Ongoing AI visibility monitoring: brand mentions and citations in ChatGPT, Perplexity, Gemini and Google AI
Which AI platforms do you monitor, and how often?
| Platforms | ChatGPT, Perplexity, Gemini and Google AI Overviews |
| Frequency | Every month as standard; weekly monitoring on request, quoted separately |
| Questions tracked | 20 on GEO Starter, 40 on GEO Growth |
| Competitors tracked | Top 3 on GEO Starter, top 5 on GEO Growth |
| Report | Written monthly report; GEO Growth adds a strategy memo |
Which metrics do you report?
Every metric is calculated from the same fixed set of questions, so you can compare month to month.
| Metric | Definition |
|---|---|
| Mention rate | The share of tested questions where the AI names your business in its answer |
| Citation rate | The share of answers that use one of your own web pages as a source |
| Cited URLs | Every page the AI used as a source for each question, yours and other sites’, so you can see which sites matter in your industry |
| Competitor share | Of all the business names in the answers, the share that goes to you and to each tracked competitor |
| Sentiment | Whether the AI describes your business positively, neutrally or negatively when it mentions you |
| Position | How high your business appears when it is named |
| Accuracy | Whether the AI states your prices, location and services correctly |
| CitedAsia Score | One score from 0 to 100 that combines mention rate, position, coverage across AI tools and accuracy |
| AI referral visits | Visits to your website from AI tools such as chatgpt.com and perplexity.ai, from your Google Analytics |
| Recommended actions | The changes we suggest for next month, in priority order, based on what moved and why |
What does a monthly report look like?
An example of the summary page, with illustrative figures rather than a real client:
| Metric | Last month | This month |
|---|---|---|
| Mention rate | 25% | 40% |
| Citation rate | 10% | 20% |
| Competitor share (you) | 18% | 27% |
| Sentiment when named | 4 positive, 1 neutral | 7 positive, 1 neutral |
| Accuracy errors found | 3 | 1 |
| CitedAsia Score | 32 | 48 |
Below the summary, the report lists every question with the answer from each AI tool, the URLs cited, the changes we made that month and the recommended actions for next month.
How do you run the tests?
With our own testing tool, checked by a person before every report. The tool asks each question on each platform in a fresh session, so earlier chats do not change the answer, and saves the full answer, the sources cited and the date. A person then reviews the results, checks the facts and writes the recommended actions.
How is the question set chosen?
- We start from what your customers really ask: your enquiries, your FAQs and the searches in your Google Search Console.
- We cover five types of question: shortlist, price, specific need, comparison and brand check.
- We write each question in your customers’ own words, in English or Chinese.
- You approve the list in writing before monitoring starts.
- The list stays fixed so results stay comparable; any change is agreed with you and noted in the report.
See how the audit tests questions for examples.
How much does monitoring cost?
Monthly monitoring is included in GEO Starter (USD 1,200 a month) and GEO Growth (USD 2,200 a month). Both have a 3-month minimum. Weekly monitoring is quoted separately. See all prices on our pricing page.
Frequently asked questions
Why monitor every month and not every day?
AI answers vary from one test to the next, so a single day says little. A fixed set of questions tested every month shows the real trend. Weekly monitoring helps around a launch, a rebrand or a correction project.
Why can the same question give different answers?
AI tools generate each answer fresh and update their sources often. That is why we test in fresh sessions, record the date and look at rates across many questions rather than one answer.
Can we change the questions or the competitors?
Yes. We agree any change with you in writing and note it in the report, so you can still compare results over time.
Do you monitor other AI tools, such as Claude or Copilot?
Not at the moment. We monitor ChatGPT, Perplexity, Gemini and Google AI Overviews.
Who owns the monitoring data?
You do, once it is paid for. Reports and test results belong to your business.
Updated November 2026 · Written by Dawn, Founder of CitedAsia · AI visibility audit · GEO services for Singapore SMEs · Measuring GEO ROI · Case studies
Find out where you stand today
Get your free AI visibility score in 30 seconds, then a full PDF report with 5 customer questions tested within 24 hours.
Is AI visibility monitoring a managed service, or do you only provide software?
It is a managed service. CitedAsia runs the tests for you with our own testing tool on ChatGPT, Perplexity, Gemini and Google AI Overviews. A person reviews every result, checks the facts and writes the recommended actions before each written report. Monitoring is monthly as standard, with weekly monitoring available on request, and is included in GEO Starter and GEO Growth.
What is the difference between a mention, a citation and a recommendation?
A mention is when the AI names your business in its answer. A citation is when the AI uses one of your own web pages as a source. A recommendation is when the AI suggests your business as an option for the customer’s need. Our reports keep these three apart, because a business can be cited without being recommended.