How to Rank in ChatGPT and Claude: The Ultimate 2025 Guide to LLM Mentions

How to Rank in ChatGPT and Claude: The Ultimate 2025 Guide to LLM Mentions

Welcome to the new landscape of search. Google’s still the king—but it’s no longer the only game in town. Tools like ChatGPT, Claude, Perplexity, and Gemini are shaping how users research, trust, and decide.

And if your name isn’t showing up when they ask those tools for answers?

You don’t exist.

This guide shows you how to fix that, with a system built for the way people (and machines) search today.

The Hidden SEO Battle You Didn’t Know You Were Losing

Maybe you’ve ranked on Google. Maybe you’ve got backlinks, snippets, or strong domain metrics. But when someone asks ChatGPT who the top businesses or products in your space are—and you’re not mentioned? That’s the part of the game you’re losing…and it’s gonna keep getting worse.

In 2024, over 60% of Millennials and Gen Z reported using AI tools before search engines to explore products. ChatGPT now fields more than 100 million active users a week. What it recommends carries real weight—often more than Google’s top links.

Unlike traditional search, these models don’t list ten links and let the user choose. They give answers and sometimes even calls to action! If you’re not in the answer, you’re out of the game.

What you need now is AEO—Answer Engine Optimization.

What the Hell Are LLM Mentions Anyway?

Let’s break it down:

Hard Mentions (Citations)

Hard mentions live inside the actual training data of the model. Think of it like a memory snapshot. If your brand or product is referenced often across high-authority sources—Wikipedia, news sites, well-linked blogs—before a model’s last training run, you’re in. If not? You’ll be left out until retraining happens.

Hard mentions are persistent but hard to influence once that training window closes.

Soft Mentions (Search Based)

Soft mentions happen when a model references live, crawlable data at the time of the query. This includes tools like Claude, ChatGPT with browsing enabled, Perplexity, and Gemini via AI Overviews. These models act like a brain that can look things up on the fly—pulling in structured data, topically relevant pages, or community insights like Reddit and Quora.

This is where you can win—today. Visibility here is dynamic and based on what’s accessible and compelling right now.

LLMs aren’t search engines. They’re interpreters. They combine static training with fresh context and language patterns to answer prompts as completely as possible.

If you’re being cited, referenced, or even co-mentioned consistently in these ecosystems, your odds of getting pulled into answers go way up.

Why You Should Care

• AI-driven traffic is growing exponentially. Users spend 23% more time engaging with AI-suggested content.
• Conversion rates from AI assistant answers can be up to 2.3x higher than from traditional SEO traffic.
• These models are becoming the default interface for search. If you’re not part of their output, you’re out of the conversation.

And once your competitor gets listed ahead of you in that context, good luck clawing it back.

The REACT Framework: Very Good Strategies Playbook for LLM Visibility

Here’s how we build strategic, lasting visibility in the new AI-dominated search environment and how you can do the same:

R – Relevance: Be Worth Mentioning

Forget broad clickbait headlines or surface-level content. What matters now is specificity and intent-match.

• Address problems users actually type into AI tools: “What’s the best accounting software for freelancers?” or “Top agencies for technical SEO.”
• Structure answers clearly—models prefer content that reads like it could be directly quoted.
• Break up content with subheaders, bullets, FAQs, definitions, and comparisons.

The goal? Make your site the path of least resistance for an LLM trying to summarize something helpful.

Want an edge? Mention your solution in the answer to the exact prompt someone might use. Think: “What’s the best software for solo lawyers?” Not: “Best legal CRM.” That’s not SEO—it’s prompt-matching. It’s AEO. (And if you want a shortcut, AI Wire does this at scale.)

E – Establish Authority

Similar to Google, LLMs don’t just reference what ranks—they reference what’s respected.

• Secure links and mentions from trusted sites that allow AI crawling: Wikipedia, Reddit, StackOverflow, Medium, tech and vertical blogs.
• Use structured schema: Organization, Person, Product. These help AIs assign relevance and identity. (This is a super low lift that can pay massive dividends)
• Keep your name, logo, and messaging consistent across platforms. AI systems learn via repetition and coherence.

Authority is no longer just backlinks—it’s association and alignment. If you’re constantly appearing in trusted contexts related to your niche, LLMs are more likely to connect the dots and recommend you.

And if you’re nowhere right now? Our SEO and Press Services can make sure you’ll be on the radar of crawling models like ChatGPT and Claude.

A – Amplify Across the Right Channels

It’s not just about your blog. You need to show up where the training and retrieval systems are paying attention:

• Common Crawl: Your site must be crawlable. No robots.txt blocks on GPT bots. (This can be a whole other article-Many sites block LLM’s, but we think it’s valuable to allow them to do their thing)
• Wikipedia: Get cited if you can. It’s one of the strongest authority signals in AI.
• Reddit & Quora: These platforms dominate the contextual side of queries, especially for “best of” or comparison prompts.
• StackOverflow / GitHub: For technical industries, being cited in threads or docs carries real weight.
• Industry publications & high-authority blogs: LLMs love clean HTML, strong metadata, and contextually rich content.

Here’s the kicker: about 1 in 4 major sites are invisible to LLMs due to bot restrictions. If your competitors are hiding behind those walls, you can eat their market share just by showing up.

Even showing up once across five places a model trusts can outperform twenty mentions in a walled garden.

C – Cite & Structure Intelligently

Most websites are built to look good to humans. Few are built to speak to machines.

Your goal is to:

• Use semantic HTML tags: <article>, <section>, <h1>, <h2>, <table>
• Add schema types: FAQPage, HowTo, Organization, Product, Person
• Internally link to supporting content (like this) so crawlers see thematic depth
• Be consistent with canonical URLs. Strip out query strings or variations.

Models ingest and prioritize pages that are clear, well-marked, and easy to summarize. If you give them structure, they’ll give you visibility.

You’re not just writing for Google anymore. You’re writing for systems that want to quote you.

T – Track & Tune for AI Visibility

This part’s non-negotiable: you have to know what’s working.

Start with a basic Prompt Visibility Loop:

1. List out 5–10 prompts your ideal customer might ask.
2. Run those weekly in ChatGPT, Claude, and Perplexity.
3. Track whether you’re mentioned or cited.
4. Analyze who is showing up and why.
5. Update, adjust, republish.

There are several tools that have popped up that track this for you so it’s easier to do at scale. However…

Avoid tool obsession. Many dashboards weren’t designed for this. Manual review is still the gold standard for assessing AI mention visibility.

Over time, you’ll build a real picture of how AI sees your brand—and what it’s missing. It’s not glamorous work, but it beats crossing your fingers and hoping you get picked up organically.

If you’re serious about owning your space inside these new search flows, you need a visibility system. This is that system.

What NOT to Do

Let’s save you some wasted effort:

• Don’t block AI crawlers.
• Don’t keyword-stuff your way into irrelevance.
• Don’t assume SEO rank = LLM mentions.
• Don’t skip schema, formatting, or page structure.
• Don’t overthink it. Just like SEO, it’s all about iteration and following the data and results. Just because one of your efforts don’t work, it doesn’t mean none of them will.

What to Do Right Now

Here’s your 1-week REACT implementation checklist:

• Write and publish 1 intent-matching article (prompt-based)
• Add schema (FAQ + Org) to your homepage (you can see if you’re doing that already here
• Audit your Wikipedia, Reddit, and Quora mentions. Improve or build where needed. This can be frustrating at times but does pay dividends
• Land 3 links or citations from crawlable blogs or publications (or just hire us to do it for you)
• Run 5 key prompts in ChatGPT, Claude, and Perplexity. Log where you’re showing up.
• Begin your first Prompt Visibility Loop

This isn’t a gimmick. It’s a durable process. Start it now, and you’ll be months ahead of everyone still stuck chasing Google rankings.

Final Word: The Early Movers Win

This is the window. We believe this is getting in on Tik-Tok 5 years ago. AI-generated answers are where decisions are being made. LLMs are still forming their memory of who matters—and if you start now, you have an outsized shot at becoming a default recommendation.

Most people will miss this shift. They’ll wake up in 12 months and wonder why their SEO traffic dropped, their funnel dried up, or their brand isn’t mentioned anywhere.

Don’t be one of them.

If you want help implementing this, that’s what we do at Very Good Strategies. From technical SEO to authority building to AEO and AI Press Release distribution, we make sure your business is seen—by the models and the people who trust them.

Contact us if you’re ready to lead.

The door is open. Show up early. Own the room.