Schema markup has been around for over a decade. It's structured data that tells search engines what information on your page means. A phone number is a phone number. An address is an address. A review rating is a review rating. For years, the main value was that Google would use it to show richer search results. But schema markup has a new, more important purpose: it makes it much easier for AI systems to extract accurate information from your pages. When an AI system pulls information from your site to include in a recommendation or an answer, clean schema markup ensures the information is extracted correctly.
The shift matters. ChatGPT and Claude sometimes pull information from websites when answering questions. Perplexity regularly pulls from sources when building answers. If your page has no schema markup, these systems have to guess what information means. They might confuse your contact information with a client's information. They might pull a phone number and attribute it incorrectly. They might extract pricing that's outdated. Schema markup removes the guesswork. It says: "This is the information. Here's what it means. Use it as you need."
Core Schema Types for Businesses
Not all schema markup is equally important. Start with the basics. Every business should have Organization schema that includes your company name, logo, contact information, and social media profiles. This tells systems who you are and how to contact you. Add LocalBusiness schema if you have a physical location. If you're in a specialised field like professional services, healthcare, or hospitality, add schema specific to your vertical.
For a wealth management firm, BroadcastService schema makes sense. For a healthcare provider, add schema for your medical services, your staff credentials, and potentially Patient schema if you're accepting new patients. For a hospitality brand with multiple properties, use HotelSchema for each property. For a professional firm, add Person schema for key team members with their roles and credentials. The goal is to give systems complete, structured information about who you are and what you do.
Many businesses implement schema halfway. They add Organization and LocalBusiness but don't go deeper. That's understandable—schema markup takes effort. But in the AI era, the businesses that give AI systems complete, clear information will be the ones that show up in answers and recommendations correctly. If you're going to implement schema, do it thoroughly. Make the effort count.
Articles and Content Schema
If you publish content—articles, blogs, guides, research—add Article schema to each piece. Include the headline, publish date, author, main image, and a description. This helps AI systems understand what your content is about and who wrote it. When an AI system is looking for sources to cite when answering a question, Article schema makes it much easier for the system to evaluate your content and decide if it's relevant.
Author schema is particularly important. Tag articles with the author's name and, if available, a link to their author bio page. This helps systems understand who wrote the piece and whether that person has authority on the topic. If you're building thought leadership—where you or your team members are becoming recognised authorities in your field—Author schema amplifies that. It tells systems not just what you wrote but who you are.
For deeper content like reviews or case studies, use Review or Case schema. For FAQs, use FAQPage schema with specific questions and answers marked up. These structured formats make it easier for AI systems to pull the exact information they need without having to parse through unstructured text. The easier you make it for systems to understand your content, the more likely they'll use it correctly in answers.
Local and Geo Schema
If you serve customers in specific locations, add GeoShape or GeoCoordinates to your LocalBusiness schema. Include your service areas. If you have multiple locations, add separate LocalBusiness schema for each one. This helps systems understand where you operate and which questions you're relevant to. If someone asks an AI system about wealth management services in a specific city, and you're marked up with that city in your schema, you're more likely to show up.
Service schema is useful for multi-location or service-based businesses. Define each service you offer, the areas you serve, and pricing if you want to include it. This is particularly useful for professional services, healthcare providers, and hospitality brands. Systems can then understand your service offerings and match them to questions more accurately.
Ratings and Aggregate Rating Schema
If you have reviews or ratings, mark them up with AggregateRating schema. Include the rating value and the number of reviews. This tells systems that your business has been rated by actual customers. When an AI system is deciding which businesses to recommend or cite, it often looks for signals of customer satisfaction. Marked-up ratings make those signals visible and verifiable.
Note that for this to work, your ratings need to be real and recent. Aggregating a 4.8 star rating that's based on reviews from five years ago might hurt more than help. Systems can sometimes tell when ratings are outdated. Fresh, high ratings are what carry weight. If your ratings are poor or nonexistent, you're better off not marking them up and instead focusing on improving them genuinely.
Implementation Best Practices
Schema markup can be implemented in several ways: directly in your HTML using JSON-LD script tags, as microdata within your HTML, or using RDFa. JSON-LD is cleanest and most widely supported. It's easier to manage and less likely to conflict with your styling or layout. If you're using a modern website builder or content management system, you probably have schema markup built in. Check what's already there and supplement it with missing types.
Validation is important. Use Google's Rich Results Test or the Schema.org validator to check your markup. Fix any errors. Invalid schema won't provide benefits and might actually hurt you. Make sure all required fields are present and correctly formatted. Common mistakes include inconsistent phone number formats, addresses that don't match your service area, or author information that's vague. Fix these before deploying.
Update your schema as your business changes. If you add a new service, add it to Service schema. If you move locations, update your address. If you get new reviews, update your aggregate rating. Schema markup is not set-it-and-forget-it. It's part of your ongoing site maintenance.
Why This Matters for AI Visibility
Google still cares about schema markup for generating rich snippets and rich results. But in the AI era, schema markup matters more because it helps AI systems understand your information correctly. When Claude or Perplexity pulls information from your page, clean schema markup ensures that information is accurate and complete. When systems are deciding whether to cite you, clear schema makes it easier for them to verify that you're a legitimate business with real services and real customer feedback.
It's not a direct ranking factor for AI systems the way it might be for Google. It's more foundational. It's about making sure that when AI systems interact with your site, they extract the right information and present it correctly. In a world where citations matter more than rankings, having correct information in those citations matters enormously. Schema markup is how you ensure accuracy.
If you haven't audited your schema in the past six months, do it now. Add any missing types. Fix any errors. Update information that's changed. Schema markup is low-effort, high-impact work. It takes a few hours to implement properly but can prevent months of misdirected visibility. In the AI era, it's a priority.
— Sam