AI visibility report for the client: what to include in it

This text is part of the guide Brand visibility in AI: how it works and how to measure it. We write it mainly for SEO and marketing agencies that manage their clients’ visibility in AI.

What should an AI visibility report contain?

Terms that should be used consistently throughout the report include reference, citation and sentiment. It’s a good idea to explain them to the client once and then stick to the same terms.

Which metrics should not be included in the report?

• Items on the list. The third character in ChatGPT may be the seventh or disappear the next time it is launched. In a study by SparkToro and Gumshoe.ai ChatGPT and Google’s AI produced the same list of brands in fewer than 1 in 100 attempts.
• Individual screenshots. One dump constitutes one sample. It does not prove either success or a problem.
• A single figure, averaged across all models. The brand may grow in Claude and stagnate in Perplexity. The average will mask this.
• Results from various sets of queries. If the set of phrases has changed during the month, the periods cannot be compared. Therefore, the measurement is based on frozen query set.

How often should visibility be reported in AI?

Once a month. A shorter period mainly highlights variability in responses, whilst a longer one signals problems too late. However, the data for the report should be based on more frequent measurements: daily or at least weekly. Only then is the monthly result an average of many measurements, rather than a fluke from a single day. Read more in the article How often should visibility be measured?.

It is worth highlighting changes to the models in the report. An example from October 2026: 7 and 8 October ChatGPT he switched to GPT-6, and free and paid users were given two different models. Results from before and after that date must be compared as separate periods. We explain this in the text GPT-6 in ChatGPT: what it means for brand visibility.

How can I explain fluctuations in results to a client?

Ideally, before the client even asks about them, in the first report. Three things that are worth stating clearly:

• AI responses are, by their very nature, variable. The model uses a random selection process when generating text and may search different sources each time it is run. See the text for details Why does AI give a different answer every time?.
• The pool of brands is more stable than their order. In the SparkToro survey, the leading brands in the headphones category appeared in 55–77% responses, although the lists almost always differed. The report therefore indicates how often a brand appears, rather than its specific ranking.
• Changes take time to come about. Changes to the website and source data take weeks to be reflected in the models. This is explained in the text Why do changes take time?.

When a customer sends a screenshot with the caption „we’re not on ChatGPT”, the answer is in the report: this is just one of many attempts, and it’s the percentage of the total that counts.

How can the results of the activities be presented in the report?

By keeping a log of changes with dates and comparing it with the results for subsequent periods. An example from the Wapiennik „Węże” case study, a lime producer based in Działoszyn:

Between the first and second reports, a statement of compliance with the standard, a table of specifications and a section on pricing were added to the website. In a report for a client, such a table, accompanied by a change log, conveys more information than a description alone. However, it must be fairly noted that a few measurements of a single brand do not fully distinguish the effect of the measures from model variability. The entire process is described From zero to 63% visibility in AI in five weeks and An AI model recommends a brand for the first time.

How can I combine an AI report with an SEO and advertising report?

Keeping them in separate sections, as they measure different things:

• SEO: search engine rankings and traffic; why this is not the same as appearing in AI-generated answers, explains Visibility in AI and SEO,
• advertisements on ChatGPT: campaign results from OpenAI Ads Manager; the advert is labelled and, according to OpenAI, does not influence the response, so its results must not be confused with its presence in the responses; more on this in the text Advertisements on ChatGPT in Poland,
• the presence of AI in the responses: measurement results for several models.

Copilot gets additional data from the free AI Performance report in Bing Webmaster Tools, which shows which Copilot subpages and AI responses in Bing are cited. For ChatGPT, Claude, Gemini and Perplexity, no such report is available from the providers.

What might the structure of a client report look like?

1. A one-page summary: the three most important changes and one decision to be made.
2. Scope of measurement: phrases, models, period.
3. Brand presence by model, compared with the previous month.
4. List of pages and subpages that models cite most frequently.
5. Competition: who appears alongside the brand and who takes its place.
6. How do models describe the brand, and do they provide incorrect information?.
7. Sources used by the models in this category.
8. A log of activities, dated.
9. Plan for the coming month.
10. Changes to the models that may have affected the results.

If models provide incorrect information about a client, it is worth mentioning this in a separate section. The text explains how to correct such errors AI is providing false information about the company.

How do you prepare an AI visibility report in Aelo?

At Aelo, measurements are carried out daily across four models: ChatGPT, Claude, Gemini and Perplexity. Once a month, a full report is generated, compiling the results from all scans: brand presence, mentions, competition and the way the models talk about the brand. The results are fed into a knowledge graph Synapsis, and on this basis we draw up plans for Tickets, which can easily be moved to the „plan” section of the client report.

Agencies can manage several brands under a single account on the Pro and Custom plans, and on the Custom plan they can use the white-label version and the API. Full details are available on the website Aelo for agencies, and the whole process is described in An AI-powered brand visibility platform: how Aelo works. You can check the client’s brand’s current status at free measurement.

Try it for free

FAQ

Sources

• SparkToro and Gumshoe.ai, a study on the consistency of brand recommendations in AI (January 2026): https://sparktoro.com/blog/new-research-ais-are-highly-inconsistent-when-recommending-brands-or-products-marketers-should-take-care-when-tracking-ai-visibility/
• OpenAI, GPT-6 and Intelligent UI for everyone (7 October 2026): https://openai.com/index/gpt-6-for-everyone/
• OpenAI: A milestone in expanding access to AI (31 August 2026): https://openai.com/index/expanding-access-to-ai-with-chatgpt-ads/
• Search Engine Land, AI Performance report in Bing Webmaster Tools (February 2026): https://searchengineland.com/bing-webmaster-tools-ai-performance-report-468751
• Aelo reports for Wapiennik „Węże” dated 24 June, 27 July and 10 September 2026 (published with the client’s consent)