Picture this: you’ve spent a few weeks building a business case for a FinOps practice. You think you’ve covered the bases. Your CFO asks: “What’s the ROI on FinOps?”

Most teams, and possibly yours, answer with a savings number.

That’s a trap!

Not because savings don’t matter, they most certainly do. But because FinOps isn’t a “let’s save 30% initiative”. It’s an operating model to govern variable technology spend, across cloud, SaaS, on-premises and now AI. The model is there to allow businesses to make faster, trade-offs with clearer accountability and fewer financial surprises. That’s literally baked into how the FinOps Foundation defines it:

“business value, decision making, financial accountability, and collaboration across engineering, finance, and business”

In 2026, the CFO questions are getting sharper, while the spend is getting weirder.

Gartner forecasts worldwide public cloud spending at $723.4 billion in 2025. GenAI spending alone is forecast to hit $644 billion, up 76% from 2024. On the other hand, Flexera’s 2025 State of the Cloud reports that cloud spend is set to increase by 28% this year. Not only that, organisations are exceeding their budgets by 17% while cloud waste continues to hover around the 27%.

That’s a lot of numbers! So when your CFO asks, “What’s the ROI”, what they are really asking is:

If we fund a FinOps practice, will we get better control and outcomes, not just a smaller cloud bill?

That’s the theme I’ve consistently taked about. FinOps isn’t cost cutting. It’s capital allocation discipline. And when you throw AI workloads into the mix you need to turn your cloud spend into an investment management portfolio problem, not just a forecasting exercise.

Do you want that business case to pass muster with the CFO? Let’s use a calculator that turns assumptions into defensible numbers.

The Multi-Dimensional FinOps ROI Model

Measuring FinOps by “savings” alone does it a disservice. You’ll systematically undercount its value.

A CFO ready model needs four value dimensions: direct savings (waste elimination plus commitment optimisation), velocity gains (faster delivery, fewer finance/engineering deadlocks), risk mitigation (budget blowouts prevented, governance guardrails, anomaly response), and innovation enablement (safe experimentation runways, better investment decisions, fewer zombie initiatives).

That’s not fluff. It matches the FinOps Foundation’s framing: “trade-offs between speed, cost and quality; guided by business value”. The calculator is just the mechanism that forces organisations to do what it wants to avoid: turn opinions into assumptions and assumptions into numbers. The language of business.

Dimension 1: Direct Savings (The Obvious One, but Still Done Poorly)

Let’s start here. It’s the low-hanging fruit: measurable, fast and it funds the rest. (We need to find the money somewhere!)

Industry data consistently shows organisations leaving money on the table. Around 27% - 32% of their cloud spend. This on idle resources, oversized instances and commitment gaps. Deloitte’s TMT Predictions for 2025 estimated that mature FinOps practices will unlock $21 billion in cloud savings globally this year alone.

That number again: $27 billion in cloud savings globally this year alone.

You don’t need to argue where your organisation sits, above or below the benchmark. What you need is a CFO grade method to estimate your addressable portion. Your slice of this giant pie.

The Calculator Method

1. Establish your baseline: Use the trailing 3 to 6 months of cloud spend, split by major service families.
2. Identify addressable waste groups: e.g. idle or abandoned resources, over provisioned compute database resources, commitment and discount underutilisation etc. This is not a holistic list, but you get the idea.
3. Estimate addressable percentage: Be conservative with this and add the realisation curve (1 to 3 months, 3 to 6 months etc.), and the owner. Don’t underestimate the last point. The owner matters more than most people care to admit

The CFO-friendly version isn’t “we’ll save money by doing this”. It’s an annualised savings range (low / medium / high), confidence level, dependencies and more importantly how these savings will be sustained. One-off clean ups tend to decay. Fast.

Here’s a reality check for you:

McKinsey’s analysis of $13 billion in cloud spend found that waiting until you hit your target annualised cloud spend to implement FinOps is invariably “too late”. Waste compounds. Always. Every month of delay means an additional 2%-3% additional waste baked into your baseline, making it progressively more difficult to remove.

Remember, Savings without an operating rhythm is just a one-off spring clean. Clutter always has a way of finding its way back.

Dimension 2: Velocity Gains (The One Finance Underestimates)

This is where most ROI models get lazy. They either claim miracles in productivity gains with no evidence or use the good old “finger in the air” guesstimate method.

Do it properly. Treat velocity as any other measurable operating outcome, not just a vibe. Let’s quickly talk about DORA’s research that provides standardised software delivery metrics, i.e. change lead time, deployment frequency, failed deployment recovery time etc. They explicitly state that these metrics predict better organisational performance.

So what does this have to do with FinOps?

FinOps reduces the hidden tax that slows delivery, e.g. time wasted on “who owns this spend?” or delayed approval to decommission idle resources because financial accountability is unclear, or “stop the line” budget escalations triggered by sudden, unexpected spikes or engineers optimising blindly as cost data isn’t accurate or timely.

The Calculator Method

1. Pick 1 to 2 velocity levers you can measure without getting into fist fights. For engineering, baseline average hours per month spent investigating costs, cleaning up tags that are non compliant, budget escalation and “war rooms” for spend spikes. Convert this to dollars based on your internal loaded cost rate (LCR). The finance team is your best friend here!. Be realistic and target a conservative improvement percentage. Don’t try and claim a 50% improvement overnight. FinOps at the end of the day is an enablement model, not a panacea to fix world hunger.
2. Measure cycle time to approve platform changes with cost impact, making commitment decisions, responding to anomalies or unexpected growth. Use your own finance model (LCR or some other criteria) to convert cycle time improvements to business value.

Dimension 3: Risk Mitigation (Cost Avoidance Costs Real Money!)

CFOs understand risk. It’s a core part of their job. What they don’t love is vague risk narratives not backed by numbers. We already have the signals that are telling us that the risk is manifesting financially. Based on Flexera’s reports, organisations on average are exceeding their budgets by ~17%. This is not a technical problem, but one of governance.

The Calculator Method

Quantify “cost overruns prevented” as a form of cost avoidance. You need to baseline the average cloud budget variance (both in absolute dollar terms and as a percentage). Determine the portion that is expected organic growth (increase in costs you cannot avoid) as opposed to unplanned or unmanaged variance (bad!). Model the impact FinOps will have on the organisation in terms of improved forecasting cadence and accountability, anomaly detection and response times along with guardrails that will stop runaway spend early.

So what do you to take to the CFO? This metric using the calculation below:

Cost over runs prevented = average monthly unplanned variance x FinOps prevention Rate (%)

So where does the FinOps prevention Rate come from? Well, don’t invent benchmarks. They don’t work and you will not be able to defend them under scrutiny. Instead, use metrics that the CFO already knows and trusts: the company’s internal rate of return (IRR).

Start small. You don’t have the data to be able to defend a large prevention rate. Instead, start with IRR + 2%. This way you are not claiming FinOps will change the world overnight. All you are saying is that you can guarantee the ROI will at least be marginally better than the next best use of the capital. That’s a bar you can most certainly clear with just basic anomaly detection and budget alerts.

However, what you do need to acknowledge and commit to is this: Quarterly measurement. Track forecast variance against actual, adjusted for approved growth. The difference is the measured variance prevention rate. In a few quarters you will have the data that either validates the assumption, or more likely support lifting it higher. What this does is put FinOps on the same footing as any other capital allocation decision. You’re not asking the CFO to accept a novel or unproven framework. You are using their framework and are much more likely to gain traction.

The other thing you need to be mindful of is the operational risk of vendor lock-in. If you don’t understand your actual costs, you will not be in a position to make data driven decisions, or negotiate effectively with providers (e.g. Private Pricing Agreements or Enterprise Discount Programmes). A lot of the organisations I worked with during my time at AWS had one thing in common: granular cost allocation that gave them leverage at the negotiating table.

This risk mitigation becomes even more apparent in AI heavy environments where costs can spike without a corresponding business activity or an increase to revenue. This is the exact pattern shift that risk mitigation attempts to tackle, as these spikes make traditional forecasting effectively useless.

Dimension 4: Innovation Enablement (The CFO Version, Not the TED Talk)

This dimension is the hardest to quantify and the easiest to, well, “lie” about.

There is no way to forecast the value of every experiment. Just model it like a portfolio. Be the Venture Capitalist.

FinOps isn’t about controlling or starving innovation. Done right, it creates the financial visibility and controls so organisations can make informed bets. It transforms cloud spend from a black box into a portfolio of investments with measurable returns. This is where unit economics becomes the link, tying spend to business outcomes using metrics that make sense for your business. You need to define those metrics: is it cost per transaction, cost per customer, cost per inference per revenue dollar, cost per citizen served?

The Calculator Method:

1. Define an explicitly approved budget for your innovation runway. Distribute it across three buckets - efficiency plays (near term), growth bets (medium term) and experimental options (long term). You must track two things - your kill rate or stop-loss discipline (yes, it’s meant to sound harsh!) and the unit economic trajectory for the survivors

An AI startup I was working with was burning through roughly $2 million annually on GPU infrastructure on model training and fine-tuning. The traditional approach using cost optimisation tools and philosophy would have mandated aggressive cuts or rightsizing. We instead, rebuilt their cloud economics and governance models to achieve an impressive 32% cost reduction, all the while maintaining experimentation velocity. The practical outcome for the founders was gaining an extra 8 months of runway without raising dilutive capital.

How the ROI Calculator Works

• Tab 1 (Inputs): Trailing spend, growth assumptions, cost of labour, current variance and waste estimates, maturity levels
• Tab 2 (Costs of the FinOps Practice): People (central team + embedded time), enablement (training, workshops, operating rhythm) and optionally, tooling costs.
• Tab 3 (Direct Savings): The money you stop wasting. The most defensible ROI component.
• Tab 4 (Velocity Gains): The time/effort saved when FinOps removes friction.
• Tab 5 (Risk Mitigation): The budget overruns you prevent by having better visibility.
• Tab 6 (Innovation Enablement): Investment decisions that arise from clear unit economics. This is hardest to quantify. Be conservative.
• Tab 7 (ROI Summary): Annual benefits (low, base, high), net benefit and the ROI percentage along with the payback period.

Take the summary, and build your business case around it. Be explicit in the assumptions you make and the returns you promise. You will be held accountable!

Build your own calculator or use the one below.

The ROI Lie: Savings Without Governance is a one-time “Sugar Hit”

The uncomfortable truth that all the FinOps tool vendors don’t want you to hear is that most organisations that claim the “30% savings” are measuring a point in time cleanup. It’s not a sustainable operating outcome.

I’ve seen this play out so many times, with almost all the customers I work with. They run a cost optimisation blitz, kill zombie resources, spend time and effort rightsizing their instances, enter into commitment based discounting programmes (by the way this is cost avoidance not true cost savings). The CFO gets a pretty slide deck for the board, everyone celebrates. Job done, right?

Guess what? Six months later, it’s all back. They rationalise it as “organic growth”, or re-investing the savings achieved. The waste that’s back is often worse, because the implicit assumption is that “we already did FinOps”. This is not just anecdotal data that I’ve personally seen. FinOps Foundation’s maturity data shows that more than 51% of organisations are stuck at the “Walk” maturity level. They’ve done the initial work, captured the easy wins, then just, stalled. They’ve stalled because savings without governance decay at roughly the same rate your cloud environment grows

Here’s something that should terrify you - If your cloud spend grows by 10% every year, which is conservative for growth companies or companies seeking to go “all in” on the cloud, and your optimisation efforts are project based rather than continuous, you’re running on a treadmill. That 10% or 20% you saved in Q1? That’s coming right back, eaten up by organic growth or configuration drift by Q3. By the end of the year, you are back to the baseline waste percentage, the difference is, the absolute dollar figure is now higher.

This is why I keep hammering the capital allocation framing. A CFO doesn’t run a “portfolio optimisation project” once a year then ignores their investments for eleven months. Any investment portfolio has continuous visibility, governance mechanisms, and decision rhythms. Your cloud spend deserves the same discipline.

The calculator model I’ve outlined isn’t just a single-use tool. It’s a model for funding operating capability. Now, it’s not exhaustive, no single tool can cater for all possible scenarios or unique business requirements, but it’s a starting point. If you only use it to green-light a 90 day project. Think about day 91, otherwise you’ll end up with a sophisticated justification for a sugar hit.

The honest ROI question isn’t “how much we’ll save”, it’s “how much will we keep saving and what governance makes that sustainable”.

Benchmark Without the Lies: Maturity patterns

One approach is to benchmark where ROI tends to show up first and is based on maturity. For example, based on the FinOps Foundation’s data, more than 51% of organisations remain stuck at “Walk”, with only slightly over 14% reaching “Run”

This is the pattern I’ve consistently seen, and what we strive to build with the customers we work with:

• At early maturity (Crawl), ROI is dominated by the easy pickings, i.e. visibility, waste elimination and basic cost allocation. The “win” is stopping the obvious leaks and setting clear resource ownership.
• At improving maturity (Walk), ROI shifts to forecasting discipline, anomaly detection and response and decision velocity. Fewer budget shocks, faster approvals less friction.
• At advanced maturity (Run), ROI becomes capital allocation quality, i.e. unit economics, portfolio style decisions and governance that keeps innovation funded without loosing control.

Did you spot what’s missing there? A magic “typical ROI%” number. That number only becomes meaningful when your inputs are real, i.e. “garbage in, garbage out”

This opportunity is particularly acute for Australian organisations, with public cloud spending surpassing $26 billion in 2025 with a growth trajectory of almost 19% compared to 2024. And Gartner’s assessment is blunt.

Most Australian companies remain immature in cloud cost management. The immaturity gap represents billions of dollars in misallocated capital across the market, and competitive advantage for those who close it first.

Your Next Move

If you want an answer to “What’s the ROI in FinOps” without bluffing, this is what you need to aim for

“FinOps pays for itself by combining measurable waste reduction with safer, faster decision making, while improving capital allocation by tying spend to unit economics and outcomes”

Yes, it’s long-winded. The point is:

Cloud spend isn’t a utility bill. It’s a portfolio. AI makes that portfolio volatile.

FinOps is the operating model that keeps you investing deliberately, instead of spending accidentally.

And yes, keep it tool-agnostic. If your ROI depends on a specific dashboard vendor, it’s not FinOps. It’s subscription

Ready to Build Your Business Case?

We can run the numbers for you, validate and stress test your assumptions and produce a CFO-ready one-page business case. Or if you prefer download the calculator template, tweak it and do it yourself. Either way, the discipline that matters, is the same.