Testing modeInterface preview. Live audits connect to the OpenForAgents scanner via reverse proxy.
← Back to blog

AI and Business

How to Check Whether Your Business Is Visible in ChatGPT

Contents
  1. What Does Business Visibility in ChatGPT Mean?
  2. Why a Single Check Is Not Enough
  3. How to Check Your Business's Visibility Yourself
  4. 1. Define Relevant Search Queries
  5. 2. Repeat the Queries
  6. 3. Record Responses and Sources
  7. 4. Calculate Visibility Metrics
  8. 5. Repeat the Assessment Later
  9. Can You Obtain Statistics Directly From OpenAI?
  10. Which Services Measure AI Visibility Automatically?
  11. How to Interpret the Results Correctly
  12. What to Do if Your Business Rarely Appears in ChatGPT
  13. Frequently Asked Questions
  14. Can I Check My Company's Visibility in ChatGPT for Free?
  15. Does a Company Have a Fixed Ranking Position in ChatGPT?
  16. How Many Queries Should I Test?
  17. Does Business Hub Show Every Brand Mention?
  18. How Often Should I Repeat the Measurements?
  19. ChatGPT Visibility Should Be Measured, Not Guessed

In January 2026, the research company SparkToro published the results of a study that challenged traditional methods of measuring brand visibility in artificial intelligence. The researchers found that when the same query was repeated multiple times, ChatGPT generated a different list of recommended companies and products almost every time. The probability of receiving an identical list was less than 1%, while the chance of getting the same recommendations in exactly the same order was even lower — below 0.1%.

The study involved 600 volunteers who conducted nearly 3,000 tests across ChatGPT, Claude, and Google's AI search. The researchers compared both which brands appeared in the answers and the order in which they were mentioned.

The findings suggest that measuring a company's presence in ChatGPT using fixed ranking positions, as in Google Search, is of limited value. However, some brands appeared significantly more frequently than others. Instead of measuring a fixed ranking position, it is therefore possible to measure how often a company appears in responses to specific queries.

This approach forms the basis of a new field of analytics known as AI Visibility Monitoring.

Specialized monitoring services and new analytics tools from OpenAI are now available for this purpose. However, an initial assessment can also be conducted independently. The important thing is to understand exactly what should be measured, how to collect the results, and which conclusions can reasonably be drawn from them.

What Does Business Visibility in ChatGPT Mean?

Before starting an assessment, it is necessary to define what counts as a company's presence in ChatGPT.

In traditional SEO, the main performance indicators include website rankings, search impressions, website visits, and commercial actions taken by visitors.

ChatGPT works differently. A company may appear in an answer without a link to its website, be used as a source of information, be mentioned alongside competitors, or be explicitly recommended to a user.

Several metrics should therefore be distinguished.

Metric What It Measures
Mention The company's name appears in a ChatGPT response
Recommendation The company is suggested as a suitable option
Citation ChatGPT provides a link to the company's website
Mention Rate The proportion of tested responses in which the company is mentioned
Referral A user visits the website through a link in ChatGPT

These metrics are related but are not interchangeable.

For example, a company may be used as a source of expert information without being included among recommended service providers. Conversely, it may appear in recommendations based on information from independent sources without a direct link to its official website.

A meaningful assessment must therefore consider not only the number of mentions but also the context in which they appear.

Why a Single Check Is Not Enough

SparkToro's findings were further supported by a larger study published by BrightLocal in September 2026.

The company analyzed 200,085 local search queries across ChatGPT, Google AI Mode, and Google AI Overviews, covering 1,300 business locations.

The researchers repeated identical queries and compared the businesses included in the recommendations.

In ChatGPT, the average overlap between lists of recommended businesses across repeated queries was only 23%. However, a business that appeared in one response was mentioned again in approximately half of the subsequent checks.

This is an important distinction. Recommendations are variable, but the presence of individual businesses can be consistent enough to measure statistically.

Another finding was that ChatGPT mentioned an average of 4.1 businesses per local search query. Even the number of options presented is therefore not constant.

Results may also vary depending on query wording, available information sources, context, and the user's location.

A single check therefore captures only one possible answer, not a company's actual standing in ChatGPT.

Full study: BrightLocal — What 200k Local AI Searches Tell Us.

How to Check Your Business's Visibility Yourself

An independent assessment does not require a subscription to a specialized service. However, the results will only be useful if the testing follows a consistent methodology.

1. Define Relevant Search Queries

The main principle is to test whether a company appears in searches describing its products or services, without necessarily mentioning its brand name.

Asking ChatGPT directly about a specific company only shows whether the system can find information about it. This is not enough to assess competitive visibility.

Instead, the queries should reflect how potential customers might actually search for the company's offerings.

They should cover the company's main areas of activity and different requirements for its products or services. It is useful to prepare several natural variations for each search intent.

It is important not to select only those queries that are already likely to produce a mention of your company.

2. Repeat the Queries

Each query should be submitted several times, preferably in a separate new conversation.

For an initial assessment, one possible approach is to use 20 search queries with five repetitions each. This produces 100 responses for analysis.

This is not a scientifically established minimum sample size. It can provide an initial indication of a company's presence, but substantially more observations may be needed for stronger conclusions.

Testing conditions should remain as consistent as possible, including language, ChatGPT mode, testing period, and settings that may influence the results.

Personalization also matters. Results obtained through a single user account cannot automatically be generalized to the entire ChatGPT audience.

3. Record Responses and Sources

For each query, record the date, wording, names of mentioned companies, source links, and the content of the recommendations.

It is particularly important to distinguish between a simple mention and a recommendation that genuinely matches the query.

The data can initially be collected in a standard spreadsheet. The key requirement is to preserve the original responses so that it is possible to review why a company was included or excluded from the results.

4. Calculate Visibility Metrics

The simplest metric is the Mention Rate.

It is calculated by dividing the number of responses mentioning the company by the total number of tested responses, expressed as a percentage.

It is also possible to calculate the proportion of responses recommending the company and the proportion containing a link to its official website.

Another useful metric is Prompt Coverage: the proportion of unique search queries for which the company appeared at least once.

Together, these indicators provide a more meaningful picture of business visibility than a single position in a recommendation list.

5. Repeat the Assessment Later

To monitor changes over time, use the same set of queries under comparable conditions.

Only then can you assess whether the frequency of company mentions is changing.

However, an increase in a small sample does not mean that visibility among all ChatGPT users has increased proportionally.

The results should be treated as measurements of a defined set of search situations, not as exact statistics for the entire platform.

Can You Obtain Statistics Directly From OpenAI?

On November 10, 2026, OpenAI introduced Businesses in ChatGPT, a system for managing business information and analyzing company presence within the platform.

One of its central features is Business Hub.

After verifying their authority and completing any necessary checks, eligible businesses may gain access to an Insights section containing aggregated statistics for the previous 28 days.

These statistics include business mentions, visits to the company's website, and thematic categories of queries in which the company appeared.

However, OpenAI uses different definitions of mentions.

For an online business, a mention is counted when ChatGPT links to its website. Simply including the brand name without a link does not necessarily count.

For a local business, a mention may be recorded when the corresponding business result is displayed.

The analytics do not provide access to individual user queries or private conversations.

Importantly, a missing metric does not necessarily mean that no mentions occurred. Data may still be collected, may not have reached the threshold required for display, or may be unavailable for the particular profile.

The ability to manage a business profile depends on the country, business type, and available verification methods.

More information about eligibility and metrics: OpenAI — Setting Up Your Business on ChatGPT.

Business Hub provides valuable additional information, but it does not replace independent testing. Its analytics use specific definitions of mentions and do not represent a universal ranking of companies.

Which Services Measure AI Visibility Automatically?

A market for specialized services that monitor business presence in generative search systems has already emerged.

Among the best-known solutions are Ahrefs Brand Radar and Semrush AI Visibility Toolkit.

These tools allow businesses to analyze brand mentions, website citations, competitor presence, and changes in results across selected queries.

However, their methodologies differ.

Some platforms use their own large datasets of search queries, while others allow users to create custom query lists for regular monitoring.

Certain services also calculate additional metrics, such as AI Share of Voice, which measures a brand's relative presence compared with competitors.

The advantage of automated monitoring is the ability to collect large numbers of responses regularly, preserve the results, and compare changes over time.

The main limitation is that these services measure responses to their own test queries, not the complete statistics of all conversations conducted by ChatGPT users.

Therefore, metrics from different platforms cannot be compared directly without examining their methodology: the composition of the query set, number of repetitions, models used, and rules for counting mentions.

When evaluating such services, the findings of the original SparkToro study on the inconsistency of AI recommendations are particularly useful.

How to Interpret the Results Correctly

The collected data can indicate how frequently a company appears in ChatGPT responses under the chosen testing conditions.

However, mention frequency alone does not demonstrate the commercial value of that presence.

Several factors should be considered for a more complete evaluation.

Query relevance. A company may be mentioned frequently, but in responses unrelated to its primary commercial offering.

Recommendation quality. It matters how accurately ChatGPT describes the company's activities, services, and operating conditions.

Competitor presence. Changes in the company's mention frequency should be compared with the visibility of other businesses in the same query category.

Website visits and inquiries. To assess commercial value, website traffic and actual customer inquiries should be analyzed separately when their sources can be identified.

It is also important to distinguish changes in recommendations from the effects of website optimization.

If a company's mention rate increases after its content has been updated, this does not prove that the update caused the increase. External information sources, changes in search systems, and statistical variation in responses may also influence the results.

More reliable conclusions require repeatable testing and observations across multiple periods.

What to Do if Your Business Rarely Appears in ChatGPT

A low frequency of recommendations is a reason for further investigation, but it does not automatically indicate technical problems with the website.

The first step is to check whether the company's offering actually matches the selected search queries.

Next, review the available information sources: how completely the company is described, which websites ChatGPT uses when recommending competitors, and whether published information contains contradictions.

Another important task is to check the technical accessibility of the official website for search tools.

Two characteristics should be distinguished.

Website information readiness describes how accessible and understandable the company's information is.

Actual ChatGPT visibility describes how frequently the company appears in responses under the selected queries and testing conditions.

Good information readiness does not, by itself, guarantee recommendations. However, identified obstacles may limit automated systems' ability to retrieve reliable information about the business.

Visibility monitoring and technical website audits should therefore be used as complementary tools.

Frequently Asked Questions

Can I Check My Company's Visibility in ChatGPT for Free?

Yes. For an initial assessment, businesses can create their own list of queries, repeat them, and record the results. Specialized services are more convenient for regular monitoring and larger datasets.

Does a Company Have a Fixed Ranking Position in ChatGPT?

No. ChatGPT does not produce a permanent ranking of companies comparable to Google's traditional search results. The composition and order of recommendations may change even when identical queries are repeated.

How Many Queries Should I Test?

There is no universal number. An initial assessment can use several dozen queries with repetitions. The broader the subject area and the higher the required level of confidence, the more observations will be needed.

Does Business Hub Show Every Brand Mention?

No. OpenAI applies specific rules when counting mentions. The available statistics may also depend on feature availability, business type, and data display thresholds.

How Often Should I Repeat the Measurements?

For regular monitoring, a monthly assessment using the same methodology can be useful. More frequent measurements are possible when their additional effort and cost are justified by the expected value.

ChatGPT Visibility Should Be Measured, Not Guessed

Research published in 2026 shows that recommendations from generative AI systems can vary significantly even when the same queries are repeated. The traditional concept of a fixed position in search results is therefore not suitable for ChatGPT.

A better-supported approach is to evaluate mention frequency, recommendation quality, and source citations using a repeatable sample of search queries.

Official OpenAI analytics and specialized monitoring tools expand the possibilities for measurement, but each uses its own methodology and has particular limitations.

For businesses, it is important not only to determine how frequently they appear in ChatGPT but also to understand which informational or technical obstacles may be limiting their presence.

A useful starting point is to check how accessible and understandable the information on your company's website is to automated systems.

Check your website's information readiness for free with OpenForAgents and receive recommendations for addressing identified issues.

← Back to blog