Current prospective panel
ChatGPT Search, Google AI Mode, Gemini, Claude web search, Perplexity and Grok web search
The governed panel is frozen before collection so the tested surfaces remain explicit and comparable within the bounded protocol.
GSA AI Discoverability Report
See what AI is telling patients about your practice.
An evidence-based diagnostic of how AI search systems find, understand, and represent your optometry practice.
GSA tests a governed panel of major AI and search surfaces across real patient questions—then gives you a concise, evidence-backed report showing where the answers are accurate, inconsistent, stale or simply missing.
Complimentary · Manually reviewed by GSA · Limited availability
6core AI/search surfaces in the prospective panel
Frozenprompts and run rules before collection
Matchedretests for material findings
Boundedevidence and limitations preserved
Your practice next
See what your practice looks like in AI search.
Get a complimentary AI Discoverability Report showing how major AI/search systems currently surface, omit or misunderstand your practice.
Want to see the deliverable first? Open the real Park Slope Eye sample beside it. The sample is based on dated tested responses and is not a universal score or a promise of future AI behavior.
Read the Park Slope Eye case study →
The test
Real patient questions. Fresh sessions. Evidence preserved.
The report is designed to reveal a practical problem: being visible to AI is not the same as being understood accurately enough for a patient to act.
Controlled protocol
Prompts and run rules are frozen before collection
GSA preserves the prompt, product surface, visible mode or model, session state, response evidence, visible sources, claims, accuracy and limitations for each governed run.
Source-aware
We preserve what the systems visibly relied on
Current first-party sources, older practice pages, third-party sources and unverified source layers are kept distinct.
Evidence-bounded
No universal grade. No ranking promise.
The report describes the responses actually tested on a dated snapshot. It does not pretend one run measures all AI behavior.
Why it matters
A patient can find you—and still receive the wrong answer.
GSA’s Park Slope Eye pilot frequently found the practice, but still documented wrong entity/location information, false service availability, stale first-party policy use and inconsistent answers across repeated branded tests.
The real question
It is not simply whether AI can find the practice. It is whether what a patient is told is accurate enough to act on.
Those are not abstract SEO metrics. They are facts that can change where a patient goes, which doctor they choose, whether they think their insurance works, or what they expect before booking.
What you receive
A concise diagnostic—not a generic AI score.
The deliverable is built to be useful to a practice owner, not to manufacture a flashy grade that overstates what one testing window can prove.
01
Findings
What patients were actually told
Candidate retrieval, factual accuracy, critical errors and branded-answer consistency—translated into plain English.
02
Sources
Where stale or conflicting information appears
Visible first-party and third-party sources are separated so you can see whether outdated information is still influencing the environment.
03
Remediation path
What GSA would investigate—and how we would prove the work
Each material finding is mapped to a bounded remediation-and-proof record: what the observation suggests, what controlled sources we would investigate or correct, how propagation is verified, and which matched AI behavior is rerun.
If GSA does the work
Finding the problem is only the first step.
Each material finding is mapped to a bounded remediation-and-proof record. GSA does not claim to “fix the AI.” We improve the information environment the systems rely on, then measure what the AI actually does.
01CorrectAddress the controlled sources and conflicts tied to the finding.
02VerifyConfirm the intended correction is actually live.
03PropagateCheck whether corrected information is publicly visible where observable.
04RetestRun the same prompt on the same surface under the same rules.
05WatchRepeat later rather than declaring victory from one favorable answer.
Why stewardship matters
The website is the authoritative foundation. Launch does not instantly erase stale sources, reconcile every outside entity signal, or determine what an AI system retrieves. That is why GSA verifies propagation and retests the same behavior after the correction.
Learn why AI visibility and AI accuracy are different →The GSA standard
Implementation evidence, propagation evidence, matched AI testing, durability testing and business outcomes are reported as separate evidence layers. The same prompt and same surface anchor matched retesting. A technical change or one favorable answer is not treated as proof of permanent model behavior, conversion, revenue or ROI.
What happens after the report
The report diagnoses. GSA stewardship does the work.
AI Discoverability Report
Shows the problem clearly
Evidence-backed findings, source observations, and the bounded remediation-and-proof path for each material issue.
Managed GSA engagement
Corrects sources, verifies and monitors
Practice Truth reconciliation, source/entity cleanup, controlled-source improvements, propagation checks, matched retesting and ongoing monitoring. A correct website is the authoritative foundation; stewardship addresses the stale sources, entity conflicts, retrieval behavior and propagation that can persist after launch. Managed clients see that ongoing work translated through Practice Pulse, GSA’s monthly insights report: what we saw, why it matters and what happens next.
Complimentary diagnostic · Limited capacity
Would you be comfortable with what AI is telling patients about your practice?
If you are not sure, GSA can run the controlled diagnostic and show you the evidence.
Get Your Complimentary AI Discoverability Report
No aggregate grade. No ranking guarantee. No obligation to hire GSA for remediation.