Snowflake charges for cross-region data transfer between cloud provider regions. US-to-US region transfers cost approximately $20/TB; Asia Pacific region pairs reach up to $140/TB. Transfers within the same cloud region are free.
When this constraint is load-bearing.
- Multi-region data residency: Organizations with data residency requirements (e.g., EU data must stay in EU, APAC data must stay in APAC) often replicate Snowflake objects across regions for local analytics access. Each replication cycle incurs cross-region egress charges.
- CDW-to-activation sync: Composable CDP architectures that sync segments from a primary CDW region to regional activation endpoints (for latency reasons) pay egress on each sync cycle. At significant segment volumes, this can compound to material monthly costs.
- Healthcare and financial services: These industries are most likely to have hard data-residency requirements that force cross-region architecture, making this constraint a first-order cost consideration.
Contrast with AEP egress model. AEP's egress limits are quota-based per profile per year — when the quota is exhausted, exports fail (or require contract renegotiation). Snowflake's egress is cost-based — exports always succeed, but costs scale linearly with volume. For cost-predictability-focused organizations, the Snowflake model is more predictable than AEP's quota model but requires FinOps governance. For cost-minimization-focused organizations, AEP's quota model is bounded, while Snowflake's can escalate with architectural decisions.
ECO mitigation (April 2025): The Snowflake Egress Cost Optimizer uses intelligent caching to reduce cross-region replication cost by up to 96% for replication-heavy workloads. Organizations with replication-driven egress should evaluate ECO before sizing this constraint as prohibitive.