Get Recommended by ChatGPT Before Your Competitors Do
Your buyers ask ChatGPT who to pick. Docket's AI visibility tracking asks the same questions with live web search, shows whether you were named and who was named instead, and turns every lost answer into a brief for what to write next. Every figure comes from a real answer.
Is ChatGPT Recommending Your Brand? See Every Answer
A buyer who picked a rival because ChatGPT said so leaves no trace in your analytics. This is the trace.
Your Buyers' Real Questions
20 by default, drafted from your website's details and your own Search Console queries. Add your own, up to 100.Share of Voice, With a Range
A low and a high figure, not false precision. Plus mention rate and prominence.Answer Gaps, Turned Into Briefs
Where a competitor was named and you were not. Each gap becomes a content brief in one click.The Sources AI Trusts
Every cited source, with its title, for the questions in your category.Wrong Facts, Sorted
Hallucinations and factual inaccuracies about you, split apart and given a severity.A Method You Can Audit
Graded against the IAB's “Measuring Visibility in the AI Era” framework. The overall tier is the minimum, not the average.
How Does AI Visibility Tracking Work in Docket?
Docket asks ChatGPT your tracked questions with live web search, reads each answer for who was named and why, then turns the gaps into work. It is measurement for GEO, generative engine optimisation: being quoted by AI answers.
- 01 · Track
Pick the Questions
Generated from your website's details and real Search Console queries, across 8 intents: informational, comparison, recommendation, transactional, commercial, local, branded and problem. - 02 · Ask
Ask ChatGPT, Live
With live web search on, like public ChatGPT. If search fails, it asks without search and labels the answer. - 03 · Read
Read Every Answer
Named or not, and where. Every brand named. Sentiment, recommendation strength, endorsements with a checked quote, citations and issues. - 04 · Act
Turn Gaps Into Briefs
A lost answer becomes a brief in one click, drawing on the sources ChatGPT cited instead. It also feeds your topic scores.
| Setting | What Docket Does |
|---|---|
| Engine | ChatGPT, with live web search |
| Cadence | A measurement stands for 7 days and refreshes automatically. Refresh measures sooner |
| Questions | 20 generated by default (4 to 40 per generation), up to 100 tracked |
| Share of Voice | Reported with a low and a high figure |
| Reviews | Found through grounded web search. A review whose source page was not retrieved is thrown away |
Mentioned or Recommended? Docket Reads the Difference
Being mentioned is not the same as being recommended. Docket records how you were named, how warmly, and what the answer got wrong, so you know which problem you have.
- Sentiment: Positive, Neutral, Negative or Absent
- Recommendation: Unconditional, Qualified, Listed or None
- Endorsements Backed by a Verbatim Quote, Checked Against the Answer
- Issues Split Into Hallucination and Factual Inaccuracy, With Severity
Why ChatGPT Names a Rival, and What to Fix
Docket reads your site the way AI crawlers do: plain HTML, no JavaScript, robots.txt obeyed, up to 120 pages. Then it explains the gaps.
- robots.txt Checked Against 8 Live AI and Search Crawlers
- Questions Grouped Into Up to 8 Topics
- Each Topic Diagnosed From 12 Written Causes, Such as Marketing Language Instead of Verifiable Facts
- Each Recommendation Is Pre-Scored From 1 to 5, So the Highest-Impact Fixes Come First
- Docket AI
Tracked Share of Voice in ChatGPT, Not One-Off Screenshots
| Topic | The Usual Way | With Docket |
|---|---|---|
| Checking | The Usual Way: Someone asks ChatGPT once, screenshots it, and calls it research. | With Docket: Up to 100 tracked questions. Each reading stands 7 days, then refreshes automatically. |
| Reading | The Usual Way: Did we come up? Sort of? Moving on. | With Docket: Named and where, sentiment, recommendation strength, citations and issues. |
| Reporting | The Usual Way: One confident percentage with nothing behind it. | With Docket: Share of voice with a low and a high figure. |
| Wrong Answers | The Usual Way: Noticed when a customer mentions it. | With Docket: Hallucinations and factual inaccuracies flagged with severity. |
| Next Step | The Usual Way: A meeting about the screenshot. | With Docket: Each answer gap becomes a content brief in one click. |
Measurement You Can Stand Behind
Built so the numbers in front of your team are ones they can explain.
ChatGPT, Asked Live
Your questions are asked with live web search on, the way buyers ask, so each reading reflects what they see.Readings You Can Compare
Docket asks controlled, repeatable questions, so movement from one measurement to the next reflects the answers themselves. Pair it with Search Console and GA4 measurement.Ranges, Not False Precision
Share of voice comes with a low and a high figure, because the same question can be answered differently twice.Recommendations, Ranked by Impact
Every GEO recommendation is pre-scored from 1 to 5, so the strongest fixes rise to the top.Your Data, Handled Carefully
Read the security page for how Docket stores and protects your data.
Questions About AI Visibility
What is AI visibility?
AI visibility is how often an AI assistant names your business when someone asks a question you should be the answer to. Docket measures it in ChatGPT: whether you were named, where in the answer, who was named instead, which sources were cited and how you were described.
How does Docket check whether ChatGPT recommends my business?
Docket asks ChatGPT your tracked buyer questions with live web search on, like public ChatGPT, and reads each answer. It records whether you were named and where, every brand named, sentiment, recommendation strength, cited sources, and any hallucination or factual inaccuracy about you.
How do I get my business recommended by ChatGPT?
Docket helps you get recommended by ChatGPT by showing where it names a rival instead of you, which sources it cites and why, then turning each gap into a content brief. Articles written to Docket's GEO checks open with quotable answers and attributed facts, which makes them easier to cite, and the next measurement shows what changed.
Is ChatGPT recommending my brand or a competitor?
Docket's AI visibility tracking answers that for your own buyer questions. For every answer it records whether you were named and where, every competitor named, how strongly you were recommended and which sources were cited. Answer gaps list the questions where a competitor was named and you were not.
Does Docket track Gemini, Perplexity or Copilot?
Docket focuses on ChatGPT today, and Gemini, Perplexity and Copilot are not measured. Each question is asked in ChatGPT with live web search on, so every reading reflects what buyers actually see, and each answer is read in full: whether you were named, who was named instead and which sources were cited.
How often does Docket measure AI visibility?
A measurement stands for 7 days, and Docket refreshes it automatically before it goes stale. Refresh measures sooner. It is not a fixed weekly slot. Each measurement asks all of your tracked questions, up to 100.
What questions does Docket ask ChatGPT?
Docket asks the questions your buyers ask. It generates 20 by default from your website's details and your own Search Console queries, real queries first, and you can add your own, up to 100. They span informational, comparison, recommendation, transactional, commercial, local, branded and problem intents.
Does Docket catch ChatGPT getting facts about my business wrong?
Yes. Each answer is read for issues, split into hallucinations and factual inaccuracies, each with a severity, because the fixes differ. Endorsements must be backed by a verbatim quote that is checked against the answer, and a review whose source page was not retrieved is thrown away.
Can I check my AI visibility for free?
Yes. The free check at /preview needs no account. It reads your homepage, writes 3 buyer questions that do not name you, asks ChatGPT each one with web search, and shows whether you were named and which competitors appeared instead. Results are cached per domain for 6 hours.
See What ChatGPT Tells Your Buyers About You
On a demo we run your questions and walk through the answers, the rivals named and the sources cited.