In our experience, the primary driver of Snowflake costs is not the compute for aggregation, but the compute required for lots of reads/scans.
We recently built a Snowflake-to-ClickHouse pipeline for a customer where aggregates are built hourly in Snowflake, then pushed into a ClickHouse table to power their user-facing dashboards.
By offloading dashboard queries to ClickHouse, they slashed their Snowflake bill by ~25%, which was worth millions to them.
(Admittedly, running aggregations elsewhere—for example, in Spark—could further reduce costs, but you would then need Iceberg to make the tables queryable in Snowflake.)