When this beats the AWS analytics stack

AWS and GCP both have working analytics stacks. The right one depends on team familiarity, ecosystem already in place, and the specific workload. The course is honest about both: this is when GCP plus dbt wins, and this is when it loses.

When GCP plus dbt wins: SQL-heavy analytics teams, Looker shops, teams already on Workspace, on-demand billing fits sporadic workloads, you want zero operational overhead on the warehouse layer.

When AWS plus Iceberg wins: hybrid Snowflake plus Athena access patterns, S3 already holds petabytes of source data, multi-engine reads (Spark, Athena, Snowflake) over the same files, strict cost discipline that prefers S3 storage tiers.

Multi-cloud is rarely worth it for analytics. The complexity tax is real and the wins are mostly imaginary unless you have a specific compliance or vendor-bargaining reason. Pick one. Get good at it. Switch only if the workload demands it.