The method
Diagnose from the foundation up.
Visibility findings are the most exciting layer, but they mean nothing if your site is blocking crawlers. So we work in strict order, access first, and every finding comes with the evidence behind it.
The six layers
Access
Can AI crawlers reach your site at all?
We check every relevant crawler. Training, search, and user-triggered fetchers are separate and independently controllable. A blanket block from the scraping panic can silently remove you from AI answers. The fix is usually small, fast, and high-value.
Understandability
Can engines tell who you are and what you do?
Clarity of naming, plain-language descriptions of what you sell, and structured data that matches what’s visible on the page. If an engine can’t summarize you accurately, it won’t recommend you confidently.
Trust
Is there evidence you’re credible?
About pages, real bios, credentials, and case studies: the trust architecture engines and buyers both look for. We build what’s missing rather than gesture at it.
Citation-worthiness
Can engines quote your content?
Structure and evidence, not volume: heading hierarchy, chunking, quotations from credible sources, statistics, and cited claims. The research is clear that this beats keyword stuffing and content churn.
Conversion
When visitors arrive, do they convert?
Visibility that lands on a leaking bucket is wasted. We check the conversion path so improved demand actually turns into inquiries and revenue.
AI Visibility
Do engines actually name you?
Active, cross-engine probing of the questions your buyers ask. No engine offers this report, and existing analytics can’t separate it out. We baseline your share of voice against named competitors and re-measure over time.
Layers and disciplines
Six layers, seven disciplines: two views of the same diagnostic. The layers are the order we work in; the disciplines are what the report grades. Four map one to one. The three that don’t, Foundation, SEO and GEO, all land in the first two layers.
Foundation · GEO (crawler permissions)
SEO · GEO (structured data, semantic markup)
Authority & Trust
AEO
SXO
LLMO
The report grades all seven. The layers set the order we work in; the sample below leads with GEO rather than repeating that sequence.
What the report covers
Seven disciplines. One report.
The layers set the diagnostic order; the report grades the seven disciplines they map to, each with its own score, findings, and one concrete recommendation. Page through what a The Citation Lab diagnostic delivers.
AI Overviews54
Perplexity54
Copilot45
Claude44
OpenAI44
Gemini30The evidence base
This is studied, not invented.
Visibility improvement, up to, reported in the foundational GEO benchmark from deliberate optimization. Quotations, statistics, and cited sources were the strongest tested methods.
Aggarwal et al., GEO, KDD 2024 →Citation-rate lift from structural changes alone (heading hierarchy, chunking, emphasis) with content held constant, measured across six engines in a 2026 study.
Yu et al., GEO-SFE, March 2026 →Guarantees anyone can honestly make. Engines diverge, change, and published research shows tactics tuned to current behavior degrade, which is why measurement repeats.
Every figure is a research result under specific conditions, never a guarantee.
The report artifact
Share of voice, measured.
We probe the questions your buyers actually ask, across the major AI engines, and count who gets named. This is the table from a visibility report: the one that shows where you really stand.
Google AI Overviews
Perplexity
Copilot
Claude
OpenAI
Gemini