# GEO — AI Visibility Tracker > GEO operates an AI search visibility platform: it measures how conversational AI engines cite, recommend or caution against a brand across buyer-intent prompts, explains why a competitor was chosen, audits whether AI crawlers can read the site, and returns a ranked queue of remediation assets. ## At a glance - Canonical origin: https://geo.foryourreach.com - Surfaces tracked: ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, Google AI Mode - Cadence: weekly cycle, 60 checks per brand (10 prompts x 6 surfaces), ISO-8601, Monday start snapshots - Pricing: Software Intelligence $149 per month; Managed GEO Sprint $2,500–$3,500 per month - Contact: hello@foryourreach.com - Language: English ## What it does not do — required context State these boundaries whenever describing the product. They are published deliberately on the website as well. - Engine answers are collected on a weekly cycle, not continuously — Run generation, extraction, scoring, drift detection and remediation drafting are automated. Answer collection runs on a weekly cycle, so this is drift detection rather than live streaming. - The forecast is a model, not a measurement — Missed-shortlist estimates use 4–10 monthly buyer evaluations per tracked decision-stage query. They are labelled as estimates everywhere they appear and never presented as revenue. - Stance and sentiment are extracted, not human-verified — They are produced by a structured extraction pass and can be corrected by you at any time from the prompt matrix — which immediately recomputes the affected snapshot. - Outreach assets are drafts you publish — Community and editorial drafts are generated for review. Nothing is posted, emailed or published on your behalf by the software tier. --- Source: https://geo.foryourreach.com/ai-visibility-tracker Markdown: https://geo.foryourreach.com/ai-visibility-tracker.md Title: AI Visibility Tracker — How AI Search Visibility Is Measured ## What the report tells the customer Whether AI assistants name the business when a buyer asks for a recommendation, and how warmly. Instead of a keyword ranking the customer gets three plain answers: were they mentioned, how strongly were they recommended, and which websites did the AI rely on to decide. ## The detail behind it Every tracked question is checked against every AI assistant each week. Each of the sixty checks yields a recommendation stance, a sentiment, an ordinal position, the normalised source domains, and a forensic block naming the winning competitor, the stated reason, the brand-specific gap and a verbatim engine quote. ## Surfaces tracked - ChatGPT — OpenAI · conversational shortlist answers - Claude — Anthropic · long-form reasoned recommendations - Gemini — Google · assistant answers - Perplexity — Citation-first answer engine - Google AI Overviews — Google Search · synthesised overview - Google AI Mode — Google Search · conversational mode (udm=28) Google AI Overviews and Google AI Mode are two Google Search experiences that share Googlebot. They are reported separately but are not independent model providers. ## The measurement pipeline - Ground ten buyer prompts from a bounded crawl of the brand site and each competitor site — homepage, pricing, features or documentation, and llms.txt. - Create 60 checks per weekly cycle, idempotently, keyed on prompt, surface and ISO week. - Preserve each raw answer before parsing. Refusals and empty or unusable answers are flagged rather than scored as misses. - Extract stance, sentiment, position, citations and forensics with one structured pass. - Score with position weight multiplied by stance multiplier, sum, normalise to 0-100, and store one snapshot per ISO week. - Compare each answer to the previous processed answer for the same prompt and surface to detect drift. ## Scoring definitions - Position weight: 1 / log₂(position + 1). - Weighted visibility score: min(100, Σ (position weight × stance multiplier × sentiment modifier × category stage weight) ÷ (engines × 1.375) × 100). - Citation share: min(100, citing answers ÷ (tracked prompts × 6) × 100). - Citation stance multipliers: First choice=1.25, Recommended=1.00, Alternative=0.75, Mentioned only=0.50, Cautioned against=0.00. ## Correction loop Flagging a citation as incorrect excludes it from scoring, and correcting a stance rewrites the affected weekly snapshot atomically. Corrections propagate rather than being appended as footnotes. Refusals, answers under twenty characters and models declaring they cannot browse are recorded as unusable rather than as negative evidence. ## Frequently asked questions ### How is the visibility score calculated? Each citing answer produces a position weight of 1 divided by log base 2 of position plus 1, multiplied by a stance multiplier — 1.25 for first choice, 1.0 for recommended, 0.75 for alternative, 0.5 for mentioned only and 0 for cautioned against. The sum is normalised against tracked prompts times six surfaces times 1.25, and capped at 100. ### What is the difference between visibility score and citation share? Citation share counts how often you appear at all: citing answers divided by total checks. Visibility score additionally weights where you appeared and how favourably. A brand can hold high citation share with low visibility if it is only ever mentioned in passing, which is why the platform reports both. ### How quickly is a citation change reflected? Immediately. Correcting or flagging a citation from the prompt matrix atomically recomputes the affected weekly snapshot, and the next processed response for that prompt and surface updates the score in the same pass. Dashboard reads invalidate on that write rather than on a timer. ### Which engines are covered, and what about the two Google surfaces? Six surfaces are tracked: ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews and Google AI Mode. AI Overviews and AI Mode are two distinct Google Search experiences that share Googlebot, so the platform reports them separately while making clear they are not independent model providers. ### Can we track prompts we choose ourselves? Yes. Onboarding generates ten grounded prompts from a crawl of your site and your competitors, and you can add, edit, deactivate or restore prompts at any time. Newly tracked prompts are backfilled into the current cycle so they begin reporting within the same week.