Explore the
questions that
shape AI discovery
People increasingly ask AI systems detailed, conversational questions as they research services, compare providers and consider their options. Visibility Prompting is Macrocosm’s structured approach to exploring that behaviour through questions shaped around real buyer intent.
Rather than relying on broad, generic prompts, we develop question pathways that reflect what prospective customers want to know, including the service they need, where they need it and what matters when choosing a provider. This brings the buyer’s perspective into focus and explores how brands are understood across different AI environments.
Visibility Prompting examines the language and context surrounding discovery, comparison and selection. It creates a practical framework for exploring how a business fits into the questions people ask and the answers AI systems provide.
What is
Visibility
Prompting?
A prompt is the question given to an AI system. In this methodology, prompts are based on the questions prospective customers might ask when researching a service, comparing providers or looking for a recommendation.
We keep the questions consistent, ask each selected platform directly and assess the answers using the same rules. This creates a clear view of how each system responds without assuming that one platform represents another.
What is Visibility
Measurement?
Visibility Measurement is the subcategory that records and assesses the results of Visibility Prompting. It tracks whether selected AI systems mention, recommend and cite a business, and whether they describe it accurately. The same buyer questions are tested over time. Results are reported as measurable rates, with each platform’s performance compared against its own baseline.
The Visibility
Quintet of
Measurement
What does Visibility Measurement reveal? AI visibility involves more than whether a business appears in an answer. Visibility Measurement separates five outcomes, so the results show what happened and how often.
Mention rate
Mention Rate shows how often an AI answer names your business in response to the agreed buyer questions. It is calculated as the number of tested answers that mention your business divided by the total number of answers reviewed. This reveals whether your brand is entering the conversation, even when it is not presented as a recommendation.
Recommendation rate
Recommendation Rate measures how often an AI answer presents your business as an option for the person asking the question. A passing reference does not count; the answer must position your business as a relevant choice. This distinguishes simple brand awareness from stronger consideration within the response.
Citation rate
Citation Rate measures how often an AI answer links to your website as a source. It shows whether the platform directs users to your site to support or verify its response. A mention without a link counts as a mention, but not as a citation.
Share of voice
Share of Voice compares your business’s mentions with all brand mentions across the tested answers, including competitors. It shows the proportion of the measured brand conversation associated with your business. This makes it possible to track how your presence changes relative to the other businesses named in the same answers.
Description accuracy
Description Accuracy measures whether AI answers that mention your business get its verified facts right. We compare each description with the agreed information about your services, locations and other key business details. The rate is calculated from answers that mention your business; if there are no mentions in the measurement period, accuracy is recorded as not available.
How do we prompt
AI systems?
A Fixed Set of Buyer Questions
We agree on around 20 questions a prospective customer might ask when researching a service, comparing providers or looking for a recommendation. The same set is used for monthly measurements so that changes in the results are not caused by changing the questions.
Each Platform Is Tested Directly
We ask each selected AI platform directly. We do not use one assistant’s answer to predict what another platform will say.
Repeated Runs Where Answers Vary
Chat assistants can give different answers to the same question. We ask each question multiple times during a measurement period and count the results across those runs. Google AI Overviews are reported separately. They are grounded in live search and provide one answer per query, so their results are not blended with repeated chat assistant answers.
Questions Asked as a Buyer Would Ask Them
We give each assistant the question without first providing the client’s website or preferred answer. This avoids priming the system and lets us observe its response to the buyer question itself.
A Baseline for Every Platform
The first measurement establishes a separate baseline for each platform. Later reports show movement against that baseline. If a platform is added later, its first measurement becomes its own baseline.
We report how often an outcome occurred and how many answers were tested. For example, “mentioned in 8 of 20 answers” shows what was counted and how the rate was calculated.
We do not reduce the results to a single AI visibility score. A single score can hide what was measured, how it was calculated and where the change occurred.
Platform results stay separate, making it easier to see where your business is appearing, where it is absent and where its description needs attention.
Rates,
not scores
The signals behind
the numbers
Five foundations that can shape how AI systems represent your business. AI systems draw on information from across the web. These five areas can influence how clearly and consistently your business is represented in their answers. They help identify opportunities for review; they do not guarantee that an AI system will mention or recommend a business.
Entity clarity
Your business name, services, locations and description should be consistent across your website and other trusted sources. Clear, consistent information makes it easier to distinguish your business from others and understand what it offers.
Independent validation
Genuine reviews, relevant press coverage and appropriate directories can corroborate information about your business. The methodology records what AI systems return; it does not rely on paid or seeded mentions to influence the results.
Useful website pages
Your website should answer the questions prospective customers ask about your services, process and experience. Specific details, including pricing where appropriate, give people and AI systems clearer information to assess.
Review freshness
Review volume and recency on relevant platforms affect the current information available about your business. Reviewing these signals can reveal whether your public reputation reflects the business as it operates today.
Crawl access and indexability
Your website needs to be accessible to relevant crawlers, without blocking directives that prevent discovery. Search Console data can help assess whether Google is surfacing your pages in its AI search features.
Visibility
Measurement
within the art
of visibility
Visibility Measurement gives the wider visibility strategy a way to observe how selected AI systems respond to your brand.
SEO supports access to and discovery of your website. GEO, AEO, AI Optimisation and LLMO address how brand information can be retrieved, interpreted and used in answers. Visibility Measurement tracks the results that appear: whether your business is mentioned, recommended, cited and described accurately.
The measurement does not control AI answers or guarantee inclusion. It provides a consistent view of what is happening, a baseline for comparison and evidence to inform the work ahead.
Why do visibility
prompting and
measurement
matter now?
People increasingly use AI assistants to explore services, compare providers and ask for recommendations. The questions they ask help shape which businesses appear in the answers and enter the shortlist. Without structured prompting, it is difficult to see how AI systems respond to the kinds of questions prospective customers actually ask.
Visibility Prompting creates a consistent way to test those questions across selected AI platforms. Visibility Measurement then records whether a business is mentioned, recommended, cited and described accurately. Traditional rankings show how pages perform in search results; this process shows how AI assistants respond to buyer questions about your business and its competitors.
AI responses can vary, and platforms update on their own schedules. Repeating the same prompts and measuring the answers over time makes results easier to compare and helps distinguish a developing trend from a one-off response. That gives businesses clearer evidence of where their AI visibility is strong, where it is missing and what may need attention.
How does
Macrocosm deliver
Visibility
Prompting?
We agree on the scope
We identify the business, its priority services, locations, relevant competitors and the AI platforms to include.
We build the buyer question set
We create and agree on around 20 relevant buyer questions. This fixed set is used for the baseline and future comparisons.
We establish the baseline
We ask each platform directly, repeat questions where answers vary and record the initial results for each platform.
We assess the answers
We apply consistent rules to count mentions, recommendations, citations and Share of Voice. We compare descriptions with verified business facts.
We report movement and opportunities
Monthly reporting shows movement against each platform’s baseline and highlights information, content and reputation areas that may warrant attention.
We review the trend
AI systems update on their own schedules. We review the results as a trend across at least three months, rather than treating one measurement as a final verdict.
Frequently
asked
questions
What is Visibility Prompting?
Visibility Prompting is Macrocosm’s structured approach to asking selected AI systems relevant buyer questions and observing their responses. The prompts are kept consistent so the answers can be assessed and compared over time.
What is Visibility Measurement?
Visibility Measurement is the subcategory that records and assesses the results of Visibility Prompting. It measures how often selected platforms mention, recommend and cite a business, and whether their descriptions are accurate.
How are the buyer questions chosen?
We build the question set around realistic customer needs, such as researching a service, comparing providers or asking for a recommendation. The questions are agreed before the baseline and kept consistent for comparison.
Why ask each platform directly?
AI systems can return different answers. Testing each platform directly records its own response instead of assuming one assistant represents the others.
Why are some questions asked more than once?
Chat assistants can vary their answers. Repeated runs show how often an outcome occurs across multiple responses rather than relying on a single answer.
How are Google AI Overviews measured?
Google AI Overviews are recorded separately because each query provides one answer grounded in live search. They are not combined with repeated chat assistant responses.
Does Visibility Prompting use tokens to improve visibility?
No. Tokens are units of text processed by language models. They are not a reliable measure of whether a business appears or is recommended. This methodology measures the answers returned to agreed buyer questions.
Can Macrocosm guarantee that an AI assistant will recommend my business?
No. Macrocosm can measure how selected platforms respond and identify areas for review, but it cannot control third-party AI systems or guarantee they will mention, cite or recommend a business.
What if an AI system recommends a competitor?
That result is recorded as part of the current visibility picture. It shows which questions, platforms and competing brands may need closer review.
