Does the brand appear?
See whether the brand is mentioned, where it appears and how AI describes it.
AI Visibility DiagnosisSee whether AI mentions you—and how it describes you
Start from the questions customers really ask. Review answers across AI platforms and devices, see whether the brand appears, where it ranks and which sources AI cites.
What we check
We only count results with a saved original answer; failed checks and missing records are excluded.
How does it compare with competitors?
Use the same question to compare how the brand and competitors are introduced.
Which pages does AI cite?
See which sources AI references and whether the brand information in them is accurate.
Is any information wrong?
Find outdated, incorrect, contradictory or negative brand descriptions.
Keep the original answer
Save the AI platform, device, question, time, original answer and screenshot.
List the next changes
Turn confirmed issues into material updates, content, publishing or re-check tasks.
AI answers change over time and between model versions. Without comparable original records from before and after, we do not claim a specific improvement.
What happens after the check
A diagnosis is not just a report; it shows what information to add and what content to create next.
- 01Confirm the brand and goals
Confirm the company, products, region and questions to review.
- 02Set the review scope
Choose AI platforms, desktop or mobile, questions and competitor brands.
- 03Save and review results
Keep original answers and screenshots, and exclude failed records.
- 04Agree the next plan
Only act on issues supported by an original answer.
- 05Re-check under the same conditions
Compare results before and after to see what changed.
What is agreed before a brand diagnosis?
This is a scoped diagnosis service requiring materials and confirmation, rather than an instant self-service score.
Agree on questions and platforms
Define the brand and aliases, audience, Xiamen or other target region, platforms and entry points, including whether web search is enabled.
Check the report contents
Keep each exact question, full answer, query time, platform/model, region, device, cited URLs and screenshot. Separate valid, failed and unknown samples.
How should the displayed case be used?
The homepage manufacturing example illustrates information organization and report structure, not an expected diagnosis score. Publish client identities, original records and changes only after permission and verification.
Content updated：
Scope, delivery and review
Confirm the scope first, then check each completed deliverable against its original evidence.
What is needed before diagnosis?
Provide the brand and aliases, website, target region and audience, real products and agreed competitors. Separate brand-prompted questions from discovery questions that do not name the brand.
What belongs in the original record?
Keep the exact question, platform/model identifier, collection time, device or entry point, full answer, cited URLs and screenshot. An answer without the brand is also a valid observation; do not select only successes.
How are failures handled?
Record failures, missing answers, absent citations and successful collection separately, with planned and valid sample counts. Unknown results are neither brand absence nor successful checks.