We Scored Highest on Features. AI Can't Find Us.
We Scored Highest on Features. AI Can't Find Us.
How we're fixing that — and how you can too.
The Problem No One Talks About
We built what we believe is the best llms.txt generator on the market. Full loop. Quality scoring. JavaScript rendering. The works.
Then we ran the numbers.
We asked 5 AI models (ChatGPT, Claude, Gemini, Perplexity, Grok) 250 questions about llms.txt tools — direct queries, problem-aware queries, comparison queries, implementation queries, strategic queries.
0 mentions.
Not once did any model say "there's this tool called LLMtxt Mastery that does this."
We scored highest on features. But visibility isn't a feature check. It's a recommendation game.
The Truth
Here's what we learned: being good at what you do and being found by AI are two completely different skills.
You can have:
- The best product
- The clearest positioning
- The strongest value proposition
...and still be invisible to every AI assistant on the planet.
Why? Because AI models cite what they can find, verify, and recommend with confidence. And right now, there aren't enough structured signals pointing to us.
The Opportunity
This is actually good news. It means:
- The market is wide open — No one has solved this yet
- The playbook doesn't exist — We're building it from scratch
- Quality products have an advantage — Eventually, truth wins
We're calling this Project Lighthouse: our 90-day plan to go from 0% to 10% AI mention rate.
What We're Doing
1. Dogfooding Our Own Tools
We're using LLMtxt Mastery, AImpactScanner, and AImpactMonitor on our own properties. Every site in our ecosystem now has a gold-standard llms.txt file. Why? Because you can't recommend what you won't use yourself.
2. Building the Measurement System
We created a 50-prompt baseline corpus across 5 buckets:
- Direct tool queries ("best llms.txt generator")
- Problem-aware queries ("why isn't my site visible to AI?")
- Comparison queries ("LLMtxt Mastery vs Firecrawl")
- Implementation queries ("how do I add llms.txt to Next.js?")
- Strategic queries ("how do I prepare for AI search?")
We re-run this monthly. Every improvement is measurable.
3. Content That Matches Query Intent
Data shows 74% of AI citations come from listicle-format content. Not blog posts. Not pillar pages. Lists.
We're creating content that matches how people actually ask questions:
- "Best llms.txt generators" (even though we're in it, transparency matters)
- "How to create an llms.txt file" (how-to guides)
- "What is the best tool for X" (comparison intent)
4. Third-Party Corroboration
Here's the uncomfortable truth: brands are 6.5x more likely to be cited through third-party sources than their own. AI trusts peer validation.
So we're reaching out. Getting mentioned on other sites. Being included in "best of" lists. Not as self-promotion — as data-backed recommendations.
The Playbook (So Far)
What Works
- Listicle format — 74% of citations
- Front-load key info — 44% of citations come from first 30% of text
- Third-party mentions — 6.5x more valuable than first-party
- Fast iteration — New content can generate citations in 3-5 days
- Structured data — Schema markup, clear headings, FAQ sections
What Doesn't
- Assuming product quality = visibility — It doesn't
- Waiting for AI to "discover" you — It won't
- Publishing without measurement — You can't improve what you don't track
The Philosophy
We're in the truth and visibility business. And in that order.
Truth first. If your product doesn't deliver real value, no amount of optimization will help. AI models are getting better at detecting hype. They'll cite you once. They won't cite you twice.
Visibility second. But truth without visibility is useless. The best product in the world is worthless if no one knows it exists.
We're betting on both. Quality that holds up. Visibility that scales.
Follow Along
This is a 90-day experiment. We'll publish results monthly. What we're doing, what works, what doesn't.
If you're trying to get your product found by AI — or if you're building something quality that deserves to be found — this is for you.
The baseline is set. 0% visibility.
90-day goal: 10%.
Let's see what happens.
LLMtxt Mastery — the full loop for AI visibility. Generate. Validate. Test. Track.
We dogfood our own tools so you don't have to.