Trackbase's Visibility Score is one daily number that shows how often and how prominently your brand appears in AI-generated answers. It combines five factors (presence, positioning, frequency, sentiment and comparative performance) instead of simply counting mentions, so it tells you who is winning the AI conversation, not just who showed up.
Most AI-tracking tools stop at a mention count. That is a useful starting point, but it cannot separate being recommended first from being named as an afterthought, or a glowing description from a critical one. This article explains what goes into the score, how it is refreshed, how to read it and where it can mislead you.
What is the Trackbase Visibility Score?
The Visibility Score is a standardized metric of how often and how prominently your brand appears in answers across the 11 AI platforms that Trackbase monitors: ChatGPT, Claude, Gemini, Google AI Overview, Google AI Mode, Perplexity, Grok, DeepSeek, Copilot, LLaMA and Mistral. It is calculated daily, so you can see how AI visibility shifts over time and which actions move the number. The dashboard shows it next to Share of Voice, sentiment and citations.
The five factors behind the score
Trackbase analyzes every AI response on five dimensions:
Presence: was your brand mentioned at all? This is the binary baseline, and the point where most simple tools stop.
Positioning: where in the answer does your brand appear? Being named first or recommended at the top carries more weight than a passing reference at the end.
Frequency: how often does your brand come up across all tracked prompts and platforms? One mention per week looks very different from twenty.
Context and sentiment: are you described positively, neutrally or negatively? "The leading platform for X" is worth far more than a neutral name-drop or a critical reference.
Comparative performance: how do you rank against competitors in the same answers? Visibility only matters relative to who else is in the room.
Why a mention count alone falls short
A raw count treats every mention as equal. Three situations show why that is a problem:
Situation | Mention count says | What is really happening |
|---|---|---|
Named first in "best tools for X" | 1 mention | Strong recommendation |
Named last in a list of ten | 1 mention | Weak, easily skipped |
Named in a warning about reliability | 1 mention | Damaging, and the count rewards it |
A brand mentioned 50 times negatively is worse off than one mentioned 10 times positively, yet a count ranks it higher. By blending all five factors, the Visibility Score reflects the impact of your AI presence, not just its volume. If you want to dig into the difference, see mentions vs citations.
How the daily refresh works
Every 24 hours, Trackbase runs a defined set of prompts across the AI models. The prompts simulate real user queries in your category. The system parses every response, scores it on the five factors and aggregates the result into your Visibility Score.
Because it runs daily, you can:
see the effect of content updates within days instead of months;
catch sudden drops before they cost you customers;
compare changes over time by following the trend line after you publish or fix something.
AI answers vary from run to run, so a single day's movement of a point or two is noise. Look at the trend over a week or more, and at the model-by-model breakdown before drawing conclusions.
What a healthy score looks like
There is no universal target, because context matters. A category-leading B2B SaaS company in a niche market may sit at 70 or above, while a new entrant in a crowded space may start in the 20s. These are illustrations, not benchmarks. What matters is your trend line and your share of voice versus direct competitors.
Three questions help you interpret a score:
Is it rising or falling over the last four weeks?
Which models pull it down? A strong score on ChatGPT and a zero on Gemini points to a different fix than a low score everywhere.
Which competitors sit above you in the same answers, and what do those answers cite?
From score to action
A score is only useful if it leads to work. Trackbase pairs it with prioritized actions and the sources AI models cite for your prompts, so you can see which pages and publishers to influence. For ideas on what to change, read how brands can influence AI recommendations and the explanation of how AI models decide which brands to recommend.
To see your own number, start with the free AI visibility checker, or try Trackbase with the 14-day free trial on the Essential and Professional plans (see pricing).
FAQ
How is the Visibility Score different from Share of Voice?
The Visibility Score describes how prominent your own brand is in AI answers, based on presence, position, frequency, sentiment and comparison. Share of Voice shows your portion of all brand mentions in your category compared with competitors. Use both: the score for your own trend, Share of Voice for your position in the market.
How often is the score updated?
Trackbase runs its prompts every 24 hours across the tracked models, so the score is refreshed daily. Because AI answers vary between runs, judge the trend over several days rather than a single day.
What is a good Visibility Score?
There is no universal target. A leader in a niche market can score high, while a new brand in a crowded category may start low. Compare your score with direct competitors in the same prompts and focus on whether it is rising.




