Schema markup has a dirty secret: most guides list every available property without telling you which ones actually matter. The result? You either implement a bare-bones schema that misses rich result eligibility entirely, or you spend hours populating optional fields that move the needle on nothing. This guide cuts through that noise. For the five most common schema types – Article, Product, LocalBusiness, FAQPage, and Organization – you'll see exactly which properties are required to qualify for rich results, which recommended properties improve quality signals, and which optional fields you can safely skip until you have time.

Google uses schema to power rich results: star ratings in search, FAQ dropdowns, product price panels, breadcrumbs, and more. But not all schema triggers rich results equally. Google's Rich Results Test checks for required properties first. If a required field is missing, the schema is parsed but the rich result is suppressed – meaning you get none of the visibility benefit.

Recommended properties are different. Google defines them as fields that "improve the quality of your search results" but does not require them for the rich result to display. Skipping recommended properties means leaving signal strength on the table, but the page still qualifies.

Understanding what structured data is and how it differs from visible content is the foundation here. Structured data speaks to machines; recommended and required properties determine how clearly that message comes through.

The practical implication: implement required properties first, every time. Then layer in recommended properties for the schemas that matter most to your business. Optional fields are a nice-to-have, not a priority.

Article Schema: Required Vs. Recommended Properties

Article schema is the most widely deployed schema type on the web, and it is also the most inconsistently implemented. A large share of published Article schemas are technically valid but stripped of the properties Google actually uses to trigger rich results in Google News, Discover, and AI Overviews.

Required Properties for Article Schema

Google requires three properties for Article schema to be eligible for rich result features:

  1. `headline` – The article title, 110 characters or fewer
  2. `image` – At least one image URL; Google prefers images wider than 1200px
  3. `datePublished` – Publication date in ISO 8601 format (e.g., 2024-11-15T09:00:00Z)

Without all three, the Article schema will not trigger Top Stories or Discover rich results. Missing image is the most common failure point – many implementations omit it entirely or reference a thumbnail too small for Google's requirements.

Recommended Properties for Article Schema

These properties do not gate rich result eligibility, but they improve how Google understands and trusts your content:

Property What It Does
author Names the author; links to an author page with Person or Organization schema
dateModified Signals freshness; important for content updated after publication
description A short summary used in Discover cards and AI extraction
publisher Identifies the publishing organization; connects to Organization schema

The author property deserves special attention. Google's E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness) framework treats authorship as a quality signal, and author schema for content creators directly supports that signal. Skipping it weakens the trust layer around your content, particularly in health, legal, and financial topics where YMYL (Your Money or Your Life) standards apply.

The difference between Article, BlogPosting, and NewsArticle also affects which recommended fields carry the most weight – the distinctions between these three types determine which rich result features each one is eligible for.

For ecommerce businesses, product schema is where required vs. recommended really bites. Google updated its product schema requirements in 2023, and sites that haven't kept up are missing out on the price, availability, and rating displays that dominate shopping results.

Required Properties for Product Schema

Google requires at minimum:

  1. `name` – The product name

That is technically the only hard requirement for a valid Product schema. But this is misleading.

For merchant listing rich results – the prominent product panels with price and availability – Google additionally requires at least one of:

  • `offers` (with price, priceCurrency, and availability)
  • `review` (with reviewRating)
  • `aggregateRating` (with ratingValue and reviewCount)

In practice, any Product schema without offers is near-useless for an ecommerce page. Implement name and offers as your combined baseline.

Property What It Does
image Product image; Google strongly favors it for Shopping panels
description Plain-text product description
brand The product brand as a nested Brand type
sku Stock keeping unit; helps Google match your listing
gtin Global Trade Item Number; improves product entity matching
aggregateRating Star rating display in results
review Individual review; feeds rich snippet display

The gtin field is worth prioritizing even though it is technically recommended. Google uses GTINs to match your product listing to its product knowledge graph, which makes your Product schema significantly more useful for ecommerce AI citation and schema visibility.

Local businesses – clinics, law firms, agencies, restaurants, service providers – have the most to gain from getting this right. Local schema directly supports map pack visibility, business knowledge panels, and local AI citations.

Required Properties for LocalBusiness Schema

Google does not publish a rigid "required" list for LocalBusiness the way it does for some other types, but the following properties are consistently flagged as essential by its documentation and validation tooling:

  1. `name` – The business name, exactly as it appears on your Google Business Profile
  2. `address` – Nested PostalAddress with streetAddress, addressLocality, addressRegion, postalCode, and addressCountry
  3. `telephone` – Primary contact number

Without these three, the schema is structurally incomplete and unlikely to support local rich result features.

Property What It Does
url Your website URL
openingHoursSpecification Structured hours; feeds rich result hours display
geo Latitude/longitude; improves map accuracy
priceRange A rough indicator (e.g., "$$") for restaurant and service business listings
image Business photo
aggregateRating Review rating display
hasMap Link to a Google Maps location
sameAs URLs for your social profiles and directory listings

The sameAs property is the most underused recommended field for local businesses. Linking to your Google Business Profile, Yelp listing, and social profiles helps search engines and AI systems confirm entity consistency – that the business schema on your site refers to the same real-world entity as your directory listings. This entity consistency is a core local SEO schema signal and directly affects how AI tools reference your business in location-based answers.

Medical clinics and specialist practices have additional schema considerations – local SEO schema for medical practices covers the overlay between LocalBusiness and healthcare-specific types.

FAQPage schema is the most powerful schema type for AI citation. When structured correctly, individual Q&A pairs get extracted verbatim by Perplexity, ChatGPT, and Google AI Overviews. The required vs. recommended distinction here is unusually clean.

Required Properties for FAQPage Schema

Google's documentation is explicit:

  1. `mainEntity` – An array of Question items
  2. For each Question:
  • `name` – The question text
  • `acceptedAnswer` – A nested Answer type
  1. For each Answer:
  • `text` – The answer text in plain language

All three levels are required. An FAQ schema without acceptedAnswer.text is invalid and will not trigger the FAQ rich result dropdown in search.

FAQPage has fewer recommended fields than most schema types because the type is tightly defined. The properties worth adding:

Property What It Does
url on Answer Links the answer back to the source page
dateCreated Signals when the Q&A was written
author on Answer Attributes the answer to a named expert or entity

The author field on individual answers matters more than most practitioners realize. AI systems evaluate source authority at the entity level, and named authorship on FAQ answers reinforces the trust signals that determine whether AI cites your FAQ content over a competitor's.

A full walkthrough of how to implement FAQ schema correctly – including how to format answer text and nest the JSON-LD – is in the FAQ schema implementation guide.

Organization schema is the identity layer of your entire schema strategy. Every other schema type becomes more trustworthy when it connects back to a well-formed Organization entity on your homepage or about page.

Required Properties for Organization Schema

Google does not list strict required fields for Organization the way it does for Article or Product. But for the schema to be useful and for Google to recognize your brand as a distinct entity – these fields are functionally required:

  1. `name` – Your official brand or business name
  2. `url` – Your canonical homepage URL
  3. `logo` – A nested ImageObject with a url to your logo

The logo triggers the brand logo display in Knowledge Panels and is used by AI systems to identify your brand entity.

Property What It Does
contactPoint Phone, email, and contact type; supports Knowledge Panel contact info
sameAs Authoritative external profiles (LinkedIn, Crunchbase, Wikipedia, etc.)
address Physical address; important for local entity recognition
foundingDate Founding year; reinforces entity history
description A short, factual description of the organization
numberOfEmployees Signals company scale
areaServed Geographic service area

The sameAs array is the single most valuable recommended field for Organization schema. AI systems use external profile links to corroborate entity claims – a brand that links to its LinkedIn page, Crunchbase profile, and Wikipedia entry from its Organization schema sends a strong entity consistency signal. This is a core mechanism behind how AI systems recognize brand entities and choose to cite them.

A deep dive into the full property set and how to structure Organization schema is in the Organization schema markup guide.

Schema Type Required Properties Key Recommended Properties
Article headline, image, datePublished author, dateModified, description, publisher
Product name + offers (price, currency, availability) image, brand, gtin, aggregateRating
LocalBusiness name, address, telephone openingHoursSpecification, geo, sameAs, url
FAQPage mainEntity, Question.name, acceptedAnswer.text author on answers, url on answers
Organization name, url, logo sameAs, contactPoint, description, address

One pattern cuts across all five types: sameAs appears as a recommended property in LocalBusiness, Organization, and Product (via brand), and in every case it performs above its "optional" label. Entity matching depends on it.

AuthorityStack.ai's AI-powered schema generator reads your actual page content to determine which properties are genuinely present – rather than pattern-matching on keywords – which is why it populates recommended fields accurately instead of generating placeholder values you then have to clean up.

The Most Commonly Skipped Required Properties

These are the failures that show up most often in schema audits across SaaS, ecommerce, and local business sites:

Missing image in Article Schema

The image property is required for Article rich results, but a significant share of Article schemas in the wild omit it. The result is technically valid JSON-LD that earns no rich result eligibility. Always include a high-resolution image (minimum 1200px wide) with width and height specified in a nested ImageObject.

Missing offers in Product Schema

A product page with only name and description in its Product schema qualifies for almost nothing. The offers block – with price, priceCurrency, and availability – is what triggers merchant listing eligibility. Every product page needs it.

Missing address Sub-properties in LocalBusiness Schema

Writing "address": "123 Main St, Austin TX 78701" as a plain string fails validation. Google requires address to be a nested PostalAddress type with each component as a separate property. Plain-string addresses are ignored entirely.

Empty acceptedAnswer.text in FAQPage Schema

Some CMS plugins generate FAQPage schema with an empty text field on acceptedAnswer. This is worse than no FAQ schema at all – it is malformed markup that signals poor implementation quality. Every answer needs actual text content.

Catching these kinds of errors before they go live is what schema validation is for. Google's Rich Results Test and Schema.org's validator catch most of them in seconds.

How Schema Property Quality Affects AI Citations

Required properties determine whether you qualify for rich results. But recommended properties – particularly author, sameAs, description, and dateModified – affect something beyond search: they determine how confidently AI systems can extract and cite your content.

AI tools like ChatGPT, Perplexity, and Google AI Overviews evaluate structured data as part of assessing source authority. A page with complete, recommended-property-rich schema sends a clearer entity signal than a page with only the required minimum. The relationship between schema markup and AI search citations is becoming one of the more reliable levers in GEO – brands that invest in schema quality consistently see stronger citation rates.

This is part of why treating recommended properties as optional in practice is a mistake. They are optional in the sense that missing them won't break your rich result eligibility. They are not optional if AI citation share is a goal.

Where Schema Property Standards Are Heading

Schema.org specifications evolve continuously, and Google's interpretation of which properties it weighs most heavily shifts with it. A few trends worth watching:

AI-specific properties are emerging. Schema.org has introduced types like Claim, MediaReview, and SpecialAnnouncement that speak directly to AI retrieval needs. As Generative Engine Optimization (GEO) matures, expect schema types designed explicitly to help AI systems evaluate source credibility.

Authorship and entity signals are gaining weight. The author property – currently recommended across Article, FAQPage, and related types – is moving toward de facto required status for content in YMYL categories. Google's quality rater guidelines treat expertise and authorship as non-negotiable signals, and schema is the machine-readable layer that communicates both.

Product schema requirements are expanding. Google has already increased product schema requirements once (in 2023), and further expansion into sustainability, return policy, and shipping detail fields is likely as Shopping becomes more competitive. Recommended fields today have a track record of becoming tomorrow's required ones.

Multi-type schema is becoming standard. Implementing a single schema type per page is giving way to stacked implementations – a page might carry Article, Organization, Author, and FAQPage simultaneously. The schema types that most affect SEO and GEO signals overlap in ways that reward comprehensive rather than minimal implementations.

FAQ

Your page may still qualify for rich results, but it will perform below its potential. Required properties gate eligibility; recommended properties improve how clearly Google and AI systems understand your content. A Product schema with only name and offers qualifies for merchant listings but lacks the image, aggregateRating, and brand data that makes those listings competitive. In AI citation terms, incomplete schemas produce weaker entity signals, which reduces the likelihood that tools like Perplexity or ChatGPT surface your content over a competitor's.

Does Google Penalize Schema That Includes Optional Properties Incorrectly?

Google does not penalize you for including optional properties, but it can penalize for inaccurate markup. If a property value doesn't match the visible page content – for example, a review schema showing a rating that doesn't appear on the page – that constitutes misleading structured data and can trigger a manual action. The risk isn't including optional fields; the risk is populating them with inaccurate values.

Which Schema Type Has the Most Impact for a SaaS Company?

For SaaS companies, Organization, SoftwareApplication, and Article deliver the most return. Organization establishes the brand entity; SoftwareApplication supports software-specific rich results including pricing and rating displays; Article powers content citation in AI tools and Discover. SaaS-specific structured data for software products covers the property set for SoftwareApplication in detail.

How Do I Check Which Required Properties My Schema Is Missing?

Google's Rich Results Test (available at search.google.com/test/rich-results) is the authoritative tool for this. It parses your page, identifies which schema types it detects, and flags any missing required or recommended properties by name. Schema.org's own validator (validator.schema.org) checks structural validity independently of Google's rich result rules. Running both gives you the full picture.

Are Required Properties the Same for All Countries?

The core required properties defined by Google's documentation apply globally, but some markets have additional expectations. In Germany, for example, price transparency requirements affect how offers should be structured for ecommerce. In the EU, structured return and shipping policy data is increasingly expected. Local regulatory context can effectively elevate recommended properties to required ones depending on where your customers are.

Does Adding More Schema Properties Improve My AI Citation Chances?

Adding properties improves AI citation chances only when those properties are accurate and relevant to your page content. More schema is not better than correct schema. The properties that most consistently improve AI citation rates are author (for content credibility), sameAs (for entity consistency), description (for extractable summaries), and dateModified (for freshness signals). Populating these well outperforms adding optional fields for the sake of completeness.

What Is the Fastest Way to Generate Complete, Property-accurate Schema for an Existing Page?

The fastest approach is an AI-powered schema generator that reads your actual page content rather than relying on manual input. Tools that require you to fill in fields manually are prone to missing properties or populating fields that don't match your content. A generator that scans the page, identifies the correct schema types, and populates only the fields actually present on the page produces more accurate output in significantly less time.

Final Verdict: What You Actually Need

The required vs. recommended distinction is real, but it is not a license to stop at required. Here is the practical rule: required properties get you in the room; recommended properties determine what happens once you are there.

Start with this minimum per type: headline + image + datePublished for Article, name + offers for Product, name + address + telephone for LocalBusiness, mainEntity + Question.name + acceptedAnswer.text for FAQPage, and name + url + logo for Organization. Then add recommended properties in this priority order: author, sameAs, dateModified, and description – because those four do the most work across the widest range of search and AI citation signals.

If you are managing schema across multiple pages or client sites, scaling schema generation automatically is worth prioritizing over manual implementation once your baseline is solid. And if you want to see how your current schema holds up against AI citation eligibility specifically, the free schema generator from AuthorityStack.ai scans any URL and produces complete JSON-LD you can paste directly into your page's head section – Generate JSON-LD Schema and skip the guesswork.