Visibility Score
Visibility Score is the Heuriso metric that shows how often, and how favorably, AI presents your brand. It accounts both for answers that leave your brand out and for the position and tone of the mentions it does get.
How does Heuriso calculate it?
For every organic prompt and AI engine, we look at up to the last 5 samples, none older than 14 days. We combine the share of answers that mention your brand with the average quality of those mentions. An answer without your brand scores 0 points; brand-specific control questions are analyzed separately.
- Your brand’s position in the answer: up to 4 points.
- How it is presented (top pick, one of several options or a passing mention): up to 3 points.
- Sentiment, meaning a positive, neutral or negative tone: up to 3 points.
Where does the overall score come from?
Prompt scores are averaged within each engine. Heuriso then combines the engines with different weights: ChatGPT 35%, Gemini 22%, Perplexity 17%, Google AI Overviews 14% and Copilot 12%. If an engine has no data, the weights of the remaining engines are scaled up proportionally.
Being cited as a source does not add separate points. Visibility Score is also not a market share and not a sales forecast.
Why you should look beneath the number
A single number makes a change easy to spot, but it doesn’t tell you what to do. A drop may affect only questions about one service, one AI engine or the way your brand is described. Start by reading the answers that differ from earlier ones. Did your brand disappear, move down the list, or pick up a caveat about its offer? Each of those needs a different explanation.
The score is most useful for comparing the same brand under similar conditions. Adding a new engine or a new set of prompts can shift the average even when the existing answers stay the same. Note such changes next to the dates of your fixes. It will help you tell a change in measurement from a change in visibility.
Common mistake: reading points as a percentage of customers
A score of 70 does not mean you reach 70% of AI users or have a 70% chance of a sale. It summarizes a set of samples according to Heuriso’s rules. Likewise, a ten-point gain does not prove on its own that a new article worked. To judge business impact you also need data on website visits, lead quality and conversions. The metric helps you choose where to look; it doesn’t replace that data.
Example
For one prompt in one engine, your brand appeared in 3 of 5 answers, and the average quality of those mentions was 8/10. The score for that prompt is 60% × 8/10 × 100 = 48 points.
Key takeaway
Compare scores across a similar set of prompts and engines. An increase can mean more frequent mentions, better presentation, or both.
Sources and methodology
Heuriso methodology: the formula, points and weights describe our product. They are not an official Google scale or an industry standard, and the weights do not represent market shares.