Using Fanout Queries to Come Up with Content Topic Ideas for GEO Strategies (A Practical Guide)

When you type a prompt into an AI assistant like ChatGPT, it doesn’t search your prompt verbatim. It breaks it down into subqueries (called fanout queries) to ground its answer in a web search (assuming it does one).

These fanout queries cover different topical realms and lines of intent, allowing the AI assistant to put together a more holistically satisfying answer.

Ultimately, your goal with a GEO strategy is to be used as a source for a generated answer, potentially contributing to a mention or recommendation or giving your site a clickable citation.

To become a source, your content generally needs to be retrievable and relevant to one or more of the fanout queries. Relevance isn’t the only factor, but it is the foundation for being considered.

In this guide, I’ll show you how to use fanout queries to identify content opportunities for your GEO strategy.

And in case you want to explore this topic through a video, I’ve got you covered:

The first question is: where can we find fanout queries?

As far as I know, you can only see fanout queries for ChatGPT, Perplexity, and Copilot. To get fanout queries for Google’s AI platforms like Gemini, AI Overviews, or AI Mode, you’d need to use a simulation tool, like Qforia.

The free and easy way to see fanout queries for ChatGPT is to use a Chrome extension.

ChatGPT Path by Ayima, for example, allows you to see the fanout queries for your ChatGPT conversations. Here is an example from an SEO news prompt:

Other extensions for this purpose include FanoutFox.

But this technique is limited to individual conversations.

A more scalable way to get query fanouts is to use a prompt-tracking tool.

I track 50 SEO/GEO-themed prompts across ChatGPT in Peec AI. This means I can visit the Fanouts section of the tool and view all of the fanout queries for my tracked prompts:

But how can we put these fanout queries to use?

You might hear a recommendation to use your query fanouts for keyword optimization, such as by inserting them directly into your content.

That’s clunky, likely to lead to awkward wording, and doesn’t align with how semantic search works today.

Instead, we want to take inspiration from our fanout query themes to come up with content topics. Because, ultimately, our goal is to be relevant for the fanout queries that matter to our target audience’s buyer’s journey across AI assistants.

The first step in getting content ideas from fanout queries is to export all of your fanout queries from your prompt-tracking tool into a CSV file.

Once you have that file, you can upload it to an AI assistant (here I’ll use ChatGPT) and use a prompt such as:

“I’m uploading an exported list of fanout queries for prompts I track. Please evaluate the fanout queries for common themes and then give me content suggestions based on those themes. My goal is to create content that’s relevant for the fanout queries and that my target audience would find helpful.”

You’ll get an output that looks something like this, with content ideas divided by topical themes, based on your fanout queries:

You could take inspiration from those themes alone, but in addition to that analysis, you’ll also get suggested content topics to create:

Since these content topics are based on fanout queries related to prompts you’re tracking (and assuming your prompts cover the buyer’s journey of your target personas for your main lines of business), they’ll be relevant to your AI visibility goals.

The next step is to actually create the content.

This is where strategy comes into play, because AI assistants can rely on multimodal sources.

That means you’ll want to create content for the topics identified from your fanout queries in the best formats to get AI visibility from. This may include written blog posts or guides, or it may include YouTube videos.

To get a sense of which types of content to create, you can look at the sources associated with individual topics in your prompt-tracking tool.

For example, for my Content Marketing Strategy topic, I might want to create a YouTube video, a holistic guide, or a LinkedIn Pulse article (based on the sources I see retrieved for those prompts):

Analyzing Google’s SERPs or AI Mode answers can be another way to infer the intent of a topic.

For example, “How AI Systems Select Sources” was one of the suggested topics for me from ChatGPT, based on my fanout query data. If I search that topic in Google and AI Mode, I largely see written guides (blog posts), which tells me that’s the type of content I should create.

What are the main takeaways?

This guide showed you how to export fanout queries from your prompt-tracking tool, upload them to an AI assistant, and get back content ideas for your GEO strategy.

The key is having good data to start with: the prompts you track should be relevant to the buyer’s journey of your target personas for your primary lines of business. Those prompts will then produce fanout queries that will be relevant to your AI visibility goals.

If you don’t have a prompt-tracking tool available, you can use a Chrome extension to view the fanout queries in ChatGPT for individual conversations.

Of course, if you want a hand with this type of work, feel free to get in touch with me. I’m an independent SEO/GEO consultant helping brands and agencies with their AI visibility strategies.

Until next time, enjoy the vibes:

Thanks for reading. Happy optimizing!

Related posts