Your competitor just got recommended by ChatGPT to 400M users last month. You weren't even mentioned.
Welcome to AI SEO warfare.
I've helped brands rank on search engines for 15+ years. But lately? Every founder asks me the same urgent question: "How do we rank in ChatGPT?" The interface has changed. The game has changed. And there's an entirely new playbook you need to master before your competitors do.
Today we’re breaking down:
TL;DR (Swipe This If You’re Busy)
"AI doesn't rank you—it remembers you. And that memory comes from Wikipedia blurbs, not your homepage." [Click to Share]
📊 Some brands are already seeing 10x+ traffic from AI referrals—one case study showed 2,300% growth by aligning content with AI Overviews.
🧠 AI Overviews are Google's new top-of-page answers—generated by Gemini, shown before ads, and often without showing your actual site.
📈 You can track your brand mentions in AI Overviews using tools like Ahrefs Brand Radar or Dejan's AI Rank Tracker.
✅ Your new playbook: summarize smart, speak conversationally, and seed content on trusted surfaces—Wikipedia, Reddit, and anywhere AI scrapes first.
📈 AI Search is exploding — ChatGPT, Gemini, and Perplexity are driving serious traffic. Canva's Head of SEO reports a 300% YoY surge.
💡 Gemini's "memory" is programmable. With dynamic thresholds for grounding, your brand description can flip on/off in real-time.
📉 Clicks convert worse—20–30% drop vs classic SEO—but volume is rising, and influence is compounding.
✅ New playbook: Forget keywords. Engineer summaries. Seed citations. Feed the machine what you want it to remember.
The AI Search Flood Is Here
Most marketers have heard of 'zero-click search' where Google answers the question right on the results page, and users never visit your site. That's been a problem for years.
But AI search? It's far more dangerous.

It doesn't just answer—it decides. ChatGPT and Gemini pick the brand, summarize the value, and move the user along. If you're not in that answer box, you're not even in the conversation.
ChatGPT Search went live on October 31, 2024, and by February 2025 it rolled out to everyone. In just a few months, it's changed how hundreds of millions find information—and make buying decisions about your products.
The Numbers Don’t Lie:
3.7 billion monthly visits
400+ million weekly active users → That's more than Bing and DuckDuckGo combined (Source: Similarweb, Keywords Everywhere)
Perplexity's catching up fast: 155M visits in March 2025, with over 100M queries per week (Source: Exploding Topics)
And the traffic converts. Canva's Head of SEO confirmed the users coming from these models aren't just curious—they're clicking, designing, and buying.

There are exactly three places AI scrapes first when forming opinions about your brand. Most companies completely ignore the second one. Keep reading to find out what they are.
WHO WINS IN AI SEARCH:
Brands with Wikipedia presence
Companies with Reddit megathreads & big social presences
Businesses cited by journalists
WHO DISAPPEARS:
Brands relying on traditional SEO only
Companies with stale social presence
Businesses with generic product descriptions
📣 TAKEAWAY: Forget the "0-click panic." Clicks are still happening. They're just coming from chat boxes instead of search results.
What AI Really "Knows" About Your Brand (Spoiler: It's Probably Wrong)
Here’s the part nobody tells you:
Gemini, ChatGPT, and Perplexity don't just "search."
They summarize, often using scraps from Reddit threads, old press articles, and scraped blurbs from the SERP.
When was the last time you checked how your brand appears in these models? You might be horrified by what you find.
Case in Point:
Gemini’s response JSON includes a field called <additional_info>—an AI-generated brand blurb stitched together from the search index.
Dan Petrovic (Dejan) ran a teardown. The results? Wildly inaccurate.
One brand was described using copy from a 10-year-old blog post and a Reddit comment—neither of which were on the brand's site.
🧠 Rule: If you're not explicitly feeding the machine, you're implicitly trusting its improv skills.

✅ Action Steps:
Search your top 3 brand terms in Gemini, ChatGPT, and Perplexity
Screenshot the summaries you get
Highlight inaccuracies → Then create micro-content that corrects them:
50-word FAQs
Updated press snippets
New blurbs on .edu sites or trusted third-party citations
Seed these surfaces where models crawl first:
Wikipedia
Reddit
LinkedIn pulse
Medium
Quora
Major media
Schema-rich pages
What happens if you don't?
Your competitors or random disgruntled people on the internet will shape your brand narrative for you. This is your life’s work. You DO NOT want to be defined by some rando online.
Learn AI in 5 minutes a day
This is the easiest way for a busy person wanting to learn AI in as little time as possible:
Sign up for The Rundown AI newsletter
They send you 5-minute email updates on the latest AI news and how to use it
You learn how to become 2x more productive by leveraging AI
From Rank Tracking to Memory Engineering
Old SEO tracked positions. Modern visibility tracks influence inside the model's mind.
If you want to show up consistently in AI answers, you need to understand the two layers of LLM memory:
Memory Layer | What It Is | How to Influence |
|---|---|---|
Core Memory | Facts baked into the pretrained model | Earn PR mentions, Wikipedia citations, schema-rich content, brand mentions across social, sustained visibility on platforms like X, LinkedIn, YouTube |
Contextual Grounding | Real-time snippets retrieved at answer time | Optimize titles/meta, get cited in top-ranking pages, compress key facts into answer-ready chunks |
Gemini’s Secret Lever:
Its retrieval engine grounds live search results only when confidence dips below a certain threshold. Default: 0.3. (Source: Google Developers Blog)
What does this mean for your brand? You can subtly influence whether Gemini pulls new info... or defaults to stale summaries burned into its brain.

It’s like SEO meets behavioral psychology. You're nudging AI when to think fresh and when to autopilot.
The 2025 AI SEO Optimization Stack
You don’t need to throw out your SEO playbook.
But you do need to layer in new tools built for AI-first visibility.
Traditional | 2025 Add-Ons |
Technical SEO hygiene |
|
Schema & internal links | Brand mentions on Social, Reddit, Quora, YouTube, Wikipedia |
Rank tracking | AI-rank visibility probes like airank.dejan.ai |
Content refresh | Snippet engineering—tight, citation-rich 50-word answer blocks designed for LLM pickup |
30-Day Brand Memory Implementation Plan
Week 1: Audit current AI visibility
Week 2: Create core brand snippets
Week 3: Seed top-priority platforms
Week 4: Measure & adjust visibility
Pro Tips to Stack Wins:
Use llms.txt to control what gets into the model's memory. Don't leave it to random crawl behavior
Embed facts that models love: dates, names, credentials, locations, use cases
Test visibility in different models weekly. Use prompts like "What is [brand]?" and "Best tools for [category]"
Engineer first impressions—models often surface the most compressed, repeat-cited, and multi-source answers
Treat Reddit, Quora, Medium, and LinkedIn (pulse is good here) like high-authority real estate. LLMs crawl them constantly. Be there.
Write clearly and directly. AI favors crisp, confident language.
Use one H1, clear H2s for sections, and H3s for subpoints.
Add a TL;DR or Key Takeaways section at the top of major content.
Answer like a chatbot: think full-sentence, natural language responses.
Include 'People Also Ask' style FAQs to match real conversational search.
This Feels Like 2005 SEO Again (That's Good News For You)
If you've ever wished you were earlier to Google... this is your shot.
Because this moment? It rhymes with 2005.
Back then, the game was decoding PageRank quirks and backlink tricks before Google patched the loopholes.
Now, the edge lives in reverse-engineering LLM behavior. Figuring out how Gemini chunks SERP data, how ChatGPT dedupes sources, how Perplexity chooses "related links."
The rules are foggy. The results are erratic. That's your opportunity.
Arbitrage Alert:
The brands that learn to plant the right "memory atoms" now will compound authority while everyone else is still fiddling with H1 tags.
Every Reddit thread that cites you = potential AI anchor
Every Wikipedia edit = a neuron in the machine's brain
Every snippet that summarizes you = a candidate for LLM recall
This is pre-algorithmic ossification. You've got 12–18 months before the window tightens and Google bakes this behavior into its broader search ecosystem.
Remember how you asked ChatGPT about your brand at the beginning of this article? Go check that screenshot again. That's what millions of potential customers are seeing right now.
The question isn't whether AI will describe your brand. It's whether you'll control what it says.
The model is learning. Are you teaching it?
If this newsletter lit a fire under you, forward it to one person who would benefit.
🎁 Get rewarded for sharing! My team grew NEW accounts to over 50 million views in just a few months. We made an AI viral hook generator so you can follow the same hook frameworks that we do.
Invite your friends to join, and when they sign up, you’ll unlock our AI Viral Hook Generator—the ultimate tool for creating scroll-stopping content.

{{rp_personalized_text}}
Copy and paste this link: {{rp_refer_url}}
Let's turn those random AI mentions into an engineered narrative that drives traffic.
Smart brands don't leave their AI visibility to chance. They systematically plant memory hooks, seed strategic citations, and build algorithmic gravity that pulls customers in.



