FinOps Isn’t a Cost-Cutting Project – It’s Capital Allocation

Around 79% of today’s IT budgets sit under operating expenditure, yet barely 15% of organisations can draw a clear line between their cloud costs and measurable business value (McKinsey, citing Gartner’s 2025 CIO Agenda). This gap drains millions in misallocated resources and lost opportunities, the inevitable result of unmanaged spending and murky performance visibility.

The hidden costs cut much deeper; they create a competitive handicap. While you scour your cloud spend for yet another round of savings, competitors are using the cloud as a strategic growth engine. Using it as a strategic tool to enter new markets in weeks, launching products at unprecedented speed, and aligning structure, process, and technology into a single performance machine

Many organisations still treat cloud costs as a utility bill, an operational expense to minimise through aggressive procurement negotiations, discounts, or usage restrictions. After enabling ~35% spend reductions across the many organisations I’ve worked with, I’ve realised something counterintuitive: the organisations that achieve the best outcomes don’t focus on cost-cutting but view their cloud spend as an investment portfolio requiring sophisticated capital allocation.

The data speaks for itself. Accenture research confirms organisations that apply the same investment management principles to cloud spending achieve between 1.2x to 2.7x better returns on their time and effort when compared to those focused purely on cost reduction. Companies that treat FinOps as a “90 day cost cutting project” achieve most of their intended savings, but then stagnate, or in a lot of cases go back to square one with organic growth. Those that embrace portfolio management principles sustain much larger (30% – 40% ) levels of optimisation while managing organic growth and accelerating innovation.

Cloud spending is not an expense problem; it’s a capital allocation opportunity.

The Investment Thesis

In 2025 the FinOps Foundation rewrote a core principle to ‘*Business value drives technology decisions*,’ expanding beyond cloud-only language. Pair that with Gartner’s projected $723.4B public-cloud spend in 2025, and it’s obvious: cloud is a capital-allocation problem, not a utility bill.

Why now? Simple. Executives finally have the data to see what each dollar spent produces. Yet most do not realise the value of this data. The first and most common question I get every time I walk into a room is “*How can I reduce my cloud costs*“. What they should be asking is “*Which cloud investment will help me accelerate my business*“

Let me be blunt: Organisation that are stuck in cost cutting mode are falling behind the curve. While they celebrate their “25%” cost reductions, their competition using investment frameworks are achieving better outcomes AND accelerating growth at the same time.

So what’s the real opportunity here? McKinsey estimates $1.6 – $3.4 trillion in cloud-enabled value by 2040. The AI boom will only help fuel this growth further. Firms that fund products with positive unit economics capture a disproportionate share of that $1.6–$3.4T; those that apply blanket cuts often raise cost-to-serve while depressing growth.

The Unit Economics Ladder Framework

So how do organisations adapt to this fundamental change transforming cloud from an expense to be controlled and minimised to an investment? You climb the Unit Economics Ladder; an approach that connects granular technical metrics to P&L outcomes.

This is how it works:

• Level 1 – Unit Costs: What’s your cost per API call, database query or compute hour? This is the land of the engineers, optimising and tweaking each individual resource and service. Example: reducing Lambda costs from $0.0003 to $0.0002 per invocation
• Level 2 – Service Costs: How much does it cost to serve a customer, or process an order, or complete a transaction end to end. This is where we start linking cloud and infrastructure costs to business operations. Example: your checkout service costs $0.45 per transaction across all infrastructure
• Level 3 – Product Line Economics: What’s your cost for net new customer acquisition? Cloud to margin contribution? This connects cloud spend directly to profitability per product. Example: Premium tier customers cost $12/month to serve but generate $89/month revenue a healthy 86% margin
• Level 4 – P&L Impact: What percentage of revenue goes to servicing your cloud costs? What the cloud efficiency ratio? This is where executives make investment decisions. Example: cloud costs at 8% of revenue for SaaS companies signals efficient operations

Here’s the reality check: From what I’ve seen most organisations never get past Level 2. They keep optimising their EC2 instances, reducing API calls while remaining oblivious to unit economics. I worked with an organisation that discovered one of the applications they were very proud of, the one they had optimised as much as it was possible, consumed 37% of their cloud resources while only generating 12% of revenue. Without the ladder framework, they would have kept pouring investment into a margin destroyer.

The one anti-pattern I see consistently is teams celebrating infrastructure savings while their cost per customer actually increases.

Stop managing costs. Start managing unit economics

The Investment Charter Approach

Understanding Unit Economics and developing a framework to measure and track them is just the first step. Implementing a process that transforms these insights into decisions is what differentiates a mature organisation from the rest

The aim is to empower employees at all levels and not create a bureaucratic nightmare. A good Investment Charter establishes at least three decision tiers

• Operational Decisions: for smaller investment technical decisions (e.g. <$10k/month, buying a Reserved Instance commitment) that engineers can make autonomously, but within guardrails. No committees and no delays
• Tactical Decisions: requiring approval from the FinOps team (e.g. up to $100k/month), within a set SLA. Clear criteria for investment and fast execution.
• Strategic Decisions: Investment committee review, like any major capital allocation decision requiring substantial investment (e.g. more than $100k/month),

Again, these are merely suggestions, each organisation will need to determine what works best for them. But this is what separates the winners: tracking the KPIs and metrics that matter.

• Unit Cost Trends: Are your per-transaction costs improving? (e.g. Target: 10% quarterly improvement)
• Service Efficiency: Cost per customer served trending down? (Benchmark against industry standards)
• Product Margins: Cloud costs as percentage of product revenue (e.g. Set targets that varies by product maturity)
• Portfolio Balance: Resources allocated proportional to revenue contribution? (e.g. spend share vs. revenue share within ±20% band, reviewed quarterly)
• ROI Realisation: Actual vs projected returns on cloud investments (e.g. Target: 80% achievement rate)

These empower teams to self govern, and when teams see and understand these metrics, they make better architectural decisions. These sound fantastic, but here’s a reality check: This isn’t a simple 90 day transformation. It takes time, typically 12 – 18 months to embed capital allocation thinking into the organisation’s culture and ways of working.

Stop asking for budget approval. Start making investment decisions

Five-Pillar ROI Measurement

Where I have seen most FinOps programs fail is when they measure cost reduction and call it a success. The AWS Cloud Value Framework gets this right; to measure Cloud ROI requires you to look at it using multiple angles not just cost.

1. Cost Efficiency: Yes absolutely, reduce waste, but as a percentage of revenue, not an absolute dollar figure
2. Productivity: Do your teams ship features faster? I have seen organisations improve delivery efficiency after establishing a Cloud Centre of Excellence. That’s velocity, not just savings
3. Resilience: incident count, MTTR, and control coverage. IBM’s 2024 global average breach cost was $4.88M; 2025 fell to ~$4.44M, but the US average rose; that’s not
4. Agility: Has your Time to Market for new features or products improved? What about experimentation velocity or Geographic expansion speed? Cloud is your competitive enabler not merely IT Infrastructure
5. Sustainability: Carbon emissions per transaction or per dollar of revenue. This will soon become mandatory in multiple jurisdictions. You need to get ahead of the curve

Smart organisations set minimum thresholds for all five pillars, rejecting investments if metrics drop below them.

Measure value, not Cost. Optimise portfolios, not line items

Making It Real: Governance That Works

It’s all well and good to theorise. Execution is where the rubber hits the road. Having a governance structure is what delivers results. As an example,

• FinOps Investment Committee: CFO and CTO to co-chair, meeting monthly for tactical reviews, quarterly for strategic rebalancing (FinOps Foundation, 2025). Just like an investment portfolio
• Distributed Ownership: Enable your team. Have a FinOps champion embedded within every product team, not a cost optimisation specialist, but someone that can help optimise their portfolio and make data driven business decisions. I’ve seen the shift from resistance to enthusiasm happen when teams realise, they are gaining control, not losing it
• Clear Escalation Paths: Keep the operational noise within the engineering teams, make the tactical decisions within the 5-day SLA, escalate quickly so strategic decisions hit the investment committee.

The thing that destroys most attempts to set this up? Setting FinOps as an IT project. Without executive sponsorship, all you end up with is pretty dashboards that nobody looks at. Dashboards don’t move money. The charter defines thresholds and owners; the committee rebalances funds quarterly; product teams execute within guardrails.

The cultural transformation is the most challenging part. Your technical teams need to be able to think commercially, while the Finance teams need to understand cloud dynamics. Executives need to treat the cloud like any other investment, with the same amount of rigour, governance and an eye on the return on investment. Get this right, and everything else will follow.

Your Next Move

You only have a small window of opportunity. While you are debating cost reduction targets, your competitors are building investment frameworks that will give them a competitive edge.

Gartner forecasts $644 billion in GenAI funding alone. AI workloads don’t follow traditional cloud economics, and require specialised frameworks. Adopt AI-specific unit metrics (e.g., $/1k tokens at P95 latency; $/training run to target loss) and put them under the same investment charter, ensuring capital will flow to teams that can prove ROI per token, not per virtual CPU. Without a mature capital allocation framework and processes that go along with it, you will end up burning cash on AI experiments with no ROI framework to guide decisions.

The $3.4 trillion cloud opportunity won’t wait; the AI revolution will only amplify the gap between leaders and laggards. Where will you be?