Improve Snowflake cost visibility with Capital One Slingshot

As organizations’ scale their cloud infrastructures, managing consumption and spending across multiple users, accounts, teams and services becomes increasingly complex. Gaining a clear view of cloud spending across the organization helps companies understand where spend occurs so they can make informed business decisions. 

Capital One’s Slingshot is a data management solution that offers Snowflake customers greater transparency into their Snowflake costs and usage. Through a suite of dashboards that provide detailed breakdowns and insights, Slingshot helps businesses gain a deeper understanding of the organization’s usage to optimize cost and performance in Snowflake.

With cloud adoption, enterprises have access to an unprecedented level of scalability and flexibility. As organizations grow, cloud costs can quickly balloon and data inefficiencies can further drive up expenses. A comprehensive view of spending across all accounts is key to helping businesses gain predictability and effectively manage costs as they scale.  

While Snowflake customers can segment data on cost and usage by account, Slingshot goes deeper on those metrics by combining data across accounts to create cost allocations that capture total spend holistically. Slingshot’s cost insights are displayed in dashboards preconfigured for each tenant. They give companies visibility into the key metrics impacting costs, performance and usage in Snowflake. Once an organization can see its unique patterns for how it spends and consumes on Snowflake, it can make more informed decisions about where to allocate resources and where to focus on cost savings.

Slingshot’s detailed dashboards help organizations pinpoint the most expensive warehouses, databases and queries. They offer a granular view into costs that allows organizations to identify, for example, warehouses with autosuspend disabled, repeated queries with high costs and unused materialized views that can be removed to save on storage costs. In Slingshot, total costs can be viewed in the cost analysis dashboard in four ways:

• Total costs: For a high-level understanding of costs, the total costs tab presents enterprise-wide costs and projections in summaries by day, week or month within the contract period. Organization in the dashboard can view total costs by account and usage type. The dashboard also displays projections on enterprise-wide costs, warehouse usage and cost, and database costs.

• Warehouse costs: Slingshot customers can get an in-depth view into their warehouse costs by day, week, month, account and environment. They also gain insight into the costliest warehouses, users and queries along with warehouse costs by department. Machine learning models generate warehouse cost projections for the contract period by comparing past costs with trends.

• Storage costs: This view organizes database expenses for cloud storage managed by Snowflake and includes details like data transfer cost by account and the costliest databases listed in descending order. 

• Serverless costs: Slingshot also helps organizations track costs associated with resource allocations and pipelines. Included in this tab are auto clustering costs, which are costs from Snowflake automatically adjusting resources, by account and data pipeline costs by account.

As the use of Snowflake grows across lines of business, tracking costs and usage across these accounts becomes more complex and challenging. Allocating Snowflake costs by team or business unit encourages accountability and enables chargebacks. Using the Cost Breakdown Report, which can be found in the cost analysis dashboard, admins can view cost attributions on shared resources across accounts and lines of business. This is accomplished through customizable tags, which each business can create and assign to their warehouses to track spend and generate reports in a way that makes sense for the business. By tagging a warehouse by team, department or project, for example, the cost analysis dashboard provides visibility into expenses in a way that supports more efficient usage and data-driven resource allocation.

Slingshot helps organizations optimize Snowflake warehouses by drilling down into usage (e.g., how many warehouses are running), warehouse performance (e.g., identifying data spillage) and contract analysis (e.g., average daily usage cost under a current contract). 

Slingshot’s dashboard gives finance teams and database administrators transparency into how resources are being used and where there is an opportunity to right size to avoid overutilization. Detailed cost breakdowns of warehouses mean organizations can compare expenses, track trends and identify warehouses that are over- or under-provisioned. These insights help teams decide how to allocate resources efficiently, in a way that optimizes spend without compromising performance.

Planning for cloud costs allows businesses to weather unanticipated demands and forecast spending more confidently. But to do so, organizations need a clear view into trends and leading cost drivers. With Slingshot, Snowflake customers can visualize usage patterns and cost spikes that may vary by season or quarter, helping them to make informed decisions for financial planning and sustainable growth in their cloud infrastructure. For example, as organizations plan for resource-intensive initiatives like machine learning model building or running large, complex queries, they can use Slingshot’s cost insights to track historical spending on similar projects and forecast future expenses.

With Slingshot’s cost insights and dashboards integrated into their workflows, Snowflake’s finance and administrative team members can optimize cloud costs and allocate resources to the benefit of the business while delivering insights needed for forecasting. With these insights in hand, stakeholders can identify areas where the cloud benefits outweigh the costs, and vice versa, ensuring the business is receiving the best value out of its cloud investment.