Schema markup — also called structured data — is one of the most powerful and least understood things you can add to a website. This guide explains exactly what it is, why it matters for both Google and the new AI search engines, and how to add it to your own pages without writing a single line of code, using free tools. It is written for non-technical website owners, so there is no jargon you will not understand.
Schema markup is a small, hidden block of code you add to a web page that spells out, in a format computers understand, exactly what the page is about. A human reading your page can tell at a glance that it is a recipe, a product or a local business. A search engine, however, is largely guessing. Schema removes that guesswork by stating it plainly — for example, “This is a Recipe. The name is Victoria Sponge. Preparation time is 20 minutes. It serves 8.”
The official name for this is structured data, and the format Google recommends is called JSON-LD. It is a tidy block of code that sits in your page’s HTML without changing anything a visitor sees. This is an important point: schema is completely invisible to your visitors. It does not change your design, your text or your layout. It exists purely to communicate meaning to machines.
There are three major reasons to care about schema, and all three are becoming more important over time.
First, rich results. Those eye-catching Google listings with star ratings, prices, FAQ dropdowns, recipe images and event dates are almost all powered by schema. Without schema, you cannot earn a rich result. A plain blue link attracts far fewer clicks than one showing gold review stars or an expandable list of questions.
Second, better understanding. When you remove the guesswork about what your page is, Google is more likely to rank you for the right searches and less likely to misinterpret your content.
Third, and increasingly the most important, AI search engines rely heavily on structured data. ChatGPT, Perplexity, Google’s AI Overviews and Microsoft Copilot all favour pages whose meaning is spelled out clearly. Schema is one of the strongest signals you can give them. As more people search by asking an AI assistant a question rather than typing keywords into Google, schema becomes the difference between being quoted as a trusted source and being ignored.
You will encounter a handful of terms. Here they are in plain English. Structured data is the proper name for schema: information laid out in a fixed, predictable shape a machine can read. Schema.org is the shared dictionary — a website run jointly by Google, Microsoft, Yahoo and others that lists every type you can describe and the details each one accepts. JSON-LD is the recommended format for writing schema, a neat block wrapped in a script tag. A type is what your page is, such as Product, Article, LocalBusiness or FAQPage. A property is a detail about that type — a Product has a name, a price and a brand; an Event has a start date and a location. That is genuinely most of the vocabulary you need.
Before adding anything, ask one simple question: what is this page, really? Pick the closest match. Use LocalBusiness for a shop, café, tradesperson or clinic — anywhere with an address and opening hours. Use Product for something you sell, ideally with a price and reviews. Use Article or BlogPosting for a blog post, news piece or guide. Use FAQPage for a page that answers a set of questions. Use Service for a service you offer, such as plumbing or accountancy. There are also specific types like Recipe, Event and Review. A single page can carry more than one type — a product page might be a Product with Review and FAQPage on it as well — but always start with the one main thing the page is. If you are unsure, the AI Schema Generator reads your page and picks the correct type for you automatically.
Here is a quick reference for the most common schema types and when to use each one:
| Your page is… | Use this type | Key details it expects |
|---|---|---|
| A shop, café, tradesperson or clinic | LocalBusiness | name, address, telephone, opening hours, geo |
| Something you sell | Product | name, image, price, availability, reviews |
| A blog post, news piece or guide | Article / BlogPosting | headline, author, datePublished, image |
| A page answering questions | FAQPage | a list of Question and Answer pairs |
| A service you provide | Service | name, provider, areaServed, description |
| A step-by-step tutorial | HowTo | name, steps, tools, total time |
| An event | Event | name, startDate, location, offers |
| A recipe | Recipe | name, ingredients, instructions, cook time |
For example, a plumber in Leeds would use LocalBusiness (with a Service for each job type), a Shopify-style store would use Product on each item page, and a how-to article like this one would use Article with a FAQPage for its questions.
Here is the news that makes this whole subject practical for non-technical people: you never have to write schema by hand. Hand-writing JSON-LD is fiddly and easy to get wrong, because a single missing comma causes the entire block to fail silently. Instead, paste your page’s web address into the AI Schema Generator. It reads the live page, works out what it is, and writes the complete, correct JSON-LD block for you, covering more than 40 schema types and choosing the right sub-type automatically. You receive a tidy block of code, and your only job is to place it on your page. If you would rather build it field by field yourself, the Schema Builder lets you do that with simple dropdowns — still no raw coding required.
Let’s use a real one — this very site. To a human, the AIWebPageSEO contact details read clearly, but to a search engine they are just words on a page:
Run through the AI Schema Generator, that same information becomes an Organization block a search engine understands instantly:
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "AIWebPageSEO",
"url": "https://aiwebpageseo.com/",
"description": "Free webpage audit and schema generator.",
"address": {
"@type": "PostalAddress",
"streetAddress": "Bridge St Tupton",
"addressLocality": "Chesterfield",
"addressRegion": "Derbyshire",
"postalCode": "S42 6BT",
"addressCountry": "GB"
},
"email": "advice@aiwebpageseo.com",
"telephone": "07562117348",
"sameAs": [
"https://x.com/AiPageSeo",
"https://www.facebook.com/Aipageseo/",
"https://www.instagram.com/aipageseo/"
]
}
Notice every line matches something genuinely true about the business — the name, the address, the contact details, the social profiles. Nothing is invented. That is exactly the rule to follow, and it is why generating from your real page is so reliable: the tool only describes what is actually there.
Once you have your block of JSON-LD, it needs to live in your page’s code. This is where two free programs come in: FileZilla, which moves files between your computer and your website, and Notepad++, which edits them safely. First, open FileZilla, connect to your website, find the page file you want to add schema to — for example index.html or about.html — and download a copy. Second, open that file in Notepad++, find the closing </head> tag near the top, and paste your entire schema block on the line just before it. Third, save the file and upload it back with FileZilla, overwriting the old version. That is the whole job. The schema is now on your page, invisible to visitors and speaking clearly to Google. If you are new to these tools, the setup guide walks you through installing and using both.
Never assume schema works just because you pasted it in, because a stray character can break the whole block and you would never know from looking at the page. Always test. Paste your page, or the schema block itself, into the Schema Debugger. It validates the code, checking it is correctly formed and follows schema.org’s rules, and it scores the markup from 0 to 100 against Google’s Rich Results requirements, telling you in plain English what is missing or wrong — for example “Product is missing a price” or “Review has no rating.” A high score means you are eligible for rich results. A warning tells you exactly what to fix, so you regenerate or tweak and test again. Re-test whenever you change the page.
A few errors quietly sink most people’s schema. The biggest is marking up things that are not actually on the page: the golden rule is that if a visitor cannot see it on the page, it does not go in the schema. Inventing reviews or FAQs that are not really there is the fastest way to a Google penalty, because schema must match reality. The second mistake is choosing the wrong type, such as tagging a blog post as a Product. The third is broken JSON, where a missing comma or bracket kills the whole block silently — which is exactly why we generate and test rather than hand-type. The fourth is forgetting to re-test after later edits, so you break the schema by accident and never notice. The fifth is faking star ratings: only mark up genuine reviews that appear on the page, because Google actively hunts for fabricated ones. Avoid these five and you are ahead of the vast majority of websites.
Schema began as a way to help Google, but it has become one of the most important things you can do for the new wave of AI-powered search. When an AI assistant answers a question, it must decide which sources to trust and quote. A page that clearly states “I am a LocalBusiness called X, in town Y, offering service Z” is far easier for an AI to use confidently than one it has to interpret from messy HTML. Two schema types are especially powerful here. FAQPage, which presents question-and-answer pairs, matches exactly the format AI engines prefer because users ask questions — this is one of the highest-impact additions you can make. Organization schema tells AI engines who you are, where you are based, and links your social profiles, building trust. You can see how visible your page is to the major AI engines right now with the AEO Checker.
To summarise the whole process: decide what your page is; generate the JSON-LD with the AI Schema Generator; paste it just before the closing </head> tag using Notepad++ and upload with FileZilla; test and score it with the Schema Debugger; only mark up what genuinely appears on the page and re-test after any change; and add FAQPage and Organization schema to help AI search. Do that on your important pages and you will have given Google and the AI engines the clear, tidy label they reward.
No. The AI Schema Generator writes the JSON-LD code for you. Your only job is to copy it and paste it into your page using a free editor like Notepad++, then upload the page with FileZilla.
No. Schema is completely invisible to visitors and does not affect your design, text or layout. It only communicates meaning to search engines and AI engines.
JSON-LD is the format Google recommends, and it is what our tools generate. It sits in a script tag in your page’s head section.
Test it with the Schema Debugger, which validates the code against Google’s Rich Results rules, scores it and tells you what to fix. Re-test whenever you change the page.
Use the type that matches what your page is: LocalBusiness for a shop or tradesperson, Product for something you sell, Article or BlogPosting for a post, FAQPage for a questions page, and Service for a service. A page can use more than one type.
Yes. AI engines favour pages whose meaning is clearly stated. FAQPage and Organization schema in particular make your page easier for ChatGPT, Perplexity, Google AI Overviews and Copilot to understand and quote.