Schema Structured Data for Home Services

Four types drive the vast majority of measurable results. LocalBusiness (or its subtype: Plumber, RoofingContractor, HVACBusiness, ElectricalContractor) establishes your entity. Service links each offering to that entity. FAQPage makes your questions eligible for rich results and AI extraction. Review and AggregateRating publish your star rating on the SERP before the click. Everything else is secondary.

Not directly, in the traditional sense. Schema does not carry ranking weight as a direct signal. What it does is reduce ambiguity: Google and AI engines become more confident about what your business is and where it operates, which increases the probability of being surfaced for the right queries. The indirect effect on rankings and AI citations is well-documented and practically significant.

Microdata is embedded inside your HTML elements as attributes; JSON-LD lives in a separate script tag. Google recommends JSON-LD for all new implementations: it does not require changing your HTML, it can be updated independently from the page design and it is far easier to validate and maintain. We exclusively deploy JSON-LD.

Google provides two free tools. The Rich Results Test (search.google.com/test/rich-results) analyzes any URL or code snippet and shows errors, warnings and eligible rich result types. Google Search Console's Enhancements section shows errors found during crawling, with affected URLs. We run both during our audit and deliver a prioritized error list.

It depends on your model. If you have a physical location in each city, yes — each location gets its own LocalBusiness entity with its own address and phone. If you operate from one office but serve multiple cities (service-area business), you declare a single LocalBusiness with an areaServed property listing your coverage. Mixing both models incorrectly is one of the most common errors we find.

Both Google's AI overviews and large language models like ChatGPT do passage-level retrieval: they extract the fragment of content that best answers a query and cite or paraphrase it. FAQPage markup pre-formats your answers as discrete, labeled question-answer pairs, making them far easier to extract than prose. A well-marked FAQ is effectively a pre-packaged citation.

Yes. Schema with factual errors — wrong phone numbers, incorrect opening hours, categories that don't match your actual services — reduces machine confidence and can suppress your entity in the knowledge graph. Google's quality guidelines also flag structured data used deceptively (marking up things not visible on the page) as a manual action risk. Stale schema is nearly as damaging as no schema.

Every time something material changes: new service added, location moved, phone number updated, opening hours changed, new review period. At a minimum, we schedule a full reconciliation audit every 90 days. We also set up automated alerts in Google Search Console so errors are caught within days of appearing, not months.

Yes, provided they are genuine reviews from real customers and are also visible on the page in a form the customer can read. Google prohibits marking up reviews that exist only in the structured data and not on the page itself. We also ensure that review collection and display practices comply with CCPA requirements for US businesses — particularly around data transparency and opt-out rights.

Yes. Voice assistants and AI overviews both rely on structured, machine-readable data to construct spoken or generated answers. A business without schema markup forces the engine to infer its details from prose, which is slower, less reliable and more likely to produce errors. Structured data is effectively the API that lets AI assistants query your business facts directly.