ChatGPT
Capture the dated returned answer, brand mention state, prominence and observed citation evidence for tracked questions.
Measurement contract
Measure Share of Voice without blending the metrics
SOV Tracker treats every scan as a dated sample. Raw Share of Voice is the competitor-relative valid mention share and is never weighted by engine market share, question popularity, question volume or prominence. Discovery Visibility and Prominence answer different questions and stay separate.
Five paid-plan engines, one measurement contract
Paid plans cover ChatGPT + AI Overviews + Gemini + Perplexity + Claude. Each result keeps engine identity, answer evidence and measurement availability explicit.
Claude
Compare Claude observations on a compatible scope instead of treating one response as a permanent ranking.
Gemini
Track whether the target brand and configured competitors appear in Gemini responses and preserve the answer evidence.
Perplexity
Measure valid mentions and retain source URLs when they are exposed in the evidence available to the scan.
AI Overviews
Track dated Google AI Overviews observations with the same measured/not-measured rules used across the product.
One metric family, separate meanings
Share, discovery, prominence and measurement availability remain separate across customer-facing surfaces.
Raw Share of Voice
SOV is the target brand's valid mention count divided by valid mentions for the target brand plus the configured competitor set. It is a percentage share; no position bonus is mixed into it. We do not weight raw SOV by engine market share, question popularity, or question volume.
Discovery Visibility
Discovery Visibility answers a different question: among eligible answered non-branded questions, how often did the target brand appear without being inserted into the question? It is not SOV.
Prominence
Prominence records where the first valid brand mention appears inside the returned answer. It remains separate from SOV and Discovery Visibility rather than being converted into a generic composite.
Measured vs unavailable stays explicit
Canonical SOV is only shown when the configured competitor universe and synchronized scope are valid. Provider failure or missing comparison scope remains unavailable rather than being manufactured as 0 or 100.
Frequently Asked Questions
How many competitors can I track?
Starter supports 10, Pro 25 and Agency 50 configured competitors. The Free Scan does not manufacture canonical SOV from automatically observed entity names.
What exactly is AI Share of Voice?
AI Share of Voice is the percentage of valid mentions belonging to your brand within the configured comparison set: target-brand mentions divided by target-brand plus configured-competitor mentions. It measures share, not prominence.
How is Discovery Visibility different?
Discovery Visibility uses eligible answered non-branded questions and asks whether the brand was discovered without being named in the question. It does not require or replace the configured competitor universe used by canonical SOV.
When is SOV not measured?
If there is no configured competitor universe, the comparison scope is incompatible, or the provider response is unavailable, SOV is reported as not measured. A provider failure is never converted into a measured zero.
Which engines are included?
The no-signup Free Scan checks ChatGPT. Paid plans include ChatGPT + AI Overviews + Gemini + Perplexity + Claude.
How often is paid measurement refreshed?
Priority questions receive a daily ChatGPT + Google AI Overviews pulse. The complete tracked set is measured weekly across all five supported engines. Canonical SOV comes only from the synchronized weekly full-engine window; the daily Priority Pulse is not presented as Daily SOV.
Explore the measurement stack
Start with a dated discovery sample
The no-signup Free Scan checks ChatGPT with five non-branded questions. Add a confirmed competitor universe in the paid product when you want canonical competitor-relative Share of Voice.