Data Costs Reduced by Half While Performance Rose

Our client operates in the highly competitive transportation sector, where profit margins are often slim. They encountered a rising trend in their monthly expenditures for cloud services and storage yet found it challenging to accurately identify what was driving these costs, making it difficult to implement effective optimization strategies.

55%

cost savings

Improved

cost visibility

Data-driven

decision-making

About the Client

Our client is a North American fleet management organization, offering comprehensive, connected solutions that meld intuitive design with seamless integration for businesses of all sizes.

Solution

We solved the client’s problem by scaling down computing power, implementing logic applications to regulate the number of active hours for data warehouses, and introducing a reserve capacity. Thanks to the innovative solutions introduced by our team of experts, significant cost savings were achieved without compromising their production efficiency.

Success Story

Industry:

Logistics

Technology:

Azure SQL | Power BI

Expertise:

BI and Analytics | Cost Optimization | Data Warehouse

Challenge Story

The Need to Reduce Data Costs While Maintaining Productivity

Our client works within the transportation industry, characterized by tight margins. They faced increasing monthly resource costs for their Azure Cloud consumption and storage but also struggled to pinpoint the sources of these expenses to effectively optimize them.

The client sought a solution to uphold their overall Azure performance while mitigating costs around managing their data. They came to Adastra with the main goals of wanting to assess resource usage, collect, process, and analyze data to find the most optimal way to manage their company and maintain their operations.

It quickly became apparent to the company that to maintain their operational excellence and manage costs more effectively, a significant upgrade in their data management practices was crucial. By turning to Adastra, they were taking a decisive step towards optimizing their operations and achieving their cost-reduction and performance goals.

Solution Story

Adastra’s Tailored Cost Optimization Assessment

Our experts devised a plan to solve the clients’ overspending on data issues, split into three phases:

Phase 1 – Assessment:

We completed a comprehensive assessment of the current Azure resources, databases and their usage patterns. This assessment was done to understand the databases’ status and identify all areas for improvement. We analyzed the structure, size, and queries, and bottlenecks to identify inefficiencies. This assessment formed the basis for the strategies used to optimize and enhance the functionality of all databases.

Phase 2 – Analysis:

During the assessment phase, we discovered some of the databases were overprovisioned. The most overscaled database was incurring a large monthly cost, while only reaching 20% of the max usage. Due to the nature of some of these databases, downscaling was not an option for all of them.

Phase 3 – Cost Optimization Implementation:

We began by optimizing the computer power of the client’s Azure SQL databases. For databases that could be downscaled, we adjusted the model to provide “just enough” power for queries to run smoothly. This reduced monthly computer power costs by 90%.

For the databases that couldn’t be downscaled, Adastra created customized solutions tailored to the type of usage. Development data warehouses do not need to run 24/7 so we created a workaround by adding a logic app that identifies when the warehouse is being used so it can pause it via API request.

Data Costs Reduced by Half While Performance Rose

[Figure 1]: A custom logic application to identify opportunities to reduce runtime for databases that can’t be downscaled.

For the second type, involving a specific data warehouse used for a refresh of a Power BI dataset, the solution was very similar, and we implemented a logic app making the data warehouse available only for a refresh of the dataset.

Data Costs Reduced by Half While Performance Rose

[Figure 2]: A custom logic application to identify opportunities to reduce runtime for a data warehouse used once a day for Power BI.

The client’s production data warehouse could not have any downtime, so we found the simple and cost-effective solution of implementing a reserved capacity for the resource. This will lead to a 37% cost reduction after 1 year.

Data Costs Reduced by Half While Performance Rose

[Figure 3]: Using reserved capacity to cut costs on a data warehouse with no downtime. 

The final step was helping the client gain better insights into their costs to allow them to run metrics not supported by the Azure portal cost dashboard. Adastra converted the files from Azure Cost Management to external tables within the data warehouse. This enabled the client to use these tables to visualize the data in Power BI, gaining a view of cost breakdowns by resource for each month.

Data Costs Reduced by Half While Performance Rose

[Figure 4]: A sample dashboard providing insights into the organization’s costs.

Benefits Story

Strengthening the Client’s Data Infrastructure: Maximizing Visibility, Savings, and Production Efficiency
The solution empowered the client to extract maximum value from their data, resulting in cost savings while maintaining production. The client now has visibility into their data spending with detailed reports in Power BI.

Other benefits of the engagement included:

  • 55.6% decrease in average daily costs
  • Improved operations
  • The ability for the client to create their own cost management reports