COVERAGE, WITHOUT THE GUESSWORK
Planned platforms. Clear status.
We’re validating collection methods and access. Listing a platform here does not mean a working integration or endorsement.
No collection integrations are live yet. BrandSeen’s walkthrough is a local simulation using fictional data.
| Platform | Status | What remains to validate |
|---|---|---|
| ChatGPT | Planned | Permitted access, collection method, evidence and costs |
| Claude | Planned | Permitted access, collection method, evidence and costs |
| Gemini | Planned | Permitted access, collection method, evidence and costs |
| Perplexity | Planned | Permitted access, collection method, evidence and costs |
| Grok | Planned | Permitted access, collection method, evidence and costs |
| Muse | Under evaluation | Exact product/surface and permitted access |
| Dot | Under evaluation | Exact product/surface and permitted access |
An observation needs a method
API responses can differ from consumer chat interfaces. Before a platform becomes available, we need to explain the source, collection method, known model/version details, regional limitations and how failures are handled.
Collection is different from an assistant connection
Measuring a platform’s answers is different from letting an assistant operate BrandSeen through an API or MCP integration. Using a model internally does not make it a supported observation source.
No substitutions disguised as evidence
If a collection fails on one platform, another platform’s response cannot be labelled as its result. A completed answer without a mention is also different from a failed request.