Do you know how to optimize database costs by query?

You open your Azure bill and see a large monthly spend on Azure SQL or SQL Server, but no clear explanation of why. There are dozens (or hundreds) of databases, and you’re unsure which application, query, or piece of code is driving the cost. Without visibility at the query level, optimization becomes guesswork.

Figure: Dashboard showing highest-cost SQL queries

Get visibility into query-level costs

When managing cloud databases, especially in Azure, cost is directly tied to resource consumption:

• CPU time
• IO usage
• Execution frequency
• Query duration

Instead of only looking at server-level metrics, use a FinOps-focused tool that shows:

• Cost per database
• Cost per query/statement
• Top 10 most expensive statements
• Percentage of total spend by query

Tools like Fortified’s WISdom allow you to drill into query-level spending and identify which SQL statements are responsible for the biggest share of your bill.

Step 1: Focus on “Best Bang for Buck” Queries

Not every expensive query should be optimized first. Prioritize based on:

• High hourly cost (e.g., $4/hour for a single query)
• High CPU consumption
• High execution count
• High percentage of total environment cost

Click into the query details and inspect:

• The SQL statement being executed
• The application sending it
• Execution metrics (CPU, reads, duration)

Once identified, locate the source in your codebase and assess optimization options.

Step 2: Optimize the Query

Common optimizations include:

• Adding the correct index
• Rewriting inefficient joins
• Reducing unnecessary columns
• Analyze execution plans
• Eliminating redundant executions

A single well-placed index can dramatically reduce CPU and execution time.

It’s not uncommon to achieve performance improvements of 80–90% for poorly optimized queries.

Step 3: Calculate and Communicate ROI

The most powerful part of query-level cost analysis is demonstrating real financial impact.

Use ROI planning features in FinOps tools to:

1. Estimate expected performance improvements
2. Model reduced CPU or execution counts
3. Calculate projected savings

For example:

• Performance gain: 89%
• Estimated savings: $500/year from a single query change

Now you can clearly communicate value:

“If I spend time adding this index and optimizing this query, we can save $500.”

Or even better:

“We implemented this fix and reduced costs by $500.”

This turns performance tuning into measurable business value, not just technical improvement.

Step 4: Make Query Cost Review Part of Your Process

To consistently control database costs:

• Review top costly queries monthly
• Include query cost analysis in performance testing
• Track cost impact after deployments
• Align DB optimization with FinOps practices

Database optimization is no longer just about speed, it’s about cloud cost efficiency.

Summary

If you’re running Azure SQL or SQL Server in the cloud and your costs are growing:

• Drill down to the query level
• Identify the top cost-driving statements
• Optimize the highest ROI candidates
• Quantify and report savings

When you optimize by query you gain control over both performance and spend.