Snowflake’s revenue grew 30% year over year in 2024. There’s no question that it’s a growing cloud–based data warehousing powerhouse. And with great growth, comes great services.
The only downside of those excellent computing, cloud, and storage offerings?
Confusing costs.
Users usually dedicate over 50% of their budget to Snowflake spend. So optimizing your storage costs is key to keeping your bill low.
This guide breaks down:
- How Snowflake’s storage actually works
- What goes into Snowflake storage costs
- Examples of how Snowflake charges for storage
- 5 pro tips on how to optimize Snowflake storage for better price and performance
Read on to cut down on your Snowflake storage spend and maintain current performance.
How Snowflake’s Storage Works
Before we get into the good stuff (a.k.a., cost savings), first you’ll need to understand how Snowflake’s storage works.
Unlike traditional data warehouses, Snowflake uses a patented “multi-cluster, shared data architecture” with three distinct but integrated layers:
- Storage layer: Automatically compresses and organizes your data into optimized columnar format micro-partitions, stored in your cloud provider’s blob storage (S3, Azure Blob, or GCP Cloud Storage)
- Compute layer (virtual warehouses): Independent clusters that process queries against the storage layer, scaling automatically based on workload
- Cloud services layer: Handles infrastructure management, metadata, security, query optimization, and transaction control
NOTE: Snowflake can usually achieve 3-5x compression, which directly improves storage costs.
This separation means Snowflake can independently scale each service based on your specific needs, resulting in better cost control and performance optimization.
How Snowflake Storage Cost Works
Good news! Monthly Snowflake storage costs are pretty straightforward. Your bill is based on a flat rate per terabyte of consumption – after everything is compressed.
For other Snowflake services, you’ll use Snowflake’s “credit” currency system, but when it comes to storage, you’ll be billed on a per TB or per month basis post-compression.
Your final Snowflake storage cost bill does vary based on:
- Region
- Cloud platform
- Duration of storage
- Data type
- Account type (Capacity or On demand)
Now, take a closer look at each of these variables to see how they influence your final bill.
Region
You can create your Snowflake account anywhere in over 35 different regions. You can use AWS, GCP, or Azure as your host. Picking outside of the U.S. usually leads to higher charges.
Cloud platform
Depending on your cloud platform (AWS, GCP, Azure) of choice, you might pay a little more or less for storage because your data might be kept in S3, Blob Storage, or Cloud storage.
For example, in the U.S., AWS storage costs are about $23 per TB per month. And remember: venturing outside the U.S. will leave you with higher storage costs.
Duration of storage
Though you’ll be charged for your data storage on a monthly basis, your final fee will be based on your data retention settings for Time Travel and Fail-safe settings – which we’ll discuss in more detail in a moment.
Snowflake does minimize storage prices here the best it can by only maintaining changes needed to restore data versions, not the full copies of tables.
Data type
The primary factor that influences Snowflake storage costs is the storage type you’re using. Take a close look at the most common types below.
Staged file storage costs
This is the cost that comes from storing files in Snowflake for bulk data loading or unloading. \Use a CSV or JSON format. You can see the estimated storage cost based on the file’s size after the file has been staged.
You don’t need to worry about data ingress fees, thankfully, but you should watch out for data egress. If you use Snowpipe or execute COPY or INSERT queries to load data, you’ll incur compute costs.
Database storage cost
This cost covers any data stored in Snowflake’s databases, including historical data used for Timetravel. Snowflake compresses all data stored in tables, optimizing Snowflake storage. That compressed file size is what’s used to calculate your total storage charges.
Time travel storage costs
Time travel storage is the perfect solution for an organization worried about accidental data deletions or truncations. It gives you access to historical versions of your data with a default retention period of either one day for the Standard Edition or 90 days for the Enterprise Edition, making it easy to revert or recall your original data.
Fail-safe storage costs
Similar to time travel storage, fail-safe storage gives you an additional seven days of data retention on top of Time Travel for absolute peace of mind.
The cost of fail-safe storage depends on the amount of data stored in multiple copies. You won’t get charged for fail-safe directly, but instead you’ll pay for the additional overall storage volume. Snowflake will only maintain full copies of tables after a table is dropped or truncated.
Zero-copy cloning cost
Zero-copy cloning occurs when you create a copy of a database, table, or schema and any data associated with it. Zero-copy cloning costs actually don’t occur until any changes are made to the cloned object because those changes create new physical data. And that new physical data needs storage space, which is what incurs your cost. This is because zero-copy cloning is a metadata-only operation.
Say you’ve just cloned a database. After you add a new table or delete rows from that database, you’ll incur a charge.
Account type
You can either subscribe to Snowflake services via On Demand or Capacity basis. With On Demand, you pay on a monthly basis. Capacity is purchased up front and a bit more expensive but allows for more flexibility.
On demand
If you’re looking at using Snowflake consistently for regular amounts of data, consider On Demand services. You’ll get less flexibility, but things are a bit cheaper compared with Capacity.
You can use On Demand services via popular cloud providers like AWS, GCP, and Azure. You’ll pay differently depending on brand, region, and tier within the brand.
Capacity Snowflake storage
Purchased capacity can be applied to any of your Snowflake services, so don’t feel limited to storage. The same as On Demand, Snowflake will compress and store your data on popular cloud platforms like AWS, GCP, and Azure.
Knowing this info, let’s look at a real-world example:
A mid-sized financial services company with 20TB of raw data might see:
- Raw data: 20TB
- After Snowflake compression: ~7TB (65% compression ratio)
- Monthly storage cost (US/AWS): ~$160/month
- With 14 days of Time Travel: Additional ~$75/month
- With Fail-safe: Additional ~$55/month
Total monthly storage: ~$290 (or ~$3,480 annually)
This shows how compression significantly reduces costs while Time Travel and Fail-safe add meaningful but manageable expenses for data protection.
5 Tips to Reduce Snowflake Storage Costs
Okay, so you know how Snowflake storage costs works, so now you can look at how you reduce that bill without hurting performance.
These are five tips that we use with our own customers to help them improve costs without sacrificing an inch of performance. The last one is our particular favorite, but we encourage you to try all of them.
Storage optimization tip #1: Use Snowsight
To do this, you’ll need to use an ACCOUNTADMIN account. Use these privileges to view account-wide and specific table data storage.
You can do this by using Snowsight or the classic web interface. To make things easier, we’re recommending Snowflake, going to Admin, Usage, Storage, and reviewing the average Snowflake data storage for your account.
Assuming you have the correct permissions, you can view the data storage from individual tables here. Do this by navigating to Databases, “db_name”, and then Tables.
If you don’t have any third-party tools that send alerts with storage usage spikes, you’ll want to go here on a semiregular basis so unexpected changes don’t catch you unaware.
Storage optimization tip #2: Review Snowflake storage history
You can also review historical storage costs using ACCOUNT_USAGE and ORGANIZATION_USAGE schemas. This will tell you about storage costs, how much data has been stored, the amount of time the data was stored, and any costs associated with storing that data.
While you’re there, review the storage history for your organization as a whole.
Navigate to the Admin section and select Usage. Go to the Usage Type drop1. -down and select Storage. You’ll need to pick a virtual warehouse – Snowflake recommends picking an X-Small – because you’ll need something that consumes compute resources when retrieving data.
Use this data to view your organization’s storage costs as a whole. Again, this is a dashboard you should review on a semi-regular basis if you don’t have a third-party tool that sends alerts if strange spikes occur.
Storage optimization tip #3: Filter usage by tags
You can also filter your organization’s data by different tags to attribute costs to certain organization units. Look for different projects, departments, or other logical units that may be inflating storage costs.
Storage optimization tip #4: Use SQL queries
Run these queries to identify your largest tables and optimization opportunities:
Find your largest tables and use this SQL code:
SELECT
TABLE_NAME,
ROUND(ACTIVE_BYTES/1024/1024/1024,2) AS ACTIVE_GB,
ROUND(TIME_TRAVEL_BYTES/1024/1024/1024,2) AS TIME_TRAVEL_GB,
ROUND(FAILSAFE_BYTES/1024/1024/1024,2) AS FAILSAFE_GB
FROM SNOWFLAKE.ACCOUNT_USAGE.TABLE_STORAGE_METRICS
ORDER BY ACTIVE_BYTES DESC
LIMIT 20;
Once you’ve identified large tables, consider:
- Implementing column-level clustering keys on frequently filtered columns
- Evaluating data retention policies for historical data
- Converting rarely-accessed data to lower-cost storage tiers
Storage optimization tip #5: Try a third-party tool
Manually reviewing your Snowflake storage works if you have the time or if you’re fine with checking performance on a quarterly or even yearly basis, but if you’re serious about optimizing your storage spend, you’ll want a tool that can automate the process.
You’re in luck: there are quite a lot of Snowflake cost optimization tools. It’s not a matter of if there’s a tool that can help, but finding the best tool for your organization.
The right tool can plug right into your system, no additional dev lift needed, and start automating warehouse sizes, alerting you to spikes in storage costs, and making your life as simple as a click of a button.
Maximize Snowflake storage cost ROI with Yuki
You don’t need to choose between performance and cost efficiency.
Yuki’s Snowflake optimization engine continuously monitors your storage usage, flags waste, and implements improvements automatically – no need to lift a finger.
Our users typically reduce Snowflake storage costs by 30–50% without impacting performance or workflows.
Request a free Analysis to see how Yuki can optimize your Snowflake storage spend.
