Schema / AI citations

This is why you shouldn't let ChatGPT run your SEO. How to do SEO for LLMs?

This is something that a lot of people are searching for right now. Key tactics: right answer first content.

So if you're targeting informational keywords, you should know that AI overviews reduce clicks by 58%. A lot of the time AI will cite you, but people won't even care.

They just want to get their answer and move on. Instead, you should target high intent keywords.

Structure content clearly. That's good advice.

Build entity authority. Also good advice.

Add structured data. Here's an interesting thing about adding structured data: it takes a while to add and have it be perfect.

You often have to maintain it, so that takes extra time. When people say structured data, they mean schema.

Hrefs just did a study. We tracked 1,885 pages adding schema.

AI citations barely moved. If you spend a lot of time adding schema for LLMs, it's just going to be a waste of time.

Google has constantly been dropping support for more schema types. Google will ignore a lot of schema.

Big waste of time right now...

URL slug, your H1, the beginning of your first sentence. To see what ChatGPT searches, copy the part of your URL that comes after forward slash C, right click inspect, go to network, paste in what you just copied.

Refresh the page, click on the orange brackets with what you just copied. Go to response.

Search this for the word queries. Right here you can see exactly what ChatGPT searched.

Keep content fresh. Funny enough, the same study I was just showing you reveal the average cited pages.

500 days old and still getting picked. You want to do SEO where you're not wasting your time on things that don't matter.

I have this article on what does matter: A I S E O G O A E O, how to get shown in LMS in 2026. You you can get it at free SEO knowledge com.

This is how to do SEO for free. Don't waste your time on things that won't get results.

One of the most common ways SEO agencies will rip off clients is with unnecessary schema implementation. Schema is code that makes you get special features in Google like this.

With no schema. With schema, you get a lot more clicks when you get the special feature in Google.

But Google has actually deprecated a lot of schema over the years, and agencies just ignore this. Now the new narrative is schema makes you get cited in LLMs.

If you hear people say you need structured data to get mentioned in LLMs, they're usually talking about schema. Ahrefs, one of the most reputable SEO software companies, just did this study.

We tracked 1,885 pages adding schema. AI citation is barely moved.

Adding schema didn't boost citations on any platform. It had a negative effect on Google's AI overviews, statistically indistinguishable from 0 on Google AI mode, also statistically indistinguishable from 0 in ChatGPT.

This was a thorough study. You can see it for yourself here.

Talk to your agency to make sure you're only focusing on schema that gives you special visible features in Google...