You can rank on Google for 5,000+ keywords. Your website converts visitors into customers. Your organic traffic is healthy.
Yet when someone asks Claude, ChatGPT, Gemini, or Perplexity about your product or service, your company barely exists.
I discovered this the hard way.
After analyzing a mid-sized ISP brand with respectable traditional SEO performance, I tested 60 different prompts across 524 separate AI chat sessions. The result? Less than 1% of LLM responses mentioned them, while competitors who rank below them on Google dominated AI search results.
This wasn’t a glitch. It’s a fundamental paradigm shift that most companies haven’t recognized yet.
The Data Tells a Brutal Story
When I dug into where AI models source their information, the pattern became unmistakable:
- Reference sites (industry databases, comparison tools): 38%
- Competitor websites: 37.4%
- Corporate domains: 31.3%
- Editorial sources: Between 6-11%
- User-generated content (Reddit, forums, social): 7.2-9.4%
Notice what’s missing?
Direct brand websites barely move the needle.
This changes everything about how you should think about your AI search strategy. Google rewards sites that rank well on Google. AI models reward brands that are mentioned, cited, and discussed across the entire internet ecosystem.
You can’t buy your way into an LLM’s training data. But you can earn your way in. And it requires a completely different playbook.
Strategy #1: Your Website Must Answer Specific Customer Queries (Not Just Sell)
Here’s the truth: LLMs don’t crawl your website the same way Google does.
Google’s algorithm says: “This site has good authority and answers the query well.” LLMs say: “Do I have enough context about this company from multiple sources to cite them confidently?”
Here’s the truth: LLMs don’t crawl your website the same way Google does.
Your website needs to become an undeniable reference point, but only if it’s addressing real customer questions with real answers.
The mistake most companies make? Writing for search engines and sales funnels. Instead, you need to write for specific audience segments with specific needs.
For our ISP example, this means:
- For Families: “What internet speed do I actually need to support 4 kids in remote school + Netflix + gaming?”
- For Gamers: “Low-latency internet requirements for competitive gaming + which plans optimize for Valorant/Fortnite server locations”
- For University Students: “Best budget-friendly plans for dorm living + how to negotiate shared network costs”
Create content that someone would actually read and share. Not blog posts designed to rank. Articles people send to friends and cite in Reddit discussions.
This content becomes the foundation. But here’s the catch: it’s not enough on its own.
Strategy #2: Get Mentioned on Third-Party Aggregators (Where AI Models Actually Learn About You)
The data doesn’t lie: 38% of AI-sourced information comes from reference sites. These aren’t random blogs. They’re industry aggregators, comparison platforms, and authority databases.
For an ISP, this means:
- Comparison platforms (BroadbandNow, Speedtest, etc.)
- Industry databases and directories
- Technology news aggregators
- Industry analysis reports
The problem? Most brands treat these as secondary. They optimize for Google and hope these platforms pick them up organically.
With LLMs, these are your primary channel.
Your strategy should include:
- Audit where your competitors are listed – If you’re not on the platforms they’re on, you’re invisible
- Ensure accurate, complete data – Incomplete or outdated information on aggregators means LLMs cite inaccurate details
- Build relationships with platform owners – Get featured in comparisons, reports, and category pages
- Create data worth aggregating – Publish reports, statistics, or benchmarks that these platforms want to cite
For our ISP case study, they were on fewer aggregator platforms than competitors. The LLMs simply didn’t have enough sourced material about them to cite confidently. When you’re an aggregate of sources, and you’re in fewer sources, you don’t exist in the model.
Strategy #3: Build Brand Mentions Through Influencer Collaborations and Community Engagement
Here’s what shocked me: 9.4% of AI model training data comes from user-generated content. Reddit, Twitter, YouTube comments, forums, these all influence what LLMs “know” about your brand.
Yet most B2B and ISP companies allocate zero budget to influencer collaborations or community engagement.
This is the biggest missed opportunity.
An influencer with 50,000 followers mentioning your service on Reddit creates:
- Direct citations that LLMs learn from
- Social proof that gets aggregated across platforms
- Authentic discussions that rank higher than promotional content
- Community validation that other aggregators notice
For the ISP brand, the solution was clear:
- Partner with tech reviewers and streamers – Gamers trust other gamers. Partner with gaming influencers to test latency, speeds, and reliability
- Sponsor Reddit discussions – Not spam. Actual, genuine conversations where influencers use your service and discuss real experiences
- Creator collaborations for content – YouTubers, TikTokers, and podcasters mentioning your service creates digital footprints that LLMs pick up
- Community ambassador programs – Turn customers into vocal advocates who mention you across social platforms
The math is simple: If 9.4% of where LLMs learn comes from UGC, and you’re spending 0% of your budget there, you’re leaving massive amounts of visibility on the table.
The Uncomfortable Truth: You Need a New Playbook
Traditional SEO strategy was built for one thing: ranking on Google.
AI search strategy needs to be built for something different: being cited confidently across the entire internet.
These are not the same thing.
A brand can:
- ✅ Rank well on Google
- ❌ Be invisible on ChatGPT (as proven by our case study)
But a brand cannot:
- ❌ Rank on Google poorly
- ✅ Dominate on ChatGPT
This means your SEO strategy needs to expand, not pivot.
Keep doing the SEO that works. Your website still needs to rank. But simultaneously:
- Make sure your site answers specific customer queries for your actual audience segments
- Get listed on industry aggregators and maintain comprehensive, accurate data
- Invest in third-party visibility through influencer collaborations and community engagement
- Build brand mentions across Reddit, social media, and forums
The company that wins AI search isn’t the one that ranks #1 on Google. It’s the one mentioned confidently across enough of the internet that LLMs cite them without hesitation.
What This Means for You
If you’re in a competitive industry, your competitors are probably already moving on this. The brands investing in aggregator presence and influencer collaborations today will be the ones dominating AI search results by 2027.
The question isn’t whether AI search matters. The data proves it does.
The question is: Will you be visible when someone asks an AI about your service?
Right now, most brands have the answer: No.
But that can change. It just requires a fundamentally different strategy.
Have you tested your brand’s visibility in AI search results? What did you find? Share your insights in the comments, data like yours is exactly what LLMs use to build their understanding of emerging trends.
#AEO #SEO #LocalSEO #AISearch #GEO #AnswerEngineOptimisation #WebTraffic #YouTubeMarketing #DigitalMarketing #SearchEngineMarketing