• Because NULL user_id values are deemed invalid traffic in post-impression server-side models.
• This would be redundant; records with NULL user_id values have been removed upstream of query processing anyway.
• Reach and frequency cannot be accurately determined for NULL user_id values, so this group is excluded to prevent NULL user_id values from skewing the results.

The correct answer is: Reach and frequency cannot be accurately determined for NULL user_id values, so this group is excluded to prevent NULL user_id values from skewing the results.

Explanation: According to Amazon’s official documentation, excluding NULL user_id values is a best practice for reach and frequency queries because reach and frequency cannot be accurately calculated on records without a user_id. Null values indicate that the user’s identity was not available, meaning their impressions or events cannot be attributed to any specific user, which would distort or inflate your metrics.

Reference content: From Amazon’s documentation playbooks and best practices – Optimal frequency analysis

“Ensure you exclude records with NULL user_id from your reach and frequency calculations. Reach and frequency must be based on known, distinct user_id values for accuracy.”