Cloud gets lean: ‘FinOps’ makes every dollar work harder
Enterprise cloud spend is growing, and using FinOps strategies can make each dollar work harder. Companies can save money, boost value, and build cross-functional cohesion.
Ben Stanton
Adam Gogarty
Paul Lee
Gillian Crossan
Ariane Bucaille
Global cloud spending will likely top US$825 billion in 2025, but ask an organization’s leadership what they spend, and it might be difficult for them to answer.1 Lots of companies either don’t know, or struggle to explain it.
As organizations increasingly rely on cloud services, the need for an effective strategy to manage cloud investments becomes paramount. Enter “FinOps” (a mash up of finance and DevOps), a set of strategies to help track and optimize cloud spending. Deloitte predicts US$21 billion may be saved by companies implementing FinOps tools and practices in 2025 alone, and this could grow in subsequent years. Some may even cut cloud costs as much as 40%. Going forward, we expect companies without a FinOps team to act swiftly to implement first steps, and we expect FinOps veterans to develop more sophisticated optimization strategies.
Cloud is complex, and complexity often creates waste
Cloud is now indispensable for enterprises. Spinning up new cloud environments often takes just a few clicks, whereas building private physical infrastructure can include procuring and installing servers and can take weeks or months to complete. It has helped democratize convenience and scalability. Cloud powers innovation at pace without an army of PhDs on staff; it helps enable industries like video on demand, ride sharing, challenger banks, and telehealth to disrupt their respective markets;2 and it underpins applications like data analytics, remote working, and of course AI.
Today, the organizations investing in software engineering are increasingly diverse. Commercial off-the-shelf products may no longer meet their needs. Automotive companies, like Daimler, have assembled teams of developers to build software platforms for electric vehicles.3 Even in industries without digital heritage, like wooden pallet distribution, there is also a growing need for custom software and expertise.4 All of this increases the cloud bill.
However, cloud is getting complex. Companies tend to juggle their private computing resources with public cloud services—a hybrid cloud infrastructure—that 73% of companies now have. Also, more than half (53%) of companies source cloud from multiple providers to take advantage of promotions, specific capabilities or avoid vendor lock-in.5 It’s also common for individual departments (think finance, HR, or marketing) to buy cloud software applications without the knowledge of the central engineering team. All of this can create complexity in areas like company data integration, compliance, and security.
In general, companies are not good at sticking to their cloud budget. Half of organizations surveyed overspent last year, and the average overrun was 15%.6 One factor is pay-as-you-go billing, which means cost is variable, and can make forecasting a greater challenge. In extreme cases, cloud engineers have inadvertently triggered thousands of dollars of cloud spend overnight.7
Cloud is not cheap; it can cost more than the equivalent private infrastructure,8 and is fast becoming a company’s largest IT line-item bill. Coca-Cola, for example, recently signed a $1.1 billion deal for its cloud needs.9 Crucially, though, 27% of spend is wasted, according to cloud leaders.10 Cloud customers have started to recognize this, and half now have a dedicated FinOps team, while 20% more should form a team within a year.11
Getting started with FinOps
FinOps is a financial management discipline. It can range from the technical, like rearchitecting cloud workloads and reviewing what can go into less accessible long-term storage, to much less technical, like negotiating discounts and credits. Its long-term impact, however, is cultural change. At its core, it’s about cross-organization responsibility and financial accountability, and aligning each cloud dollar spent with the business value it generates.
Starting with FinOps is all about planning: reviewing the current strategy, evaluating any tagging and alerting structures, and then defining key performance indicators.12 A practical first step is to focus on visibility, which can mean cataloguing current cloud resources and exploring how they may align with organizational needs. For this, cloud providers offer resource monitoring tools as well as specific tools focused on cost, or there are third-party FinOps platforms like that can provide more granular metrics.
However, interpreting dashboard data may require dedicated FinOps specialists and practitioners; who are often in-demand people. Also, companies with multiple cloud providers may need a dashboard for each. A single integrated portal can be challenging as data feeds from each provider will vary. Finally, FinOps tools can bear a considerable cost (around 3% to 5% of the cloud bill at the high end), so a company should understand their cloud economics before deploying one.
Starting out with FinOps: Preliminary measures
Organizations starting their FinOps journey will likely focus more on preliminary measures to help cut waste, ensure that resources are allocated optimally, review their contracts, and take advantage of potential credits and discounts.
Waste and consumption: We expect that FinOps novices will find getting rid of waste a great place to start. FinOps tools and dashboards might help a company pinpoint underutilized or idle resources to be right-sized or shut down, for potentially instant cost-savings. Examples of waste can include oversized virtual machines, redundant storage instances, orphaned resources, or duplicate data. The companies most adept at FinOps may use predictive analytics to forecast usage, and automated governance scripts that can dynamically adjust capacities. Plus, this can often be done by the central cloud engineering team, so the task can be completed quickly.
Structure and tiers: Cloud services are not all created equal. Computing and storage instances can span a range of qualities and price points. Companies could assess the caliber of their provisioning (allocating and managing cloud resources effectively), and whether it adequately fits the need of an application. It may be that some applications perform well on more cost-effective instances. For example, to help mitigate seasonal volatility, an event ticketing website may choose an instance which does not use the full CPU continuously, but occasionally needs to burst to align resourcing with web traffic.13
Incentives: Cloud platforms offer discount programs that can lead to substantial savings. Some platforms allow users to commit to a consistent amount of usage in exchange for lower rates. For some companies, directly renegotiating with their cloud service provider can be a fruitful approach. Cloud companies tend to be receptive to this, welcoming the chance to exchange discounted rates for multi-year contract commitments.
Getting serious: Refining with broader approaches
In 2025, experienced FinOps practitioners may continue to refine the above, but are also expected to advance their approaches to cost observability and control.
Accountability: As cloud is vital across all parts of many businesses, each department (and team) should be financially accountable for their spend. Departments should be given oversight and responsibility for costs that can be directly attributed to them, via either a chargeback model (charging a department directly) or “showback” model (showing a department its cost burden).14 This can require a robust tagging strategy, which assigns resource costs to specific teams or projects, ideally auto-tagging based on predefined rules. Ideally, this could create a culture where teams throughout the organizations feel engaged in cloud cost reduction.
On-prem: FinOps communities like the FinOps Foundation have started to encourage discourse around on-premises infrastructure, which is often opaque to users, as part of the overall equation.15 Companies should be thinking about cost across their entire IT estate. But this can be complex, as the central cloud team may need to liaise with branch offices and infrastructure sites and may find that they use a variety of hardware and software tools for local needs. Strategies for on-prem cost reduction can include cancelling redundant licenses and extending lifecycle of hardware.
Sustainability: FinOps also intersects with the growing “GreenOps” movement. GreenOps describes a set of cloud management strategies which optimize for sustainability. Granular metrics delivered by FinOps reporting tools can help with measurement of energy consumption, carbon emissions, and other sustainability goals.16 In the wake of major reporting regulations like the EU’s Corporate Sustainability Reporting Directive,17 tracking energy and carbon metrics and subsequently improving them can be an incredible by-product of investing in FinOps.
FinOps makes a tangible difference
FinOps practices have been instrumental for many companies working to achieve cloud cost savings:
- Airbnb:18 Travel app Airbnb generated a $63.5 million saving in cloud costs. Part of its approach was to shift storage to a lower-cost service tier and switch out its homegrown backup system for a cloud provider’s alternative.
- Sky Group:19 Media and entertainment company Sky discovered it had spent a full year’s cloud budget within six months. It deployed a first-party FinOps tool to identify $1.5 million in savings and implemented visibility dashboards, which enabled $3.8 million in savings in the subsequent year.
- The Home Depot:20 The home improvement retailer built a dedicated cloud cost team in 2022 and identified “tens of millions of dollars” in savings compared with the previous year.
- Lyft:21 Ride-sharing app Lyft cut cloud costs per ride by 40% in six months, with a spreadsheet-based software tool to track billing data made available to the entire company. This led to a wave of right-sizing programs.
- WPP:22 Advertising firm WPP saved $2 million after just three months of FinOps deployment, which eventually scaled to a 30% annual cost reduction on its yearly cloud spend. It leveraged a range of tools and techniques, including autogenerated sizing recommendations.
A broad range of organizations are now actively investing in FinOps. For example, members of the FinOps Foundation—a non-profit organization promoting best practices in cloud financial management—include Walmart, Mastercard, and American Airlines.23
Bottom line: Cloud unit economics
The emergence of FinOps reflects a need for better visibility, improved budgeting, and proactive control of cloud expenditures, which has continued to grow as organizations increase their reliance on cloud.
Going forward, global IT spending is set to rise (exceeding $5.1 trillion in 2025,)24 in part driven by digital transformation and AI. On top of that, private infrastructure still accounts for around half of all workloads, which, if migrated to public cloud, could significantly inflate the cloud bill. Additionally, high interest rates (at least compared to the last decade), have caused companies to focus on profitability and cost reduction, with a particular appetite to remove cost variability. In other words, the stage is set for FinOps’ growth.
FinOps should be viewed as a long-term practice, integral to operational strategy. It should not be seen as a simple fix. It starts with cost reduction, but it can eventually transform cloud spending from a mere line item into a strategic asset and enabler.
For some of the most advanced companies, the end-goal may be to create a “cloud unit economics” model. This approach quantifies the costs associated with each unit of cloud service used—per application, workload, or gigabyte of data processed—and aligns this with the resulting business metrics, such as revenue, cost per delivery, cost per booking, and cost per ride. This more granular insight can help companies make effective decisions about IT in the context of their whole business, helping to ensure each unit of spend is trackable to the bottom line.
For some companies, costs saved may be reinvested in new growth opportunities, such as scaling through new cloud services, or accelerating a product roadmap.
Cloud will always be complex, and it may never be inexpensive, but companies that can apply FinOps should make cloud more valuable to the bottom line than ever.
