How to track your visibility in AI search

Measuring AI search visibility is awkward because there is no single equivalent of a Google ranking report. A business can be mentioned in ChatGPT, cited in Perplexity, appear in Google AI results or be absent from all three, and the platforms do not expose the same data. 

Clicks only show part of the picture. A brand can be recommended without receiving a visit, while a page can be cited without the brand being mentioned prominently. That makes it easy to overstate performance if you rely on one metric or one platform.

For the AI SEO work I do, I use a repeatable process that combines prompt tracking, citations, competitor visibility, Search Console, GA4 and lead data. This article explains that process and the measurements I use to judge whether visibility is improving over time.

Key takeaways

• AI search visibility cannot be measured with rankings alone because mentions, citations, traffic and leads happen across different platforms.

• Track a fixed set of prompts so you can compare visibility over time rather than relying on one-off results.

• Record brand mentions, citations, cited URLs and competitor appearances for every prompt.

• Use specialist AI visibility tools when manual tracking becomes too time-consuming.

• Search Console can show visibility from Google AI features, while GA4 shows identifiable AI referral traffic and leads.

• Compare your results with competitors to understand whether your share of visibility is improving.

• Establish a baseline and review the same core data each month.

• Use the AI search visibility tracking template to keep the process consistent.

In this article, you'll learn:

1. My process for tracking AI search visibility

2. How to choose the prompts worth tracking

3. What to record for each AI search prompt

4. How to use the AI search visibility tracking template

5. When to track manually and when to use AI visibility tools

6. How to track Google AI visibility in Search Console

7. How to track AI referral traffic and leads in GA4

8. How to benchmark your AI visibility against competitors

9. How to establish a baseline and report against it

10. How to interpret the results

11. What AI visibility data cannot tell you

12. FAQs about tracking AI search visibility

My process for tracking AI search visibility

I use the same basic workflow each time:

1. Choose the prompts worth tracking. These need to reflect the services, problems and recommendations that matter to the business.

2. Run the prompts across the main AI platforms. Record whether the brand appears, whether the website is cited and which competitors are mentioned.

3. Track the results over time. A single response means very little. The value comes from repeating the same tests and comparing the pattern month to month.

4. Add first-party data. Search Console shows what Google is surfacing in its AI features, while GA4 shows identifiable referral traffic and conversions.

5. Compare visibility with competitors. This shows whether your presence is improving relative to the businesses competing for the same searches.

6. Review the data together. Mentions, citations, traffic and enquiries answer different questions, so I keep them separate rather than roll everything into one score.

AI search visibility tracking sits alongside the wider SEO reporting I already use. The next step is deciding which prompts are worth tracking in the first place.

Choose the prompts worth tracking

The quality of the tracking depends on the prompts you choose. I focus on prompts that reflect how a potential customer might actually research a service, compare options or look for a provider.

I normally group them into five types:

• Service prompts: “AI SEO services”, “SEO consultant for an e-commerce business”

• Problem prompts: “why is my website losing organic visibility?”, “how can I get my business mentioned in ChatGPT?”

• Recommendation prompts: “best SEO consultant in Dublin”, “who can help a small business with AI search visibility?”

• Comparison prompts: “SEO consultant vs SEO agency”, “traditional SEO vs AI SEO”

• Location prompts: searches where geography affects the answer, such as “local SEO consultant Ireland” or “SEO company Dublin”

The exact wording matters less than the intent behind the prompt. I keep a core set stable so I can compare results over time, but I also include a small number of variations to see whether the same businesses keep appearing when the wording changes.

I also avoid filling the list with prompts that are easy to appear for but have little commercial value. Tracking whether a business is mentioned for an obscure informational question may be interesting, but it tells me far less than whether it appears when someone is comparing providers or looking for help with a service.

The same principle applies to SEO strategy: measure the searches that matter to the business, not every query that happens to generate an impression.

Record what appears for each prompt

For every prompt, I record the result in the same format:

This approach makes the data comparable from one reporting period to the next.

I also separate mentions from citations. If an AI platform names your business but does not reference your website, that is different from a response that uses one of your pages as a source. Both matter, but they tell you different things.

The cited URL is especially important. If the same service page or article keeps appearing, that tells you which parts of the site AI platforms are relying on. It also shows whether visibility is concentrated in one page or spread across a wider content cluster.

I keep competitor mentions in the same row rather than analysing them separately later. That makes it easier to see which businesses repeatedly appear for the same commercial prompts and which topics they are strongest in.

Manual tracking vs AI visibility tools

You can start with a spreadsheet and run the prompts yourself. For a small set of prompts, that is often enough to establish a baseline and understand what the different platforms are returning.

Specialist AI visibility tools become more useful when the prompt set grows or when you want to track several competitors across multiple platforms. They can automate recurring checks, store historical results and report changes in mentions, citations and share of visibility.

I use them for scale, not as a replacement for judgement. The software still needs a sensible prompt set, the right competitors and a reporting structure that reflects the business.

Typical features include:

• prompt tracking across multiple AI platforms

• brand and competitor mention monitoring

• citation and source tracking

• share-of-visibility reporting

• historical trend data

• cited-page analysis

No tool gives a complete view of every AI response, so I use tool data alongside manual checks, Search Console, GA4 and lead data rather than treating it as a standalone source of truth.

Use my AI search visibility tracking template

I use a structured tracker so every prompt is measured in the same way from one reporting period to the next. I’ve turned that process into a Google Sheets template you can copy and use for your own business.

The template includes:

• the prompts you want to monitor

• the AI platforms tested

• brand mentions and website citations

• cited URLs

• competitors appearing for each prompt

• monthly changes in visibility

• AI referral traffic and leads

Make a copy of the AI search visibility tracking template

Start with a manageable set of prompts and record the first results as your baseline. Run the same core prompts again each month so you can see whether your visibility is changing rather than comparing unrelated searches.

Track Google AI visibility in Search Console

Google Search Console now has a Generative AI performance report for AI Overviews and AI Mode. Google started rolling it out in June 2026, so it may not yet appear in every Search Console property.

I use it to track:

• how often pages from the site appear in Google's AI results

• which URLs are being surfaced

• visibility by country and device

• changes in impressions over time

The Generative AI performance report gives you first-party data that manual prompt checks cannot provide. It can also expose patterns worth investigating. If one group of pages is gaining AI impressions while another is not, I want to know what separates them.

Google includes these AI impressions in the main Search Console performance data as well, but the dedicated report lets you isolate the generative AI component.

I keep this data separate from rankings and clicks when reporting. A rise in AI impressions tells me Google is surfacing the site more often in AI results; it does not tell me whether those appearances produced visits or enquiries. That comes from the analytics and conversion data covered next.

Pages already appearing in Google's AI features can also become candidates for deeper review when working on Google AI Overviews optimisation.

Track AI referral traffic and leads in GA4

Prompt tracking tells me whether a business appears in AI answers. GA4 tells me what happens when someone clicks through to the website.

In Reports → Acquisition → Traffic acquisition, I use Session source / medium to identify sessions attributed to AI platforms. Google defines the Traffic acquisition report around the source of each session, including referral traffic.

I then look at:

• sessions from identifiable AI sources

• the landing pages receiving those visits

• engagement on those pages

• lead or enquiry key events

• changes in AI-referred traffic over time

The landing page matters as much as the traffic source. GA4 lets you combine Session source / medium with Landing page + query string, which shows which pages AI-referred visitors entered through.

For a service business, I care more about ten AI-referred visits that generate an enquiry than 200 visits to an informational article that go no further. I therefore track lead events alongside sessions rather than reporting AI traffic as an isolated number. GA4 includes key-event data in its acquisition and landing-page reporting.

AI referral numbers will understate total AI visibility. When a browser or platform does not pass an identifiable referral source, GA4 can classify the session as direct or otherwise fail to attribute it to the AI platform.

The same landing-page and conversion analysis also forms part of how I use Google Analytics for SEO, so AI referral traffic can sit within the reporting setup you already use rather than requiring a separate analytics system.

Benchmark your AI visibility against competitors

AI visibility means more when you compare it with the businesses competing for the same prompts.

I run the same prompt set across the same platforms and record which competitors appear, whether they are cited and which pages are being used as sources.

For each competitor, I track:

• how often the brand appears

• how often the website is cited

• which prompts trigger the mention

• which URLs receive citations

• which topics the competitor dominates

• changes in visibility from one reporting period to the next

The comparison often exposes gaps that raw mention counts hide. A business might appear for 40% of its tracked prompts and still be underperforming if its main competitor appears for 75%.

Competitor data also shows whether the gap sits around a specific service or topic. If a rival keeps appearing for technical SEO prompts while your business performs better for local SEO prompts, the pattern tells you where your visibility is concentrated.

I use the same approach in broader SEO competitor analysis, but AI search requires a fixed prompt set because there is no single ranking table to compare.

Establish a baseline and report against it

The first reporting cycle creates the baseline. I record the prompt set, platforms, competitors, mentions, citations, cited URLs, AI referral traffic and lead data at the same point in time.

From there, I compare the same core data month to month.

The baseline matters because individual AI responses move around. One extra mention in ChatGPT tells me very little. A three-month pattern showing more mentions, more citations and more AI-referred enquiries tells me far more.

I keep the reporting simple:

• percentage of tracked prompts where the brand appears

• percentage where the website is cited

• share of visibility against named competitors

• pages receiving the most citations

• AI referral sessions

• leads or enquiries from identifiable AI traffic

I also keep the core prompt set stable. New prompts can be added when a service or market changes, but constantly rewriting the test set makes trend data unreliable.

For most businesses, I review AI visibility monthly and look for direction over several reporting periods rather than reacting to short-term movement. The same principle applies to SEO performance reporting: changes are easier to interpret when the measurement method stays consistent.

How I interpret the results

The numbers only become useful when you look at the pattern behind them.

If a business is mentioned often but rarely cited, the brand has visibility but the website is not being used as a source very often.

If the website is cited regularly but sends little traffic, I look at where those citations appear and whether the AI response gives the user enough information without requiring a click.

If AI-referred traffic is growing but leads are not, the problem may sit with the landing page, the offer or the match between the prompt and the page receiving the visit.

If one article or service page receives most of the citations, the site may have AI visibility concentrated around a narrow topic rather than across the wider service offering.

If competitors appear consistently for one group of prompts and you do not, I treat that as a topic-level gap rather than a sitewide visibility problem.

The interpretation matters because each pattern points to a different next step. Some gaps are content-related, some are technical, and some sit further down the conversion path. That distinction is useful when deciding whether the next piece of work belongs in SEO content creation, a technical SEO review or the commercial pages themselves.

What AI visibility data cannot tell you

AI search tracking is still less precise than traditional SEO reporting.

The same prompt can produce different answers across platforms, sessions and dates. Some AI tools show citations, others do not. Referral data can also be incomplete, which means GA4 will not capture every visit that started with an AI platform.

I therefore avoid treating one prompt result or one visibility score as definitive. I look for repeated patterns across:

• several prompts

• more than one AI platform

• multiple reporting periods

• citations and source URLs

• referral traffic

• enquiries or other conversions

The data is most useful when it shows direction. If the same competitors keep appearing for the same prompts, or the same pages keep earning citations month after month, the pattern is much more useful than a one-off mention.

Understanding how generative AI search produces and assembles answers also helps explain why AI visibility is less stable than a conventional keyword ranking.

FAQs about tracking AI search visibility

Can I track whether my business appears in ChatGPT?

Yes. Use a fixed prompt set and record whether your brand appears, whether your website is cited and which competitors are mentioned. Repeat the same checks over time rather than relying on one result.

Can Google Search Console track AI Overviews and AI Mode?

Yes. Google introduced a Generative AI performance report in June 2026 for Search Console properties that have access to it. The report shows impressions, pages, countries, devices and trends for AI Overviews and AI Mode.

How often should I check AI search visibility?

Monthly is enough for most businesses. AI responses vary too much for daily checks to tell you much. Keep the same core prompts and competitors so month-to-month comparisons remain useful.

What is the difference between an AI mention and a citation?

A mention means the AI platform names your business. A citation means the platform uses or links to a page from your website as a source. A business can have one without the other.

Can GA4 track traffic from ChatGPT and other AI platforms?

GA4 can identify some visits from AI platforms when referral data is passed correctly. I use Session source / medium, landing-page data and conversions to see which AI-referred visits reach the site and what those visitors do next.

Do I need an AI visibility tracking tool?

Not at the start. A spreadsheet is enough for a small prompt set. Specialist tools become more useful when you need to monitor more prompts, competitors and platforms or keep historical data automatically.

Is AI visibility tracking part of SEO?

Yes. AI search is another way people discover businesses and content, so I include AI visibility alongside rankings, organic traffic and conversions when assessing AI SEO performance.

Need help measuring your AI search visibility?

If you want to know where your business appears across AI search, which competitors are showing up instead, and whether that visibility is turning into traffic or enquiries, I can help you build the measurement around your actual services and commercial priorities.

My AI SEO services cover the research, tracking and analysis needed to understand where the gaps are and what deserves attention next.

Get in touch if you want me to review your current AI search visibility.

About the author

Alessandro Boscolo Conway — Hello Digital

I'm a Dublin-based freelance SEO and digital marketing consultant with over 20 years of experience, including time on Google Ireland’s Search Quality team.

I run Hello Digital, a consultancy that helps startups and small businesses across Ireland grow online through clear strategy, expert delivery, and practical support.

I've worked with over 50 Irish companies to improve their visibility, generate better leads, and grow sustainably through SEO and digital marketing.

I'm a certified Google Partner and a trusted advisor to e-commerce brands, local services, and fast-growing startups.

• Based in Dublin, 20+ years of experience

• Former Googler, certified Google Partner, SEO strategist, and performance marketer

• Trusted by 50+ Irish startups, e-commerce brands, and local businesses

• Learn more about Hello Digital