COF-C03 Question 337
Single answerData clean roomsA retail company and an advertising partner want to analyze overlap between their customer bases and measure campaign performance without exposing each other's raw customer-level data. They decide to use Snowflake Data Clean Rooms. During design review, the security team states that analysts from either company must not be able to directly query the other party's underlying tables, but both sides still need approved collaborative analysis results. Which approach best meets this requirement?
- A
Create a secure data share of the raw customer tables to both parties and rely on masking policies to prevent misuse
- B
Load both companies' customer tables into a shared schema and grant SELECT only to analysts who sign a data handling agreement
- C
Use Snowflake Data Clean Rooms so each party keeps control of its own data and only approved analysis templates or policies can be run to produce governed results
- D
Export hashed customer identifiers from both parties to external object storage, join them outside Snowflake, and re-import only the aggregated output
Show answer and explanation
Correct answer: C
Explanation
The best answer is to use Snowflake Data Clean Rooms to enable governed collaboration without exposing raw underlying tables. This is the core value proposition of clean rooms: multiple parties can collaborate on sensitive data while maintaining control over what analyses are permitted and what outputs are returned. In practice, this is more appropriate than raw data sharing, colocating data in a common schema, or exporting identifiers for external matching. For SnowPro Core, candidates should understand that Snowflake Data Clean Rooms are intended for privacy-enhancing analytics use cases such as audience overlap, attribution, and measurement, where partners need insights but should not receive unrestricted access to each other's source data. This aligns with Snowflake best practices around secure collaboration, governance, and minimizing raw data exposure.
- A. Incorrect.
Incorrect. Secure data sharing allows data to be shared without copying, but sharing raw customer tables still exposes direct table access to the consumer account. Masking policies can protect specific columns, yet this does not provide the same controlled collaboration model as a clean room, where approved analyses are governed and raw underlying datasets are not directly exposed for arbitrary querying.
- B. Incorrect.
Incorrect. Placing both datasets in a shared schema and relying on trust or agreements does not satisfy the requirement that analysts must not directly query the other party's underlying tables. This approach weakens data governance and increases the risk of unauthorized access or unintended disclosure.
- C. Correct.
Correct. Snowflake Data Clean Rooms are designed for privacy-preserving collaboration. Each party can retain control over its own data while allowing only governed analyses to run, typically through approved templates, policies, and clean room controls. This supports use cases such as audience overlap and campaign measurement without giving either side unrestricted access to the other's raw records.
- D. Incorrect.
Incorrect. Moving data to external storage and processing it outside Snowflake adds operational complexity and security risk, and it bypasses the native governance and collaboration capabilities Snowflake provides. Data Clean Rooms are specifically intended to avoid this kind of raw-data movement while enabling controlled joint analysis.