Schema Markup in 2025: What Still Works and What Google Ignores

Which Schema Types Still Drive Measurable Rich Results

Schema markup implementation decisions in 2025 should be driven by one question: does this schema type produce a rich result that measurably affects click-through rate? Google's documentation lists dozens of supported schema types. The set that reliably produces visible SERP features is much smaller. Investing implementation time in schema that Google processes but does not surface as a rich result adds maintenance overhead, increases the risk of markup errors, and diverts effort from the types that actually move metrics.

The schema types with consistent, documented rich result production as of 2025 are: FAQ, HowTo, Article, Product, and Review. Each has specific implementation requirements, specific eligibility criteria, and a measurable impact on CTR when implemented correctly.

FAQ Schema

FAQ schema produces accordion-style rich results that expand inline in the SERP. When triggered, a single result can occupy the vertical space of three to four standard results. CTR impact is well-documented: pages with FAQ rich results consistently show higher click share for informational and navigational queries.

Eligibility requirements are strict. Google requires that the FAQ content appear on the page in a user-visible format - the schema cannot describe content that does not exist in the HTML. Each question must be unique. Answer length should be concise; extremely long answer strings are less likely to display in the accordion format.

HowTo Schema

HowTo schema produces step-by-step rich results with optional image support for each step. It is effective for procedural content: installation guides, configuration walkthroughs, and any content that maps naturally to an ordered sequence of discrete steps. The key implementation requirement: each HowToStep must have a meaningful name property and a text property containing the full step description.

Article Schema

Article schema supports Top Stories carousel eligibility, Google Discover indexing, and byline display. The minimum viable Article implementation requires: headline (matching the H1, max 110 characters), image (minimum 1200px wide), datePublished, dateModified, and author with a nested Person or Organization entity. Missing dateModified is the most common Article schema error - Google uses it to assess content freshness.

Product and Review Schema

Product schema with nested Review and AggregateRating properties produces star rating displays in product-adjacent SERPs. This is one of the highest-impact JSON-LD implementations for e-commerce. The eligibility requirement most implementations get wrong: Google requires that review scores reflect genuine user-generated reviews or editorial reviews with clear attribution. Self-serving ratings will not produce rich results and may trigger a manual action under Google's review snippet policy.

What Google Has De-Emphasised in 2025

Two schema types that received significant implementation attention in prior years have diminishing returns in 2025: Breadcrumb and Speakable.

Breadcrumb schema was historically implemented with the expectation that it would increase SERP real estate by displaying the URL path as a visual breadcrumb trail. In 2025, Google's breadcrumb display is largely determined by site structure and internal linking - the schema annotation accelerates detection but does not materially affect whether breadcrumbs display. The CTR impact of breadcrumb display itself is minimal for most query types.

Speakable schema, intended to mark content for Google Assistant audio readout, has seen no significant adoption by Google's voice surfaces and is not eligible for rich results in web search. Resources spent on Speakable implementation in 2025 produce no measurable SEO return.

The question is not whether Google supports a schema type. The question is whether Google surfaces it as a SERP feature that affects user behaviour. Those are different lists.

How to Implement Article Schema Correctly

Article schema errors divide into two categories: structural errors that cause Google's parser to reject the markup entirely, and field-level errors that produce a technically valid schema with missing or incorrect data. Rich Results Test will catch structural errors. Field-level errors require manual review.

The correct JSON-LD structure for Article schema in 2025:

• @context: always https://schema.org
• @type: Article, NewsArticle, or BlogPosting depending on content type
• headline: exact match or close match to H1; never exceeds 110 characters
• image: array of ImageObject entities, each with url, width, and height
• datePublished and dateModified: ISO 8601 format (2025-04-19T09:00:00+00:00)
• author: nested Person entity with name and url pointing to an author page
• publisher: nested Organization entity with name and logo

FAQ Schema Best Practices and Common JSON-LD Errors

FAQ schema is the highest-leverage implementation for informational content, but it is also the most commonly misconfigured. The best practices below address the errors that appear most frequently in structured data audits.

Each Question entity must have a name property containing the full question text and an acceptedAnswer containing an Answer entity with the text property. Limit FAQ schema to questions that appear visibly on the page - Google's quality guidelines explicitly require this. Do not implement FAQ schema on every page indiscriminately; pages already appearing in Featured Snippets often see reduced SERP real estate when FAQ annotations are added.

Common JSON-LD errors to avoid:

• Unescaped quotes in string values - break JSON parsing silently
• Missing @type on nested entities - produces an incomplete schema graph
• Incorrect date format - must be ISO 8601, not natural language dates
• Duplicate @type declarations - CMS plugins often inject Article schema automatically; adding custom schema creates conflicts
• Schema describing content not on the page - both a technical and policy error; Rich Results Test will not catch it

Testing with Rich Results Test catches structural and field-level errors for supported schema types. Run every implementation through it before and after deployment. For bulk validation across a large URL set, use Google Search Console's Rich Results report to identify pages with schema errors at scale.

The rich results landscape in 2025 rewards precise, policy-compliant implementation over breadth of schema coverage. A single correctly implemented FAQ or Article schema block on a high-traffic page produces more measurable impact than a dozen technically malformed annotations spread across the site.

Use the Schema Validator to test your JSON-LD markup against current schema.org specifications and check for field-level errors that Rich Results Test does not surface.