---
title: "AI visibility checker: a manual test that works | PostCliq"
description: "How to check AI visibility in ChatGPT and Google AI Overviews: a ten-prompt protocol, a scoring sheet, the free tools and what each one really shows."
url: "https://postcliq.com/blog/check-ai-visibility"
---

# AI visibility: how to check whether ChatGPT and Google AI mention your business

AI visibility is whether assistants name or cite you when buyers ask about your category. Check it in twenty minutes a month with ten fixed questions, a scoring sheet and a few free reports.

By [Nirmal Gyanwali](https://postcliq.com/blog/author/nirmal-gyanwali) Updated 15 October 2026 8 min read

On this page

1. The short answer
2. What does AI visibility actually measure?
3. How do you check AI visibility manually?
4. What should you record for each answer?
5. Which AI visibility checkers and reports are worth using?
6. Why does being cited matter if clicks are falling?
7. What does PostCliq's AI readiness check cover today?
8. What to do next
9. Frequently asked questions

## The short answer

Key takeaways

- AI visibility is whether ChatGPT, Perplexity, Gemini, Copilot and Google's AI features mention or cite your business for the questions buyers actually ask.
- A useful check asks the same ten buyer questions, in fresh chats, on the same assistants, on the same day each month.
- Record four things per answer: mentioned or not, cited or not, which of your pages, and who was cited instead.
- Free AI visibility checkers give a snapshot, and Bing and Google now report some AI data, but no tool shows a complete or stable picture.
- A score only means something as a trend, and with thirty answers a month each answer moves the score by more than three points.

The reason to check is simple. When Google shows an AI summary, people click less. In the Pew Research Center's study of US adults in March 2025, users clicked a traditional result on 8% of searches with an AI summary against 15% without one, and clicked a link inside the summary itself on [just 1% of visits](https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/) . If fewer buyers reach your site, being named inside the answer matters more.

Below is the protocol, what to record, the tools worth using, and what PostCliq does and does not check. For the wider strategy of earning citations, read our guide to [generative engine optimisation](https://postcliq.com/blog/generative-engine-optimisation) , and for related guides see [AI search and visibility](https://postcliq.com/blog/topic/ai-search) .

## What does AI visibility actually measure?

AI visibility measures presence inside generated answers: whether an assistant names your brand, whether it links to your pages as a source, and how you compare with competitors for the same questions. It is not a ranking. Answers vary by prompt, day, account and location, so any measure is a sample, not a census.

Separate three things that tools often blur together:

| What | Definition | Best free way to measure |
| --- | --- | --- |
| Mention | The assistant names your business in its answer | Manual prompt panel |
| Citation | The assistant links to one of your pages as a source | Manual prompt panel, Bing AI Performance for Copilot |
| AI referral | A person clicks through from an assistant to your site | GA4 sessions by source, counted as a floor |

Source: PostCliq analysis

A mention without a citation still shapes a shortlist. A citation without a mention is common for how-to content, where the assistant uses your page but names no one. Referrals are the only one of the three that lands on your site, and they are small for most businesses. Measure all three, and do not report one as if it were another.

## How do you check AI visibility manually?

Write ten questions a buyer would ask about your category, without your brand name in them. Ask each one in a fresh chat with no history, on the two or three assistants your buyers use, on one day. Score each answer on the same sheet. Repeat monthly with identical wording.

The prompts matter more than the tool. Use the language a buyer uses, not your internal terms, and include your location where buyers would.

| # | Prompt pattern | Example for an Australian accounting firm | What it tests |
| --- | --- | --- | --- |
| 1 | Best providers in a place | Who are the best small business accountants in Brisbane? | Local shortlist |
| 2 | Situation-led recommendation | I run a 20-person construction company in Brisbane. Who should I talk to about tax? | Fit for a segment |
| 3 | Selection criteria | What should I look for when choosing a small business accountant? | Whether your advice is quoted |
| 4 | Price | How much does a small business accountant cost in Australia? | Whether your pricing page is used |
| 5 | Comparison | What is the difference between a bookkeeper and an accountant? | Educational content |
| 6 | Specialism | Which accountants in Queensland specialise in construction? | Niche authority |
| 7 | Problem-led | How do I get ready for EOFY as a small business? | Timely content |
| 8 | Alternatives | What are alternatives to a big four firm for a small business? | Category position |
| 9 | Brand check, unleading | What do people say about \[your brand\]? | Reputation summary |
| 10 | Head to head | \[Your brand\] or \[competitor\]: which is better for a small business? | Direct comparison |

Source: PostCliq analysis, adapted from the AI visibility check: ten buyer questions

Watch out

Never add "including \[your brand\]" to a prompt, and never test while logged in to an account that has chatted about your business before. Both inflate the result.

Which assistants? Start with where your buyers are. Similarweb data shows ChatGPT's share of visits to generative AI platforms fell from about 76% to about 53% between June 2025 and May 2026, as total visits grew 70% to 9.5 billion a month ( [Similarweb](https://aisearch.similarweb.com/blog/gen-ai-stats/) , vendor panel data). ChatGPT is still the largest, but testing only ChatGPT now misses close to half the traffic. For most Australian B2B and service businesses, ChatGPT, Google (AI Overviews and AI Mode, which launched in Australia on [8 October 2025](https://blog.google/intl/en-au/products/explore-get-answers/ai-mode-in-australia/) ) and one of Perplexity or Copilot is a sensible panel.

## What should you record for each answer?

Record the date, assistant, prompt, whether you were mentioned, whether you were cited, which of your URLs was cited, which competitors were named, and which sources were cited instead. The last column is the most useful, because the pages cited instead of yours are the pages to compete with or get mentioned on.

| Field | Example | Why it matters |
| --- | --- | --- |
| Date and assistant | 30 Sep 2026, ChatGPT | Answers drift, so every row needs both |
| Prompt number | 4 | Keeps wording identical month to month |
| Mentioned | ✓ | Are you on the shortlist |
| Cited | ✕ | Did it use your page as a source |
| Your cited URL | none | Which page earns citations |
| Competitors named | Firm A, Firm B | Share of the shortlist |
| Sources cited instead | A directory listing, a Reddit thread | Where to earn a mention next |

Source: PostCliq analysis

### A worked example: what a score can and cannot tell you

Say you run ten prompts on three assistants. That is 30 answers. You are mentioned in 9, so your mention rate is 9 divided by 30, or 30%. Next month you are mentioned in 12: 40%.

It looks like a ten-point gain. It is three answers. Each answer is worth 3.3 points of the score (100 divided by 30), and assistants give different answers to identical prompts on different days. Our rule of thumb is to treat a change of three answers or fewer as noise, and to look for a direction held over three months before reporting it upward. If you need a finer measure, run each prompt twice and double the sample to 60 answers, which halves the weight of each answer to 1.7 points.

That arithmetic is also why a single screenshot of ChatGPT naming you, or not naming you, is not evidence of anything. Show your director the trend line, not the screenshot.

## Which AI visibility checkers and reports are worth using?

Use free tools for a snapshot and platform reports for trends. Ahrefs offers a free checker that samples several AI platforms for a brand name. Bing Webmaster Tools reports citations in Copilot. Search Console shows impressions in Google's AI features. None of them replaces a fixed prompt panel on the questions your buyers ask.

Here is what each one actually does, as of 30 September 2026:

- **Ahrefs AI Visibility Checker (free).** You enter a brand name. It reports mentions across ChatGPT, Gemini, Perplexity, Copilot and Google AI Overviews, with the top five topics, cited domains and cited pages. The free version is a one-time snapshot, and Ahrefs says ongoing tracking needs its paid Brand Radar product, from USD 199 a month ( [Ahrefs](https://ahrefs.com/ai-visibility-checker) ). Useful as a first look. The prompts are Ahrefs', not your buyers'.
- **Bing Webmaster Tools AI Performance (free, public preview since 10 February 2026).** Shows total citations, cited pages and the grounding queries behind them, across Microsoft Copilot, AI summaries in Bing and select partners ( [Bing Webmaster Blog](https://blogs.bing.com/webmaster/February-2026/Introducing-AI-Performance-in-Bing-Webmaster-Tools-Public-Preview) ). It is the only first-party citation report we know of, and it does not cover ChatGPT.
- **Google Search Console generative AI performance report (free, rolling out).** Shows impressions from AI Overviews, AI Mode and generative features in Discover, by page, country and device. It does not show clicks, click-through rate or the prompts ( [Search Console Help](https://support.google.com/webmasters/answer/16984139) ). Clicks from AI features are counted inside ordinary Web search results.
- **GA4.** Sessions with sources such as chatgpt.com or perplexity.ai. A floor, not a total, because some assistant visits arrive without a referrer and show as direct.

> Shows the total number of citations that are displayed as sources in AI-generated answers during the selected time frame.

Microsoft, Bing Webmaster Tools AI Performance documentation, February 2026

Many other "AI visibility checker" tools rank for this search, and the category is young. Before paying for one, ask three questions: which assistants and models does it query, are the prompts yours or its own, and does it show the raw answers so you can check them? If it cannot answer the last one, you cannot audit its score.

## Why does being cited matter if clicks are falling?

Because the clicks that remain go disproportionately to cited brands. Seer Interactive found brands cited in Google's AI Overviews earned 35% more organic clicks than brands that were not, even as organic click-through fell across the board. Citation does not replace traffic. It decides whose traffic survives.

What people clicked on Google, US adults, March 2025

- Clicked a traditional result, no AI summary 15%
- Clicked a traditional result, AI summary shown 8%
- Clicked a link inside the AI summary 1%

Source: [Pew Research Center, July 2025](https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/)

The Pew data is US only, one month, and AI features have expanded since. Ahrefs' larger dataset points the same way: an AI Overview was associated with a 58% lower click-through rate for the top result, across 300,000 keywords ( [Ahrefs, February 2026](https://ahrefs.com/blog/ai-overviews-reduce-clicks-update/) , vendor study, modelled against a projection). Against that, Seer's figure that cited brands got [35% more organic clicks](https://www.seerinteractive.com/insights/aio-impact-on-google-ctr-september-2025-update) is the reason to track citation at all. Seer is clear it cannot prove citation causes the higher click-through, so treat it as a correlation.

For the traffic side of this story, and why no one can yet give you a trustworthy "traffic lost to AI" number, read [what AI Overviews do to your search traffic](https://postcliq.com/blog/google-answers-the-question) .

## What does PostCliq's AI readiness check cover today?

PostCliq checks whether your own pages are ready to be read and cited. It does not ask assistants about you, count mentions or track citations. On up to 50 public pages it checks text readable without JavaScript, valid structured data naming your organisation, robots.txt rules per AI crawler, and whether claims sit near a source.

Being precise about scope matters in a category full of scores. Here is what the free audit's AI readiness category does and does not do:

| Check | Covered | What it looks at |
| --- | --- | --- |
| Text readable without JavaScript | ✓ | Word counts in the raw HTML of each audited page |
| Structured data parses and names the organisation | ✓ | JSON-LD validity; Organization, LocalBusiness or WebSite entities |
| AI crawler policy | ✓ | robots.txt verdict for each named AI agent, shown as a policy choice |
| Claims backed by visible sources | ✓ | Numeric and award claims without a nearby source |
| AI assistant visits | ◐ | With the read-only GA4 connection: visits from recognised assistants, shown as a floor |
| Prompt panel of buyer questions | ✕ | Not run; use the manual protocol above |
| Mention or citation tracking | ✕ | Not measured |
| CDN or firewall bot settings | ✕ | Not visible from public pages |

Source: PostCliq audit methodology, September 2026

The readiness checks answer "could an assistant read and trust this page". The manual panel answers "does an assistant name us". You need both, and a tool that claims to answer the second from the first is overreaching. The full scope sits on the [AI search visibility page](https://postcliq.com/ai-search-visibility) .

## What to do next

1. **Download the questions.** Get the [AI visibility check: ten buyer questions](https://postcliq.com/resources/ai-visibility-check-ten-buyer-questions) and adapt them to your category and location.
2. **Run the first panel this week.** Ten prompts, three assistants, fresh chats, one day. Fill in the scoring sheet.
3. **Read the "cited instead" column.** List the five sources assistants use most in place of you. Those are your next pitches, listings or partner pages.
4. **Turn on the free reports.** Verify your site in Bing Webmaster Tools for AI Performance, and check whether Search Console shows the generative AI report for your property.
5. **Fix the page basics.** Work through the [checklist for getting cited by ChatGPT](https://postcliq.com/blog/how-to-get-cited-by-chatgpt) , starting with crawler access.
6. **Run a** [**free PostCliq audit**](https://postcliq.com/#try-postcliq) for the AI readiness checks on your own pages and the issues that stop visitors becoming enquiries.
7. **Repeat in 30 days** with identical wording, and report the three-month trend, not the monthly number.

## Frequently asked questions

### Is there a free AI visibility checker?

Yes. Ahrefs offers a free AI Visibility Checker that reports mentions of a brand name across ChatGPT, Gemini, Perplexity, Copilot and Google AI Overviews, as a one-time snapshot. Bing Webmaster Tools' AI Performance report is free and shows Copilot citations. A manual prompt panel with ten buyer questions costs nothing but twenty minutes, and it is the only one that uses your buyers' questions.

### How do I check if my business shows up in Google AI Overviews?

Search your buyer questions in Google while logged out, in a private window, and note whether an AI Overview appears and whether it cites your pages. Search Console's generative AI performance report, where available, shows impressions from AI Overviews and AI Mode by page. It does not show clicks or the queries, so pair it with the manual check.

### How often should I check AI visibility?

Monthly is enough for most businesses. Answers change day to day, so checking weekly mostly measures noise. Run the same prompts on the same assistants each month, ideally on the same weekday, and judge the trend across three months. Check again after a major change, such as a site rebuild, a robots.txt edit or a new service page.

### Why does ChatGPT give different answers to the same question?

Language models generate each answer fresh, and when they search the web, the results and sources they retrieve vary with timing, location and account history. That is why one answer is weak evidence. Use fresh chats with no history, keep the wording identical, and treat your score as a sample. Doubling the prompts reduces the weight of any single answer.

### Can I improve my AI visibility quickly?

Some fixes are fast. If robots.txt or a firewall blocks AI search crawlers, fixing it can take effect within about a day for OpenAI's search systems. Rewriting page openings to answer questions directly takes weeks to be recrawled. Earning mentions on other sites, which correlates most strongly with AI visibility, usually takes months. Start with access, then content, then reputation.

### Does PostCliq track my brand in ChatGPT?

No. PostCliq does not ask ChatGPT or other assistants about your business, and it does not count mentions or citations. Its AI readiness checks look at your own public pages: whether text is readable without JavaScript, whether structured data is valid, how robots.txt treats each AI crawler, and whether claims have visible sources. Use the manual prompt panel for mentions.

## Ready to be cited? Check your pages free.

PostCliq checks whether assistants can read your pages, which AI crawlers you allow, and what stops visitors becoming enquiries.

Free. No credit card.

![Nirmal Gyanwali, founder and chief executive of WP Creative.](https://postcliq.com/authors/nirmal-gyanwali.jpg)

Written by

[Nirmal Gyanwali](https://postcliq.com/blog/author/nirmal-gyanwali)

Founder and CEO, WP Creative · Sydney, Australia

Nirmal Gyanwali has spent more than twenty years building websites, and most of that time watching them fail at the one job the business actually bought them for, which is turning a visitor into an enquiry. He founded WP Creative in Sydney in January 2016.

- [LinkedIn](https://www.linkedin.com/in/nirmalgyanwali/)
- [YouTube](https://www.youtube.com/@nirmalgyanwali)
- [Instagram](https://www.instagram.com/nirmalgyanwali)
- [TikTok](https://www.tiktok.com/@nirmalwpc)

## Keep reading

- [**Generative engine optimisation (GEO): a practical guide for marketing teams**](https://postcliq.com/blog/generative-engine-optimisation)
- [**How to get your business cited by ChatGPT: a practical checklist**](https://postcliq.com/blog/how-to-get-cited-by-chatgpt)
- [**Zero-click search: Google is answering the question, so fewer people click**](https://postcliq.com/blog/google-answers-the-question)

[More on AI search and visibility](https://postcliq.com/blog/topic/ai-search) · [All articles](https://postcliq.com/blog)