How to get Recommended by LLMs - The Basics
Although traditional Google searches still dominate the bulk of internet queries, it's becoming increasingly obvious that getting mentioned in LLM recommendations is the new battleground for brands. What's less obvious is HOW to get recommended by LLMs. With a rapidly changing environment, the line between fact and myth is often blurred, so let's start with the basics.
Phase 0: Understanding LLMs
Before we can influence LLM responses, we need to understand how the average user might encounter them.
Foundation Models vs RAG Systems
Foundation Models serve as the base for LLMs – GPT-5, for example, when not utilizing Web Search. These models are already trained with astronomical amounts of data to learn language and reasoning but their actual knowledge is static. The models themselves do not perform searches and synthesize new data or content, so all of your GEO efforts won't really have a direct impact on the current version of these models.
Retrieval-Augmented Generation (RAG) is the name of the method or system that takes a foundation model and "retrieves" new data, such as through search queries, before processing a response. A real world example of this would be GPT-5 with Web Search enabled or Perplexity. These responses are where your Generative Engine Optimization (GEO) efforts can pay off.
RAG systems are also becoming more common; for example, even for free users, ChatGPT's web browser platform now automatically performs searches for chat prompts that warrant them.
Phase 1: Visibility into your Visibility
Brand visibility or AI visibility, in GEO terms, refers to how and how often your brand appears in LLM responses.
Setting Up Proper Tracking
Setting up proper tracking foundations is crucial for understanding your current visibility and for measuring what actually moves the needle moving forward. Consider a tool like trackerly.ai to begin tracking LLM responses to common real world user prompts in order to understand how AI talks about your brand (if at all), where you fit in amongst your competitors, and what sources are being leveraged to inform those responses.
Tip: A good place to start is with Google Search Console. Look at your most common queries and convert them into AI prompts.
For example, a Google search of "best gaming headphones 2025" might translate into a more natural-sounding "What are the best gaming headphones in 2025?" when prompted by the same user in ChatGPT/Perplexity. For best results, track many variations of your most important prompts.
Phase 2: Help AI Help You
Much like how Tom Cruise pleaded with Cuba Gooding Jr. in Jerry Maguire, help AI help you. LLMs synthesize data in an entirely different way than traditional SEO, and GEO begins with making your content more LLM-friendly.
AI Crawler Accessibility
Confirm with your internal IT team and your server host that nothing is blocking AI crawlers from accessing your content. Common blockers include:
- robots.txt restrictions: Remove any "Disallow" rules for common crawlers such as GPTBot or PPLX-Bot
- JavaScript: While this is changing, many AI crawlers still do not execute JavaScript, so dynamic content can be skipped over entirely. Consider using SSR or prerendering for important GEO-related content.
- Firewall / bot protection tools: Services like Cloudflare or Akamai can block AI Crawlers. Consider whitelisting trusted agents.
- Web server or CMS blocking: Review .htaccess and NGINX configs to see which, if any, AI crawlers are blocked
- Rate Limits or CAPTCHA: Crawlers may trigger security tools and appear as DDoS threats. Consider allowlisting trusted bots.
- Meta tags: Tags like
<meta name="robots" content="noindex, nofollow">can discourage crawling
Content Structure Matters
Aim for:
- Clear and consistent formatting of H1 and H2 headings
- Opt for lists and bullet points over long paragraphs
- Concise section summaries
- FAQ schema
- Multiple schema types
Create AI-Friendly Guidelines
Develop content that's structured, concise, consistent across platforms, and directly answers questions. Start creating content specifically for AI citations—think about how an AI crawler would extract and reference your information. Even better, make the subheading the actual question you are answering.
Phase 3: External Credibility
Now that your own content has been optimized and tracking foundations have been properly set up, the next step in improving your brand visibility is strengthening the external voices that impact how LLMs talk about you.
Citation Analysis and Strategy
With your new citation data, which third party sources are heavily influencing responses? Is social content being heavily referenced? Editorial/advertorial websites? Reviews on reference sites?
Focus your efforts on strengthening your presence in the areas that matter.
How credible external sources talk about you are much more influential to your AI rankings than backlinks. But even in this context, the AI landscape is rapidly changing, and what's important one day may not be as important tomorrow.
For example, Reddit, once the Holy Grail for GEO-minded brands, is seemingly becoming much less cited and influential, for reasons that may have nothing to do with the content itself.
Citation tracking is an ongoing necessity, and not a one time audit.
The Bottom Line
As LLM recommendations continue to grow in both authority and ubiquity, understanding and strengthening your AI visibility is non-negotiable.
In a landscape where the rules can flip overnight with a policy change from Google or a new model update, having the tools and teams in place to quickly understand your visibility and what's directly impacting that visibility will be the difference between those that get uplifted by AI search and those that get buried by it.