[ Methodology ]
How XanLens works
GEO is probabilistic, not deterministic. Ask AI the same question 10 times, get 10 different answers. We measure how AI engines see your brand — not how Google ranks your website. Different problem, different methodology.
What we measure
When someone asks ChatGPT, Gemini, or Claude for a recommendation in your space, does your brand come up? That's what we test.
Brand recognition
Does the AI know your brand exists? Can it describe what you do accurately, or does it hallucinate features you don't have?
Category visibility
When users ask for "the best tools" in your industry, does your brand make the list? Or do only your competitors show up?
Recommendation strength
Being mentioned is one thing. Being recommended with positive sentiment and accurate details is another. We measure both.
Competitive positioning
How does your brand stack up against competitors in AI responses? Who gets mentioned first? Who gets recommended more often?
How we score
The GEO score (0–100) reflects your real visibility across AI engines. We test multiple query types that mirror how real users interact with AI — from direct brand searches to open-ended discovery questions.
Multi-prompt testing
We don't send one query and call it a day. Each audit runs dozens of prompts across different intent types to build a complete picture of your AI visibility. A brand might be known by name but never recommended — our prompts catch that.
Coverage matters
Appearing in one type of query isn't enough. We measure how consistently your brand shows up across all the ways users discover products through AI. High quality in a narrow slice still means low overall visibility.
Hallucination detection
AI engines sometimes invent facts about brands — wrong pricing, fake features, incorrect descriptions. We detect these hallucinations per-response and flag them. False visibility is worse than no visibility.
Deterministic results
Every query runs at temperature 0 — no randomness. Run the same audit twice, get the same results. This makes scores reliable for tracking improvement over time.
GEO score vs Web Presence
We separate AI visibility from web visibility because they measure different things. A brand can rank #1 on Google and still be invisible to ChatGPT.
GEO Score
How AI engines respond when asked about your brand or category. Direct measurement of AI visibility.
Web Presence
How discoverable your brand is on the traditional web — the source material AI engines learn from and cite.
Grade scale
Most brands score below 40. That is not a bug — it is the current state of AI visibility. Most businesses have not optimized for how AI engines discover and recommend products.
Strong AI visibility. Consistently mentioned and recommended across query types.
Good visibility with gaps. Known to AI but missing from some query categories.
Moderate. AI has some awareness but rarely recommends. Competitors dominate.
Weak. Limited or outdated knowledge. Mentioned only in direct queries, if at all.
Invisible. AI engines don't know the brand or actively recommend competitors instead.
AI engines
Each AI engine processes information differently. What Gemini knows, ChatGPT might not. Multi-engine coverage gives you the full picture.
Gemini
Live
ChatGPT
Live
Claude
Live
Grok
Live
Perplexity
Live
DeepSeek
Live
Meta AI
Live