What to Know About Cloud Data Warehouse Cost (and How to Fix It!)
Undoubtedly, the cloud has emerged as an ideal platform for a data warehouse – the tremendous success of Snowflake and AWS Redshift validates that. Cloud’s instant accessibility, scalability, and elasticity allow organizations to reduce the risk as they explore the value of their data and use it to measure performance, identify challenges, and find new growth opportunities. Data warehouses like Snowflake and Redshift have delivered tremendous value charged through usage-based pricing, where you only pay for what you use.
Cloud Investments Are Not Paying Off
Despite the clear benefits, a common concern among several early cloud adopters has been unpredictable and surging cloud costs to keep their data analytics running. Even Mike Scarpelli, the CFO of Snowflake – a company that has been incredibly successful with the usage-based pricing model – admitted that fast-growing companies couldn’t forecast their usage growth, and a subscription-based model can be much easier to budget for.
WSJ recently published an article highlighting that many CIOs have yet to see the ROI on cloud investments – and some, to the contrary, have seen their costs increase. “I must have had 10 conversations last week where CIOs were bemoaning that they have run out of money or blown their budget off,” said John Roese, Global CTO of Dell Technologies, in the same article.
And two weeks ago, AWS and Microsoft admitted that customers have realized their cloud costs are out of control, and helping them control costs will enhance long-term loyalty.
While the FinOps team in an organization does everything it can to slash cloud data warehouse costs, when the incremental savings from these initiatives hit the ceiling, orders of magnitude savings can only be made by changing the underlying architecture.
This is Where Yellowbrick Shines!
After years of perfecting our efficient, high-performance architecture on-premises, Yellowbrick now brings this to the cloud. Optimized for AWS, GCP, and Azure compute instances, organizations increase efficiency and lower cloud resource consumption, allowing you to do more with less cost.
When compared to the market leader Snowflake’s architecture, Yellowbrick delivers the highest ROI with:
- Granular Scaling: Due to Snowflake’s inefficient data distribution per node, when your workload runs out of compute resources, the only way forward is to double the nodes (aka t-shirt size), driving up cost. With Yellowbrick, however, you can incrementally add a node and keep your costs in control while getting linear scalability.
- High Concurrent Queries (and User) Support: Snowflake recommends that each cluster support up to 8 concurrent queries. If you need more, create more virtual data warehouses and direct additional queries elsewhere. Yellowbrick’s Direct Data Accelerator technology and intelligent workload management optimize cloud resources, supporting up to 150 concurrent queries per cluster.
- Extremely Low Cost Per Query: Yellowbrick compiles the query to machine code optimized for the OS, database, networking, and storage. This means we are often 1/5 of the cost of Snowflake but with higher performance.
- Adaptive Resource Allocation: Snowflake has no workload management. So, the only way users/workloads can keep impacting each other is to have a separate cluster. Yellowbrick’s intelligent resource manager optimizes the underlying compute resources and delivers multi-dimensional performance for both real-time ad-hoc and slow batch queries.
Snowflake praises usage-based pricing because you only pay for what you use. Their entire business model is built on encouraging greater consumption since this leads to greater business value. But if your architecture is inefficient and lacks basic workload management features to govern usage properly, then usage comes at a massive cost.
Want to Learn More About Reducing Data Warehouse Cloud Costs?
Register today for the LinkedIn Live panel: “For Crying out Cloud: Avoiding Poor ROI from Your Cloud Analytics” on Nov. 17 at 11 a.m. ET.
In this panel, Kevin Petrie, VP of Research, Eckerson Group, joins Yellowbrick’s Mark Cusack (CTO) and Heather Brodbeck (VP, RevOps) to learn how organizations can leverage operational and architectural efficiency to deliver the highest ROI for their cloud analytics initiatives.
Register for the panel HERE.
