What's the Best Generative Engine Optimization Strategy?

Most marketers still write blog posts with search engines in mind. But LLMs (Large Language Models) have changed the game. Tools like ChatGPT, Perplexity, and Claude don’t just show a list of links. They generate direct answers.

That means if your content isn’t the clearest, most useful answer to a specific question, it probably won’t show up.

This post breaks down the best Generative Engine Optimization (GEO) strategy. There are different strategies (check out our blog for more). But if you only have one thing to prioritize, it’s to write clear, structured answers to actual questions people ask. 

SEO-First Strategy 

Before we discuss the best GEO strategy, it’s essential to review how we typically approach SEO strategy pre-LLMs.

Picking high-volume, low-difficulty keywords.

• Optimizing content for those terms.

• Hoping to rank in SERPs.

Keywords are typically short, like SEO vs. GEO. But even long-term keywords are still short compared the the questions that users typically ask in generative AI chats like ChatGPT. For example, a long-tail keyword is “how much does AI cost.”

Now, queries on generative engines are much longer and with a lot of context. For example, the questions could be something like these:

• “How much should a small business expect to spend monthly if they want to automate customer service using AI?”

• “What are the typical costs for a startup using AI tools for content creation and lead generation?”

• “How expensive is it to integrate AI into an existing SaaS platform without hiring a full ML team?”

As you can see, the questions have a clear intent and a specific user. For example, you can tell that the question is from a small business owner who wants to implement AI in customer service. These types of queries won’t appear in traditional SEO tools. And even if they did, you might see that the search volume is very low.

Pro tip: Use Anvil to understand what kind of queries people ask about your brand. 

The Best GEO Strategy 

Answering a clear, structured question from a specific user is the most effective GEO strategy. 

Generative engines reward:

• Clarity

• Factual structure

• Direct responses to natural-language prompts

When you write the clearest, most helpful answer to a common question—even if it has low search volume—you increase your chances of:

• Being cited in Perplexity.

• Appearing in ChatGPT’s responses.

• Being paraphrased or sourced in Claude.

The title of this article is a prime example of a question that people might ask in LLMs. 

I know it can be anxiety-inducing to write new blog posts without obvious data like search volume and keyword difficulty. But trust me, it works.

As someone who has had an SEO-first strategy for the past decade, it was hard for me to roll the dice and write without a lot of SEO data. But the more you write about long, conversational questions, the more you’ll get your own data.

For example, you’ll see real, first-hand impressions and clicks from your own site instead of outdated, third-party SEO tools.  For example, I saw a question on Reddit, “How to track your brand in ChatGPT.” I wrote a blog post about it and started to discover more keywords and questions related to it. 

The question didn’t have any search volume on Ahrefs or Surfer SEO.

How to Find the Right Questions to Write About

Here are five ways to uncover real, intent-rich prompts your audience is already asking.

1. Talk to Your Users

This is the fastest way to uncover gold.

If you run onboarding calls, support chats, or customer interviews, you’ve probably already heard:

• “We’re trying to use AI to [X], but how much will it actually cost us?”

• “Can I use this without needing an engineer?”

• “Is there a free version that does this?”

Write those questions down. These are exactly the kinds of queries people type into ChatGPT.

Pro tip: Ask your team to log every “How do I…?” or “Can I…?” question. Turn the most repeated ones into blog posts.

2. Use Google and Bing’s “People Also Ask”

Search your main topic and expand the “People Also Ask” box. You’ll find questions like:

• “Is AI expensive to run?”

• “How much does OpenAI charge for API usage?”

• “What’s the cheapest AI tool?”

  
  

3. Use GEO Tools to See What AI Says

Luckily, nowadays, there are new tools that can help you find out what kind of questions people ask in LLMs. Tools like Anvil help you track:

• What questions your brand is showing up for.

• Which tools or sites ChatGPT is pulling citations from.

Whether Claude or Perplexity is paraphrasing your messaging accurately.

You’ll see exactly what LLMs say about your brand—and what they say about your competitors. That makes it easier to spot question gaps and under-covered topics worth writing about.

4. Check Forums and Reddit

Forums like Reddit are incredible sources of raw, real questions. Search for your product category or problem area, and filter by:

• Threads with 10+ comments.

• Posts asking for help choosing between tools.

• Common pain points that repeat across different subs.

You can also use the Trends or Answers tab in Reddit. You can enter the topics on Trends, and it will provide the latest and most popular threads to comment on.

Answers, on the other hand, is currently in beta. It’s essentially Reddit’s internal LLM, where you can ask questions. It will provide you with more subreddits featuring real questions.

5. Look at Your Own Search Bar and Support Tickets

If you’ve got an on-site search function, see what people are typing in. These are your low-hanging content opportunities. Same goes for:

• Zendesk queries

• Live chat logs

• Internal Slack discussions between sales and support

Even if the questions aren’t phrased perfectly, you’ll see patterns. Those are the prompts to write toward.

Note that this is just one of the strategies. Showing up on LLMs requires the right technical infrastructure on top of the right content. 

The Best of Both Worlds 

Most SEO strategies start with keywords. GEO starts with real questions—the kind your users are already asking in ChatGPT, even if they don’t show up in your keyword tool.

You don’t have to choose between the two. The smartest content teams are doing both: creating keyword-optimized content for search and writing helpful, conversational answers for generative engines.

If you can clearly answer the real questions your users are asking—even when search volume says “zero”—you’ll start showing up where it matters: inside the tools people trust to make decisions.

FAQs about GEO Strategy

How is GEO different from traditional SEO?

Traditional SEO is focused on ranking in Google SERPs using short keywords, backlinks, and metadata.

GEO (Generative Engine Optimization) is focused on being cited or included in AI-generated responses from tools like ChatGPT, Claude, and Perplexity. 

Should I stop doing keyword research if I focus on GEO?

No. Keyword research is still useful, especially for identifying demand. But GEO adds another layer—answering questions that don’t yet show up in keyword tools but are still being asked in LLMs. The best strategy is to do both: optimize for search and for conversation.

What types of questions should I prioritize for GEO content?
Prioritize questions that are:

• Clear and specific

• From a defined user persona (e.g., founder, marketer, engineer)

• Focused on a real-world problem or use case
  Even if the volume is low, they often show up in LLM prompts—especially if no one else has answered them well yet.

How do I know if my brand is showing up in ChatGPT or Perplexity?

You can test it manually by prompting ChatGPT or Claude with relevant questions and seeing if your brand appears. Or, use a tool like Anvil to automatically monitor where and how your brand is mentioned across multiple AI engines.