Trust & compliance

The trust foundation brands can buy with confidence.

Real consumer research is only valuable when it is built on consent, protection, and quality. Here is what we commit to participants and customers.

Consent-based collection

Every study begins only after a participant gives consent for that study — joining the community does not mean consenting to all research.
Participants can pause, leave, or withdraw consent at any time, and can contact us about requests related to their own data.
Before starting, participants clearly see who is running the study, the task, what is and is not recorded, and the reward rules.

Data that enters analysis is desensitized — brands never see a participant’s identity.
Customer-facing reports are presented in aggregate, anonymized form and do not point to any identifiable person.
Sensitive content is filtered out by established rules and is not included in research records.
We combine experience metrics across products into anonymized, aggregated category benchmarks so you can see where you stand against similar products. A benchmark is produced only when enough products contribute, contains statistical percentiles only, and includes no identifiable product, customer, or personal information. Using the platform constitutes agreement to participate in anonymous category benchmarks in this way.

We design our data handling around the shared principles of China’s PIPL and the EU’s GDPR, following data minimization, purpose limitation, and traceability.
Access is role-based, actions are logged, and the process is traceable and reviewable.
These are built-in mechanisms; we can provide compliance details tailored to a specific procurement process on request.

Business in each region runs in line with local requirements, and data is managed along regional boundaries.
The China region is operated by ComRatings (Shanghai Fuwang Info Tech.) under license from OneSource AI, and China-region customers contract with ComRatings.

Every report is automatically quality-checked before delivery, so conclusions rest on validated data.
Data that fails checks or is incomplete does not enter analysis — no decisions on bad data.

For compliance questions, contact support@onesourceai.net