It’s called LLM SEO (or LEO). And it’s already sending 100,000s of users to your competitors.
Here's how you can get ahead:
What is LEO?
It’s about optimising your brand and content to become the answer in LLM platforms like ChatGPT, Perplexity, Claude, and now most crucially, Google’s AI Mode.
This isn’t just a new traffic source.
It’s a whole new search ecosystem.
We’re actively managing SEO for 30+ sites.
And we’re already seeing 5–15% of their monthly traffic coming from LLM platforms (and growing MoM).
Most of the traffic (around 95%) is from ChatGPT, with the rest from Perplexity, Claude, and others.
This isn’t theoretical, it’s happening now.
Google is also rolling out AI Mode as we speak.
AI Mode adoption is growing 4% DoD. At this rate, it could hit 100% adoption by mid-September.
But I expect Google to eventually make it the default regardless.
So if you’re not part of these AI answers, you’re not just missing out on a few clicks…
You’re missing out on the future of search.
And while traditional best practices still matter, the argument that LEO is “just SEO” is overly simplistic and naive.
LEO is about relevance engineering, not just rankings.
Here’s how it actually works:
1. Brand mentions
LLMs verify trust by how often your brand is mentioned elsewhere.
Tally is getting 1,000s of new customers every month from their now #1 referral source, ChatGPT.
“We’ve invested in showing up [in forums, Reddit posts, blog mentions, and authentic UGC] by sharing what we learn, answering questions, and being human.”
2. Brand equity
LLMs personalise answers based on brand trust and sentiment.
- Monitor reviews
- Respond to feedback
- Build a positive reputation
A brand that’s trusted by users is trusted by AI.
Tip:
Use a tool like to track your visibility, understand the sentiment, compare to your competitors, and get actionable insights to improve your AI mentions.
(Launching this week only to waitlist signups) Mentions.so
3. Being the source
LLMs favour authoritative, data-rich content.
Write and publish like a reference source, not a blogger.
- Update content regularly
- Keep a neutral, authoritative tone
- Avoid excessive personal anecdotes
- Back up with credible sources and data
- Write factual, unbiased, and data-rich content
Tip:
Most LLM platforms cite their sources.
Use this data to reverse-engineer how competing content and brands are getting featured in the answers instead of you.
4. Passage-level SEO
LLMs pull passages, not pages.
Structure content so each paragraph can stand alone as the best snippet of information for a question.
- Include concise answers where relevant
- Write short, clear and focused paragraphs
- Use subheadings to segment topics logically
- Avoid burying key insights in long-winded text
- Include structured data (schema) where possible
- Make each passage easy to extract and cite by AI
Example:
5. Covering the conversation
LLMs “fan out” queries into sub-questions.
Don’t just answer your main keyword, cover all related questions.
- Research sub-questions users ask
- Add FAQs to address related questions
- Use supporting pages for each sub-question
- Link to related topic clusters with internal links
- Monitor forums and socials for emerging questions
- Structure your site so LLMs can easily find answers
Tip:
Question research is the new keyword research.
Tools like Semrush or AlsoAsked help you find related questions your audience is asking.
6. Crawlability
If LLMs can’t find your content, it’s invisible.
- Allow crawlers in robots.txt
- Check Cloudflare settings
- Submit XML sitemaps
- Monitor page speed
- Use clean HTML
Tip:
Check server logs to see how search engines and AI crawlers (e.g. GPTBot or BingBot) are accessing your site.
Start with an audit:
- Your current visibility across all LLMs
- Prompt and content gaps vs your competitors
- Technical issues preventing AI from citing your site
Use a tool like to find this information (Launching this week only to waitlist signups).
Then create your 90-day optimisation roadmap.Mentions.so
That's a wrap!
If you enjoyed this thread:
1. Follow me @jakezward for more 2. Repost below to share it with others
I track Google search trends so you don’t have to.
7 content trends I’ve found this week:
1. Content Repurposing
One piece of content can generate dozens. AI tools are making it easy to turn blogs into threads, podcasts into clips, webinars into reels. Create, then multiply.
What it means for you:
Founders: Build or integrate repurposing engines for niche formats, e.g. SaaS explainers to LinkedIn posts, or long-form YouTube to TikTok. Speed and accuracy are the moat.
Marketers: Stop creating from scratch. Build workflows where one asset fuels ten. Show ROI by mapping how a single piece scales across channels and stages.
Creators: This is true leverage. Try: “How I turned one video into 12 posts,” or “My AI content flywheel.” Teach your system and show results. People are hungry for scale without burnout.
2. AI Story Generator
AI tools are getting better at creating full narratives: stories with plot, pacing, character arcs, and emotional beats.
What started as gimmicky short tales is evolving into scripts, books, and branded storytelling at scale.
Marketing is getting the same acceleration as Vibe Coding (e.g. Replit, Bolt, Lovable). What once took TEAMS now takes ONE smart Vibe Marketer and a stack of AI agents.
What it means for you:
Founders: Entire marketing departments are being compressed into lean, automated systems. A single person can now ideate, test, and launch campaigns in hours (not weeks).
Marketers: Vibe marketing isn’t about funnels. It’s about fast feedback loops. Real-time testing. Hundreds of angles launched via agents. The right stack lets you outpace entire teams.
Creators: This is a movement. Document your workflows. Share tool stacks. Break down how one person outperforms an agency. The space is exploding, and people want a guide.
2. Programmatic SEO
Using automation to create 100s or 1,000s of long-tail pages based on templates and data.
Often targeting high-intent, low-competition keywords at scale (e.g. location pages or glossaries).
Confession: I’m obsessed with tracking search trends.
7 AI trends I’ve found this week:
1. AI Girlfriend
People are turning to AI for connection. AI is filling real emotional gaps; companionship, validation, even intimacy. It’s weird, growing fast, and increasingly normalised.
What it means for you:
Founders: AI companionship is an emotional use case that’s outpacing productivity. Expect spin-offs: AI best friends, AI mentors, AI therapists.
Marketers: Even if you’re not in the AI space, this is a signal: emotional resonance matters. People aren’t just searching for tools, they’re searching for connection.
Creators: This is attention-worthy content. Think: “I spent 24 hours with an AI girlfriend.” Cultural commentary, humour, or deep dives will all perform well.
2. AI SEO
The blending of traditional SEO and LLM SEO (or LEO) to show up in AI-generated answers like Google’s AI Overviews, ChatGPT, and Perplexity responses.
It’s about optimising your brand and content to become the answer in LLM platforms like ChatGPT, Perplexity, Claude, and now most crucially, Google’s AI Mode.
This isn’t just a new traffic source.
It’s a whole new search ecosystem.
We’re actively managing SEO for 30+ sites.
And we’re already seeing 5–15% of their monthly traffic coming from LLM platforms (and growing MoM).
Most of the traffic (around 95%) is from ChatGPT, with the rest from Perplexity, Claude, and others.