Schema markup for a local business: what to add, what to leave out

Why it matters more now

Search engines have used structured data for a decade to produce rich results — the stars, the breadcrumbs, the FAQ dropdowns. The AI engines use it for something more basic: to work out what entity a page is about. A model reading a plumber’s home page sees prose; a model reading the same page’s JSON-LD sees a LocalBusiness named X at address Y offering services Z in area W, connected by identifiers to a Google Business Profile and a set of directory listings. The second is what gets resolved into a Knowledge Graph entity and recommended. The first is a guess.

The nodes a local business page should carry

• Organization or LocalBusiness. One node, one @id, identical on every page. Use the most specific subtype that is honestly you — Plumber, Electrician, RoofingContractor, HVACBusiness, GeneralContractor, ProfessionalService — with name, url, logo, address or areaServed, geo, telephone, email, openingHoursSpecification, priceRange, and sameAs to your real profiles.
• WebSite. The site as a whole, with the Organization as publisher.
• WebPage (or AboutPage, ContactPage, CollectionPage). The page, with its name, description, breadcrumb, primary image, dates, author and a speakable specification.
• BreadcrumbList. The path to the page, matching the visible breadcrumb.
• Service with an OfferCatalog on service pages: what is offered, by whom, to what area and audience.
• FAQPage wherever the page has a FAQ, with the same questions and answers that are visible.
• Person. Who is accountable for what is published — the owner or the author — referenced from the Organization and from every WebPage. E-E-A-T is a question about accountability, and this is the answer in machine-readable form.
• ImageObject for the logo and the page’s primary image, with the creator and license stated.

All of it in one JSON-LD graph in the head, with @id values that do not change between pages, so the Organization on the contact page is unmistakably the Organization on the roofing page.

The mistakes that get you ignored or penalized

• Rating yourself. An AggregateRating or Review markup on your own business, sourced from your own site, is “self-serving” under Google’s guidelines and is ignored at best. Reviews live on the platforms that collected them.
• Invented anything. A founding date you are not sure of, a rating, a price range, a number of employees. If it cannot be verified, leave it out.
• The wrong type. Marking a service-area contractor as a Store, or a general contractor as a Plumber because it looks better for a search. The type should match the Business Profile category.
• Markup that says what the page does not. A FAQPage with questions that are not on the page; a Service the page never mentions. Google treats it as spam.
• Someone else’s @id. Copying a template and shipping with another company’s identifiers, or with example.com in the URLs. It happens constantly.
• sameAs to nowhere. Links to social profiles that were never created. Only profiles that exist, and only ones that are yours.

A minimal example

A single node for a roofing contractor, shortened. A real page would carry the WebSite, WebPage, BreadcrumbList, Service and Person nodes alongside it.

{
  "@context": "https://schema.org",
  "@type": "RoofingContractor",
  "@id": "https://example-roofing.com/#organization",
  "name": "Example Roofing",
  "url": "https://example-roofing.com/",
  "email": "[email protected]",
  "areaServed": [
    { "@type": "City", "name": "Franklin", "containedInPlace": { "@type": "State", "name": "Tennessee" } },
    { "@type": "AdministrativeArea", "name": "Williamson County, Tennessee" }
  ],
  "sameAs": [ "https://www.google.com/maps?cid=…" ],
  "openingHoursSpecification": [ { "@type": "OpeningHoursSpecification",
    "dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday"], "opens": "07:00", "closes": "18:00" } ],
  "founder": { "@type": "Person", "name": "Owner Name", "jobTitle": "Owner" }
}

Every value in it is something the page and the Business Profile also say. That is the rule.

Validate it

Google’s Rich Results Test shows what Google can read from a URL and which rich results it is eligible for. The Schema Markup Validator at validator.schema.org checks the markup against the vocabulary regardless of Google. Search Console’s enhancement reports show errors across the whole site as Google crawls it — an ImageObject without a creator, a BreadcrumbList that points at nothing, a FAQPage with an empty answer. Fix what they show; they are the only feedback you get.

The free scan reads a page’s structured data and reports which of these nodes are present and which are missing. Every site we build ships with all of them from the first version — it is the third step of the AI SEO program.