Which AI engines we track, how we compute share of voice, brand detection, and sentiment, and how we join it with your real Google Search Console data.
This page explains how JustSEO measures AI visibility, so you can judge the numbers for yourself. We track a defined set of answer engines, detect brand mentions in their responses, and join the result with your real Google Search Console data. Every figure below comes from that pipeline, not from estimates bolted on afterwards.
We send your tracked prompts to the answer engines people actually use and read back what each one says:
Engine coverage depends on your plan, and the exact model behind each engine can change as providers ship updates. DeepSeek is used only to help generate written reports; it is not one of the answer engines we score for visibility.
For each answer, we check whether your brand or domain appears. We do this with whole-word matching rather than raw substring search, so "Notion" is not counted inside "notional" and "Box" is not counted inside "boxing". We normalize case, accents, and separators, and for domains we match the registrable root. This deterministic check is the floor; an LLM review can be layered on top for harder cases.
Share of voice is your slice of the total AI visibility for a prompt. We score your brand and each competitor in the same answer, then divide your visibility score by the combined score of everyone mentioned and express it as a percentage. If only your brand appears, your share of voice for that answer is 100 percent; if you are absent, it is zero. Averaging across prompts and engines gives your overall share of voice.
When your brand is mentioned, we look at the language around the mention and label it positive, neutral, or negative. Words like "best", "leading", or "recommended" push a mention positive; words like "avoid", "limited", or "poor" push it negative. Mentions with no clear signal stay neutral.
AI visibility is only half the picture, so we pair it with your real search data. When you connect Google Search Console, we read your actual impressions, clicks, and query data for the same site and show them next to your AI results. That keeps the analysis grounded in traffic you can verify in your own account.
You can run tracking on demand whenever you want a fresh reading, and tracked projects refresh on a recurring schedule so trends build over time. To keep costs predictable, each plan has a daily limit on how many AI calls it can make; larger plans track more engines and more prompts.
AI answers are not fixed. The same prompt can return different wording from one run to the next, so treat any single reading as a sample rather than a verdict, and watch the trend across many runs. We measure the engines and prompts you configure, not every AI system on the internet, and sentiment from surrounding language is a useful signal, not a precise score. We would rather show you a number we can explain than one that looks certain but is not.
Ready to see it on your own site? Start free or read the JustSEO Academy.
