FullSchema started as my own tool
I'm Kadir Yaz. For years I researched and wrote schema markup for the websites of the brands I manage. FullSchema is that work, turned into an app anyone can use.
Updated
In digital marketing since 2010
I've worked in digital marketing professionally since 2010. These days I consult for companies on digital marketing and on bringing AI into the way they work.
Built from what worked on real sites
On the websites of the brands I manage, I wanted every page to carry complete, correct schema. Getting there took years of research: working out the right format for each kind of page, building it, publishing it and keeping what held up.
Some of those practices gave really good results, so I turned them into software. I'm convinced a lot of people need this right now: marketers, founders, agencies, brand owners, small businesses and anyone else who looks after a website.
Why not just ask a chat assistant?
I tried that. Paste a link into a chat assistant and you get a long block that looks right. The problems are the ones you can't see: gaps filled with plausible guesses, like a rating, opening hours or a price that isn't current, and nothing checks the values against the live page. Most people paste it in and assume the job is done.
FullSchema works the other way around. It crawls the live page, its source code and your site's own data, and writes schema only from what it finds there. Code then checks prices, phone numbers, dates and coordinates against the page and drops any value it can't find. Longer descriptions that aren't in the page's own words are dropped too. The result is validated against schema.org and Google's rules for each rich result before you see it.
AI does one part of the job: reading the page the way a person would. Everything it writes is checked by code before it reaches you.
Tested on real business pages, again and again
FullSchema is tested on a fixed set of 250 real business pages: online stores, clinics, hotels, restaurants, law firms, SaaS companies and more, from 27 countries, in 13 languages and on 19 website platforms. So far we've generated schema for them close to 5,000 times.
Each of those pages has its own answer key, a list of what its schema should say, and test results are compared with it. When a result gets something wrong, the fix goes into the rules for every site, not a patch for that one page.
How the guides are written
The setup guides are drafted with AI help and reviewed by me. Every fact about another company's product, like a menu path, a setting or a price, is checked against that company's own help pages before it goes live, and each guide shows the date it was last updated.
The schema markup study is our own data, with the method and a CSV of the raw numbers published alongside it.
Tell me what would make it more useful
FullSchema has one goal: the most useful result for your page. If a result looks off, an install won't stick, or you want something it doesn't do yet, please write. Questions and feedback are always welcome.
See it on one of your pages
Your first pages are covered by the free balance. Paste one and look at the block before you decide.