContractor
MedicalBusiness, Physician, Dermatology, Optician
HomeAndConstructionBusiness — broader contractors

Use the most specific subtype that fits. A Sugar Land pediatric dentist should use Dentist, not LocalBusiness or MedicalBusiness. A Katy HVAC contractor should use HVACBusiness. The specific subtype unlocks vertical-specific SERP features (booking buttons for dentists, menu pulls for restaurants, "Now Hiring" badges for service businesses) that the generic type does not.

The minimum viable LocalBusiness JSON-LD

{
  "@context": "https://schema.org",
  "@type": "Dentist",
  "name": "Sugar Land Pediatric Dental",
  "url": "https://example.com",
  "telephone": "+12815550101",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "1234 Highway 6",
    "addressLocality": "Sugar Land",
    "addressRegion": "TX",
    "postalCode": "77478",
    "addressCountry": "US"
  },
  "geo": {
    "@type": "GeoCoordinates",
    "latitude": 29.6196,
    "longitude": -95.6349
  },
  "openingHoursSpecification": [
    { "@type":"OpeningHoursSpecification","dayOfWeek":["Monday","Tuesday","Wednesday","Thursday","Friday"],"opens":"08:00","closes":"17:00" },
    { "@type":"OpeningHoursSpecification","dayOfWeek":"Saturday","opens":"09:00","closes":"13:00" }
  ],
  "areaServed": ["Sugar Land","Missouri City","Stafford","Richmond","Rosenberg"],
  "priceRange": "$$",
  "image": "https://example.com/og.jpg",
  "sameAs": [
    "https://www.facebook.com/example",
    "https://www.instagram.com/example",
    "https://g.page/example"
  ]
}

Notes that matter: areaServed is your single biggest geographic signal — list all the Houston-metro suburbs you serve. sameAs connects your site to your Google Business Profile, social, and review sites (this is how Google reconciles entities). Use real coordinates, not made-up ones.

Schema #2: FAQPage — dead for snippets, alive for AI

Google killed FAQ rich snippets for most sites in 2023. The instinct was to remove FAQ schema. That was the wrong move.

FAQ schema continues to be one of the highest-citation signals for Google AI Overviews, ChatGPT Search, and Perplexity. The structured Q&A pairs are the easiest data for an LLM to ingest and quote. We've watched FAQ-heavy pages get cited by Perplexity at roughly 3x the rate of un-FAQ'd pages on otherwise identical content.

If you read our piece on SEO without an agency for Houston SMBs, this is the second-highest-leverage on-page item after title tags. Every service page, every industry landing page, every blog post should have FAQ schema mirroring a visible FAQ section.

The FAQ JSON-LD pattern

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "Do you accept Texas Medicaid for pediatric dental?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Yes. We accept Texas Medicaid (DentaQuest and MCNA networks) and Texas CHIP at our Sugar Land location. Call us at (281) 555-0101 to verify your specific plan."
      }
    },
    { "...repeat for 5-7 Q&A pairs..." }
  ]
}

Critical rule: the JSON FAQ must match a visible FAQ section on the page verbatim. Google and the AI crawlers will flag mismatches and discount your schema entirely. This is why agencies that auto-generate schema from a separate database often get their schema ignored.

Schema #3: Service — the most underused

The Service type is criminally underused. It explicitly lists what you offer and where, and Google uses it for service-area expansion in the local pack.

{
  "@context": "https://schema.org",
  "@type": "Service",
  "serviceType": "HVAC repair and installation",
  "provider": {
    "@type": "HVACBusiness",
    "name": "Cypress AC & Heating Co."
  },
  "areaServed": [
    { "@type":"City","name":"Cypress" },
    { "@type":"City","name":"Katy" },
    { "@type":"City","name":"Spring" },
    { "@type":"City","name":"The Woodlands" }
  ],
  "hasOfferCatalog": {
    "@type": "OfferCatalog",
    "name": "HVAC Services",
    "itemListElement": [
      { "@type":"Offer","itemOffered":{"@type":"Service","name":"AC repair"} },
      { "@type":"Offer","itemOffered":{"@type":"Service","name":"AC installation"} },
      { "@type":"Offer","itemOffered":{"@type":"Service","name":"Heating system repair"} },
      { "@type":"Offer","itemOffered":{"@type":"Service","name":"Duct cleaning"} }
    ]
  }
}

This is the schema that gets you cited in answers like "what HVAC services does Cypress AC offer?" — a question type that's becoming the dominant search pattern with AI Overviews. Without explicit Service markup, the AI has to infer your offerings from prose, which it will do less reliably.

Schema #4: Review and AggregateRating

Two related schemas. Both useful. Both abuse-prone — Google has aggressively de-promoted self-claimed AggregateRating since 2019.

The legitimate use today: reflect your Google Business Profile rating, not a self-claimed one. The schema below tells Google "this page corresponds to the GBP listing with these reviews" — Google reconciles and may surface stars in the SERP.

{
  "@context": "https://schema.org",
  "@type": "Dentist",
  "name": "Sugar Land Pediatric Dental",
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": "4.9",
    "reviewCount": "237",
    "bestRating": "5",
    "worstRating": "1"
  }
}

Only use real numbers. Pulling stars and counts from your actual Google Business Profile is fine. Inventing them violates Google's structured data policies and can earn a manual action that tanks your entire site.

For the broader review-collection playbook, see the Houston customer testimonials playbook.

Schema #5: Speakable — the AI-citation unlock

This is the one almost nobody implements and it's the highest-leverage schema for the 2026 search environment.

Speakable schema marks specific CSS selectors or paragraph IDs on the page as "best suited for text-to-speech extraction." It was originally a Google Assistant signal in 2018. In 2025–2026, the AI overview systems started using Speakable signals as a hint for which sentences to quote in AI-generated answers.

{
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "...",
  "speakable": {
    "@type": "SpeakableSpecification",
    "cssSelector": [".tldr", ".key-takeaway", "h2 + p"]
  }
}

What to mark: your TL;DR section, your "key takeaways" paragraph, your one-sentence "what we do" pitch. The first paragraph under each H2. The answers in your FAQ.

This is the implementation that produces visible AI-citation lift inside 4–6 weeks. Most Houston SMB sites have zero Speakable markup. The competitive advantage is significant precisely because almost nobody is doing it.

The implementation sequence for an existing Houston SMB site

Week 1, Day 1: Add LocalBusiness (specific subtype) to your homepage. Validate at validator.schema.org and search.google.com/test/rich-results.
Week 1, Day 2: Add FAQPage to your top 3 service pages and homepage. Match visible FAQ sections verbatim.
Week 1, Day 3: Add Service schema to each service page. List areaServed with all Houston suburbs you actually serve.
Week 2: Add AggregateRating (pulled from real GBP numbers) to the LocalBusiness block. Refresh quarterly.
Week 2: Add Speakable to TL;DR sections, key-takeaway paragraphs, and FAQ answers.
Week 6: Pull Search Console "Performance" report. CTR should be up 15–35% from baseline on the modified URLs.

Total time: roughly 8–12 hours of focused work for a small site. Total cost if you DIY: zero. Total cost if you hire it out: $1,500–$3,000 one-time. Expected payback: typically inside the first month for any SMB with steady organic traffic.

For the broader on-page audit (title tags, internal linking, page speed) that pairs with schema implementation, see why your Houston SMB website is slow — speed and schema together compound.

Common Houston SMB schema mistakes

Using generic LocalBusiness instead of a vertical subtype. Costs you 10–20% on CTR.
Self-claimed AggregateRating with invented numbers. Risks a manual action. Use real GBP numbers only.
FAQ JSON that doesn't match visible page content. Schema gets ignored or, worse, the page gets demoted.
WordPress plugins that bloat schema with 200+ lines of irrelevant data. Google ignores most of it; the page weight cost remains.
Only adding schema to the homepage. Service pages and blog posts need their own schema — that's where most clicks land.
Forgetting to validate after every edit. One missing comma breaks the whole block.

Frequently asked questions

Which schema types should a Houston SMB implement first in 2026?

In strict implementation order: LocalBusiness (or a specific subtype like Dentist, AutoRepair, Restaurant), FAQPage, Service, Review/AggregateRating, and Speakable (for voice-search and AI ingestion). The first three are table stakes. The last two move the needle in 2026 specifically because of AI-generated search results from Google's AI Overviews, ChatGPT Search, and Perplexity.

Does schema markup actually improve Google rankings?

Schema does not directly improve rankings. It improves click-through rate on results that already rank — via rich snippets, knowledge panel pulls, and AI-citation eligibility. The lift typically shows up as 15–40% higher CTR on the same SERP position. For SMBs in competitive Houston search verticals (HVAC, dental, dealerships), that CTR lift is the single highest-ROI on-page change available.

What's the difference between LocalBusiness schema and Dentist or AutoRepair schema?

They're related types. LocalBusiness is the parent. Dentist, AutoRepair, HVACBusiness, Restaurant, and 40+ others are specific subtypes that inherit from LocalBusiness. Always use the most specific subtype that applies — Google's algorithms give specific subtypes more weight for vertical-specific SERP features (booking buttons, menu pulls, service-area highlights).

How is Speakable schema different and why does it matter in 2026?

Speakable schema marks the sections of a page that are best suited to text-to-speech extraction. Originally built for Google Assistant in 2018, it became newly relevant in 2025–2026 because AI overview systems (Google AI Overviews, ChatGPT Search citations, Perplexity) use Speakable signals to choose which sentences to quote. Marking your 'How we work' or 'What we charge' sections as Speakable measurably increases AI-citation rate.

Can FAQ schema still produce rich snippets in 2026?

Yes, with caveats. Google narrowed FAQ rich-snippet eligibility in August 2023 to only authoritative health and government sources for end-user-facing rich results — but FAQPage schema still gets ingested by AI Overviews, ChatGPT Search, and Perplexity at high rates, and still feeds Google's Knowledge Graph. SMBs should keep FAQPage schema for AI-citation reasons even if the visual rich snippet is gone.

What's the fastest way to add schema markup to an existing Houston SMB site?

Hand-write a single JSON-LD block per page, embed it in the <head>, and validate with Schema.org's validator and Google's Rich Results Test. Avoid WordPress schema plugins that bloat the output. A complete LocalBusiness + FAQPage + Service implementation is roughly 80–120 lines of JSON per page, takes 2–3 hours per template, and produces measurable SERP changes inside 6 weeks.

Sources & further reading

The Houston SMB SEO checklist (without an agency)
Why your Houston SMB website is slow — and the 1-day fix
Houston customer testimonials playbook
5 conversion killers on Houston SMB websites
Run a free 90-second audit on your site
Schema.org validator and Google Rich Results Test (external)

Dimitri Dimitrovski · Founder, WhiteBoxForge
Every site WhiteBoxForge ships has hand-written JSON-LD schema. It's the highest-leverage SEO change you can make in 2026 and almost no Houston SMB agency does it properly. The gap is the opportunity.