Search is changing. Increasingly, people are turning to AI tools such as ChatGPT, Google Gemini, Perplexity and Microsoft Copilot to research products, compare suppliers and ask for recommendations.
Instead of presenting a page of search results, these platforms generate a direct answer. That answer may mention your brand, recommend a competitor, quote information from your website or rely on a third-party publication as its source.
Your AI visibility dashboard powered by Rankscale helps us monitor where and how your brand appears within these answers.
However, AI visibility is not represented by one number alone. The dashboard contains several metrics, each revealing something different about how AI platforms understand, describe and recommend your business.
Here is what they mean.
Topics and search terms: what are we actually tracking?
Before looking at the results, it is important to understand the topics and search terms being monitored.
Topics
Topics are the broader areas in which we want to measure your brand’s AI visibility.
They are normally based on your services, products, audiences and commercial priorities. For example, a wine estate might have separate topics for:
- South African red wine
- South African white wine
- Wine tasting experiences
- Restaurants and accommodation
A business can perform very differently across different topics. It may be frequently recommended for one service but barely appear for another.
Dividing the tracking into topics helps us identify where the brand is already well established and where there are gaps or opportunities.
Search terms
Within RankScale, “search terms” are the individual prompts or queries that are run through the selected AI platforms.
Although the software calls them search terms, they should not be confused with traditional SEO keywords.
A traditional keyword is usually something a person types into Google, such as:
commercial insurance broker Devon
An AI search term or prompt represents a question, need or scenario being tested within an AI platform, such as:
best commercial insurance brokers in Devon
RankScale runs these selected terms across AI platforms and records what appears in the generated answers.
These prompts give us a consistent benchmark. They do not represent every possible question someone could ask about your industry, and they are not necessarily a record of exact questions asked by individual users. Instead, they are a carefully selected sample of commercially relevant searches that allow us to measure changes over time.
The topics and prompts selected at the beginning of a campaign are a starting point. They can be refined as business priorities change or as we learn more about how customers search for your services.
Visibility Score
The Visibility Score is the main headline measure of your overall presence in AI-generated answers.
RankScale describes this as an aggregate score from 0 to 100, combining factors such as:
- How frequently your brand appears
- Where it appears within an answer
- How prominently it is presented
- How consistently it appears across the tracked prompts and AI platforms
A higher Visibility Score generally means the brand is appearing more frequently and more prominently.
However, this number should not be viewed in isolation. Two brands could have similar visibility scores for very different reasons. One might be mentioned frequently but positioned near the bottom of the answers, while another might appear less often but be presented as a leading recommendation when it does appear.
The greatest value comes from tracking visibility over time, comparing it across important topics and looking at it alongside the other metrics.
Detection Rate
Detection Rate shows the percentage of tracked AI searches in which the brand was found at all.
For example, if a brand is detected in 12 out of 20 tracked results, its Detection Rate would be 60%.
This tells us about the breadth of the brand’s visibility. Is it appearing across a meaningful proportion of the questions prospective customers might ask, or only in a small number of isolated answers?
A low Detection Rate does not necessarily mean an AI platform knows nothing about the brand. It means the brand was not included in the answers generated for the particular topics, prompts, locations and platforms being tracked.
Mentions
Mentions show how many times the brand was named within the AI-generated answers.
A brand can be mentioned:
- As a direct recommendation
- As one option within a list
- In comparison with another business
- As an example
- In passing within a wider explanation
A mention does not necessarily include a link to the brand’s website. It simply means the AI platform included the brand in its answer.
Mentions help us understand whether the brand is part of the conversation. However, we also need to look at position and sentiment to understand how prominently and positively it appeared.
Share of Voice
Share of Voice compares the brand’s mentions with those of its tracked competitors.
For example, if the monitored brands received 100 mentions between them and your brand accounted for 25, your Share of Voice would be 25%.
This turns mentions into a competitive measure. It helps answer questions such as:
- Which brands dominate AI-generated discussions in this market?
- Is your business appearing as frequently as its competitors?
- Is one competitor consistently receiving most of the attention?
- Is your share of the conversation increasing over time?
Mentions are the raw count. Share of Voice is your proportion of the competitive conversation.
Position
Position shows where the brand appears within the AI-generated answer when it is included.
If an AI platform recommends several businesses or products, appearing first or second is likely to be more valuable than appearing much later in the response.
This is not the same as achieving position one in Google. It refers to the brand’s placement within the generated AI answer.
RankScale broadly interprets positions as:
- Positions 1 to 3: strong visibility
- Positions 4 to 6: moderate visibility
- Position 7 or lower: weaker prominence
- Not found: the brand did not appear
Position is important because simply being mentioned does not tell us how influential that mention is. A brand presented as the leading choice has a very different level of visibility from one included as the eighth alternative in a long list.
Top Three Visibility
Top Three Visibility shows how frequently the brand appears within the first three positions in the tracked AI answers.
These first three placements are especially important because users are more likely to notice and consider the brands presented near the beginning of a recommendation.
For example, a brand might have a high Detection Rate because it appears in many answers, but a low Top Three score because it is usually placed further down the list.
Looking at Detection Rate and Top Three Visibility together helps us distinguish between simply being present and being presented as a leading option.
Sentiment Score
Sentiment measures the tone AI platforms use when describing the brand.
It assesses whether the language surrounding a brand is:
- Positive
- Neutral
- Mixed
- Negative
A brand might be described positively for its expertise, service quality or product range. Alternatively, an AI response might repeatedly mention concerns about price, customer service, availability or reputation.
Sentiment matters because visibility is not automatically beneficial. Being mentioned frequently is less valuable if the brand is being described negatively or inaccurately.
We therefore look beyond the overall score to understand:
- Which positive qualities AI platforms associate with the brand
- Whether the brand is being represented accurately
- Whether the same concerns appear repeatedly
- Where outdated or incorrect information may be influencing the answers
This can help inform website content, reputation management, PR activity and wider brand messaging.
Citations
Citations are the source links included within AI-generated answers.
When an AI platform generates a response, it may link to websites that support the information it has provided. These could include:
- The brand’s own website
- Industry publications
- News websites
- Directories
- Retailers
- Review platforms
- Research organisations
- Competitor websites
- Government or professional bodies
A citation is therefore not necessarily a link to the brand being mentioned.
For example, an AI platform could recommend a particular wine estate but cite an article from a respected wine publication as the source supporting that recommendation. The estate receives the mention, while the publication receives the citation.
This distinction is extremely important.
Your brand’s AI visibility does not depend only on what is published on your own website. It can also be influenced by what trusted third-party websites say about you.
Why citations are so valuable
Citation data helps us identify which websites AI platforms appear to trust when answering questions about a particular subject.
If the same publications, directories or industry sources are repeatedly cited, these domains may be valuable targets for:
- Digital PR
- Editorial coverage
- Reviews
- Awards
- Industry listings
- Partnerships
- Expert contributions
- Updating existing articles that currently mention competitors but not your brand
Our aim may be to help the client’s own website become a cited source. However, it can be equally valuable to secure accurate, authoritative coverage on the third-party websites AI platforms already use.
This is one reason AI SEO extends beyond publishing more content on your own website. It also involves building authority and strengthening the brand’s presence across the wider web.
Mentions and citations: what is the difference?
The simplest way to understand the distinction is:
- A mention tells us that the AI platform named your brand.
- A citation tells us which online source the AI platform linked to when supporting its answer.
A brand may be mentioned without being cited. It may also be mentioned because a trusted third-party source has written about it.
This creates two important opportunities:
- Improve the client’s website so its own pages are more useful, authoritative and easy for AI platforms to understand and cite.
- Build the brand’s presence on the external publications, directories and websites that AI platforms already trust.
Why results vary between different AI platforms
ChatGPT, Gemini, Perplexity, Copilot and other AI platforms do not all produce the same answers.
Each platform may:
- Use different AI models
- Access different sources
- Interpret a prompt differently
- Refresh its information at different times
- Place different weight on brands, publications and other sources
Results can also change from one run to another because generative AI is probabilistic. It creates an answer rather than retrieving one permanently fixed result.
This means small day-to-day movements should not automatically be treated as meaningful changes. We focus more heavily on:
- Longer-term trends
- Repeated patterns
- Comparisons across platforms
- Performance across priority topics
- Consistent gains or losses against competitors
The dashboard is most valuable as an ongoing monitoring tool, not as a single snapshot.
How we use this information
The purpose of monitoring AI visibility is not simply to produce another set of scores.
We use the information to identify where and why a brand is being included, overlooked or outperformed within AI-generated answers.
Depending on what the data reveals, our work may include:
- Making the brand and its offering clearer to AI systems
- Improving important service and product pages
- Creating content that answers relevant customer questions
- Strengthening the website’s topical authority
- Improving structured data and other technical signals
- Correcting inaccurate or outdated information
- Developing content that is easier for AI platforms to extract and cite
- Identifying trusted third-party websites for PR and outreach
- Building stronger brand mentions across the wider web
- Analysing how competitors have earned their visibility
- Addressing recurring negative or mixed sentiment
Different metrics point towards different actions. High mentions but few citations may indicate a source-authority gap. A good Detection Rate but weak positioning may show that the brand is known but rarely treated as a leading choice. Strong visibility in one topic and poor visibility in another may reveal a content or authority gap around a particular service.
Your dashboard is a starting point, not a final verdict
Initial AI visibility results establish a baseline.
Low visibility at the beginning of a campaign is not necessarily a cause for alarm. It shows us where the brand currently stands and helps identify the areas where improvement is most needed.
The topics, prompts and competitors being tracked define the context of the results. Your scores do not represent your visibility for every possible AI search. They represent your performance within the commercially relevant areas selected for monitoring.
Over time, the dashboard allows us to measure whether your brand is:
- Appearing more frequently
- Earning more prominent positions
- Gaining competitive share of voice
- Being described more positively
- Becoming a trusted source
- Being supported by stronger third-party coverage
- Appearing across more of the AI platforms your customers use
AI search is still evolving, and the way platforms discover, assess and recommend brands will continue to change. Monitoring these different signals gives us the evidence needed to respond strategically rather than relying on guesswork.
Need Help?
If you want to understand how your brand is appearing in AI-generated answers, where competitors are gaining visibility and what needs to change to improve your position, our AI-enhanced SEO service can help. We combine AI visibility monitoring with strategic content, technical SEO and wider authority-building activity to help your business remain discoverable as search evolves. Find out more about our unique approach to Adaptive SEO or contact us to discuss how we can help your brand become more visible, trusted and frequently recommended.