Energy crisis puts pressure on data centers: how FinOps and observability help reduce costs in the cloud

By Heber Lopes, Head of Products and Marketing at Faiston

In the last two years, electricity costs have skyrocketed and are directly affecting the operation of data centers OE in both public clouds and private infrastructure. In Brazil, the electricity bill should still have an increase of 6.3% by the end of the year, according to Aneel 5 above the expected inflation for the year.

This increase weighs on the budget of companies that depend on data centers, since electricity accounts for about 32% of their operating cost

nearly a third of the expenses of maintaining servers and cooling systems are directly linked to the energy tariff.The situation becomes more critical amid the expansion of the digital economy and a global energy crisis: in Europe and the United States, electricity demand has risen again after stagnant years, driven by the accelerated construction of data centers and other large consumers.

Given this scenario, Brazilian and global companies have increasingly tightened their belts and applying FinOps to regain control over cloud spending. In fact, at least 59% of companies already have dedicated FinOps teams to manage these costs, according to the 2025 State of the Cloud survey, published by SC Cloud.

And the most relevant point for a FinOps team is precisely the total visibility of the environment: it is necessary to identify where each penny spent in the cloud is going, assigning costs per project, department or application. Thus, FinOps implements metrics and transparent panels that show, for example, how much each application or team consumes resources and what the associated invoice (eliminating surprises and the small“”.

With the data in hand, comes the active optimization phase.Underutilized workloads are restructured or consolidated, adapting the provisioned capacity to the real demand (rightsizing practice). If servers are with low CPU or memory usage consistently, they can be migrated to smaller instances or shared virtual machines, reducing both cost and power consumption without affecting performance.“zumbis”, such as storage volumes without read/write or forgotten virtual machines, are shut down or removed as soon as identified.

Observability: key piece for FinOps

However, executing an effective FinOps strategy depends on continuous monitoring & this is where observability comes in. This concept offers a holistic view of the cloud environment, correlating performance metrics, event logs and transaction tracking. Modern tools can cross technical telemetry information with cost data, generating powerful insights: if a database account has skyrocketed 40% in a month, logs can reveal that a given poorly optimized SQL query is consuming excessive resources.

Similarly, metrics show whether a virtual machine is oversized by revealing that its average CPU utilization is only 20%. From these alerts, FinOps and engineering teams can make informed decisions whether to adjust instance size, optimize code, or shut down unnecessary services.Observability provides the raw data and technical context FinOps needs to act accurately.

And the synergy between FinOps and observability creates a virtuous cycle of continuous improvement. Each anomalous peak of use or cost captured enables corrective actions, and capacity rescheduling. This integration allows quick responses to financial incidents: as important as responding to technical failures is reacting to “cost incidents” situations in which a bug or misuse causes unexpected expenses.With granular monitoring, budget deviations that would previously only be perceived at the end of the month can now be identified and mitigated in days or hours.This ensures that the company pays for the performance it really needs to deliver, nothing more.

For leadership, the lesson of the energy crisis is clear: controlling spending in the cloud has become a matter of business survival. The period of “cheque blanks to scale applications has fallen behind ¡Now, each workload needs to justify its consumption with value delivery. Implementing FinOps is to create a culture of shared responsibility, in which technical and financial teams speak the same language and make decisions based on data. Already the observability ensures the fine visibility necessary to execute these decisions on a day-to-day basis, adapting capacity and correcting directions with agility. Together, these practices form the dynamic duo that transforms the cloud from a cost center into an unpredictable platform.