How to Calculate Snowflake SaaS Pricing in PostgreSQL

What components make up Snowflake SaaS pricing?

Snowflake bills on three levers: compute credits (warehouses & tasks), storage (compressed bytes stored), and data transfer (egress to the internet or another region). Understanding each lever helps you build cost models that are easy to query and forecast.

How do I store Snowflake usage data in PostgreSQL?

Export the views METERING_HISTORY and STORAGE_USAGE_HISTORY from Snowflake to CSV or Snowpipe-into-S3, then COPY the files into a PostgreSQL table named snowflake_usage. This table will drive every cost query you write.

CREATE TABLE example

CREATE TABLE snowflake_usage (
usage_date date,
resource_type text, -- COMPUTE, STORAGE, TRANSFER
credits_used numeric,
bytes bigint,
region text,
comment text
);

How do I calculate compute costs?

Multiply credits_used by your contracted credit rate. Storing that rate in a config table lets you change pricing without rewriting queries.

SELECT usage_date,
SUM(credits_used * cr.credit_rate_usd) AS compute_cost_usd
FROM snowflake_usage su
JOIN cost_rates cr ON cr.resource_type = 'COMPUTE'
WHERE resource_type = 'COMPUTE'
GROUP BY usage_date
ORDER BY usage_date;

How can I estimate storage costs?

Snowflake charges per average compressed byte-hours. Convert bytes to GB-months, then multiply by the storage rate.

SELECT date_trunc('month', usage_date) AS month,
ROUND(SUM(bytes) / 1024^3 / 24 / 30, 2) AS avg_gb,
ROUND(SUM(bytes) / 1024^3 / 24 / 30 * cr.credit_rate_usd, 2) AS storage_cost
FROM snowflake_usage su
JOIN cost_rates cr ON cr.resource_type = 'STORAGE'
WHERE resource_type = 'STORAGE'
GROUP BY month;

How do I break down cost by customer or workload?

Tag Snowflake warehouses with the same customer_id as your Customers table. After ingest, join the tag to customer metadata.

SELECT c.name,
SUM(su.credits_used * cr.credit_rate_usd) AS total_cost
FROM snowflake_usage su
JOIN customers c ON c.id = su.comment::int -- tag holds customer_id
JOIN cost_rates cr ON cr.resource_type = su.resource_type
GROUP BY c.name
ORDER BY total_cost DESC;

Best practices for controlling Snowflake spend

Use auto-resume/auto-suspend on warehouses, schedule warehouse sizes by workload, and archive infrequently used tables to cheaper storage tiers. Monitor daily costs with the queries above and automate Slack alerts when spend spikes.