How to Get Cited by LLMs

Hello Riddlers!

Today's newsletter sponsor, is myself 😄

Check out my python course --> Python for SEO

Last week, I attended SEO IRL, and it was such a great event! The talks were full of insights, and in today’s newsletter, I want to share some of the best tips I picked up.

But instead of covering everything at once, I decided to focus on one specific talk because it truly deserves a closer look. The best part? It’s incredibly actionable!

The talk: “Truth, Not Traffic: How to Win AI Search” by Steve Toth.

So buckle up!

Disclaimer: I'm not affiliated with Steve. Also, the information listed here are both notes from Steve's talk and my personal thoughts, so not everything written here is something Steve said or agrees with, some are my personal thoughts and opinions.

Chatgpt Deep Research

If you're not familiar with the Deep Research feature in chatgpt it's basically "an AI agent integrated into ChatGPT, which generates cited reports on a user-specified topic by autonomously browsing the web for 5 to 30 minutes." [1]

You can use this feature by selecting it from the drop down menu in Chatgpt:

I usually use Vertex AI (and recently Gemini deep research) when I need in-depth research to be done on a topic.

Steve focused his talk on tips to optimize for Deep Research.

Why is deep research tool important? in my opinion, deep research is a great tool to figure out how Chatgpt thinks about you.

Here's a quote from Steve's slides:

"ChatGPT Deep Research gives asks up follow-up questions after the initial query. Optimizing for these refinements ensures your content answers these likely subsequent questions, maintaining visibility in the conversation"

Refinement Questions

I know right now we're all obsessed with "query fan-out" which is a very cool concept, but refinement questions that Agentic search generates can act as valuable clues as well on what areas you need to cover in your content to increase your potential of being cited in Chatgpt.

Here's an example of what refinement questions are in deep search:

So this for example gives me insights of what to talk about on my newsletter landing page.

You can also switch models, to get more of those refinement questions [2].

Then what?

Steve built a GPT that takes all those refinement questions you have gathered from different models then:

1. Cluster them by theme
2. Remove duplication
3. Find URLs on your website that can be used to answer those questions.
4. Generate a list of content topics to cover.

Here's an example from Steve's deck after all the collected questions were clustered and prioritized for "employer of record software":

Don't forget the activity tab!

I went ahead and provided answers to the refinement questions for "best SEO newsletters". Once you do, you'll see a tab on the right, the "activity" tab.

Once it's done, search for "searched for" to see how those refinement questions resulted in few "query fan out" searches.

So now you also have a list of keywords/searches that you can optimize for. HOWEVER Steve does not recommend you optimize for every single possible query (and I agree).

So because there can be so many of those queries, and they can change every time you search, here's what you should do instead from Steve's deck:

• Identify the common themes of fan-out queries and optimize pages towards that. 
• Target specific fan-outs on the passage level of those pages instead.

Common refinement themes

Steve shares that based on his observations, these are the common themes you'll get refinement questions about:

I LOVE this slide, why? because I've been doing something similar in my content strategy for a while now and this just will take my work to the next level.

These themes can serve as "type of pages" you should have on your website. So for example comparison pages between you and each competitor, or [your product] for [specific type of customers] like [SEO tools] for [agencies].

And That’s a Wrap (Almost 😄)

Honestly the talk has way more topics covered than I highlighted here! Feel free to scroll through the deck yourself.

One thing was very clear to me during this talk, the people saying GEO is just SEO have not done the work themselves before. There's a lot of "ADDITIONAL WORK" you need to do to:

• get cited in LLMs,
• to get "properly" cited in LLMs
• to respond to deal breakers and convert users minds ON THE LLM interface

I'm told that this talk is built on Steve's course "Win more qualified buyers from LLM search with answer engine optimization", so check it out if you want to learn more (again, I'm not affiliated with Steve)

Thanks for reading so far and see you next newsletter!

Support the SEO Riddler!

• Sign up for my newsletter if you're not already.
• Share the newsletter and invite your friends to signup. Help me reach 2k signups by end of 2025 please 🙂
• Provide feedback on how I can make this newsletter better!!!
• Buy me coffee.
• If you're an SEO tool or an SEO service provider, consider sponsoring my newsletter. I'm also open to other partnership ideas as well.

Disclaimer: LLMs were used to assist in wording and phrasing this blog.

Sources

[1] Deep Research | Wikipedia

[2] How to Rank in Chat Deep Research - Paying Attention to Refinements