Warehouses Compared for Analytics Engineers
Warehouses compared looks at how Snowflake, BigQuery, Databricks SQL, Redshift and DuckDB differ in architecture, pricing model, SQL dialect and typical use. Knowing the differences helps you choose a platform, write portable SQL and predict cost and performance behaviour.
Stage: Advanced and Modern Topics · Part of the Analytics Engineer roadmap · Last reviewed 19 November 2026
What to learn in Warehouses Compared
- Describe how each platform separates storage from compute
- Compare pay-per-query pricing with provisioned compute pricing
- Translate a date function between two SQL dialects
- Explain when DuckDB suits local development and testing
- Choose a warehouse for a given team size and workload
Warehouses Compared interview questions
How can pricing models change how you write SQL?
With per-bytes-scanned pricing, selecting fewer columns and filtering partitions directly lowers cost. With compute-time pricing, query duration and warehouse size matter more. Optimise for whichever the platform charges.
What is DuckDB good for in analytics engineering?
DuckDB is an in-process analytical database that runs locally and reads files such as Parquet. It is handy for fast development, unit testing SQL logic and small-scale analysis without a cloud warehouse.
Where Warehouses Compared fits in the Analytics Engineer roadmap
Warehouses Compared belongs to the Advanced and Modern Topics stage (advanced level), which takes about 4 hours. Go further.
Other units in this stage
Back to the full Analytics Engineer roadmap