When technology no longer depreciates: CapEx vs OpEx in cloud computing

Companies large and small decommissioned information technology infrastructure when they moved to cloud computing.

That’s hardly news. Cloud and as-a-service offerings have been mainstream for years. Enterprises begin shifting functions such as email, CRM and file storage to cloud from around 2011. by 2015 cloud was in widespread use across most industries and sectors.

Cloud is the new normal

Moving applications and tech infrastructure outside the business is often cheaper and more efficient than owning and maintaining infrastructure.

For years this was especially true where technology isn’t strategic. Today, cloud is the first option most businesses consider when investing in new apps or services.

Yet it still isn’t always clear which approach has the lowest total cost of ownership because information technologies TCOs (total cost of ownership) are hard to measure. Few businesses can tell you how much specific self-hosted applications cost to run.

There have been some notable cases of businesses moving back from cloud to on premise computing.

In theory, cloud costs are easier to gauge – you get a bill each month. There is nothing to depreciate. Yet some organisations have found there are hidden costs and, in cases, escalating costs associated with cloud. "Sticker shock" is not unknown.

The financial shift from CapEx to OpEx

The most profound change driven by cloud adoption is not technological, but financial. Historically, purchasing technology infrastructure—servers, storage arrays, network gear—was Capital Expenditure (CapEx). The assets appeared on a company's balance sheet and were systematically deducted against income over several years via depreciation. This created predictable financial structures and tax implications.

When an organization shifts to the cloud, technology moves into the Operational Expenditure (OpEx) column. Cloud resources (like AWS, Azure, or SaaS subscriptions) are treated as a monthly utility bill.

Where OpEx pays off

Beyond the balance sheet, the OpEx model provides a critical strategic advantage: agility. When you treat technology as a monthly expense rather than a depreciable asset, two significant strategic hurdles are removed:

1. Elimination of sunk cost thinking: With CapEx, large upfront investments encourage businesses to stick with aging technology just to fully depreciate the asset, slowing down innovation. OpEx allows teams to instantly pivot, decommission or downgrade services as business needs change, leading to far more responsive operations.
2. Faster Time-to-Market (TTM): New projects no longer require months of procurement cycles, budget approvals for hardware and physical installation. A development team can spin up a complete, global infrastructure in minutes, collapsing the time required to launch new products and services.

It's not that simple: Cloud costs can escalate

While a monthly cloud bill seems straightforward, measuring the true TCO of a cloud project is often more complex than calculating the cost of an old-style data centre.

Sticker shock typically comes from three areas:

1. The hidden cost of unused resources: (People sometimes refer to "zombie virtual machines"). Cloud resources are often provisioned but left running when not needed (known as "zombie instances"). These run up massive, non-value-add bills because billing is continuous. A traditional server might just sit idle after being bought but the cloud taxi meter keeps running. This is where modern disciplines like FinOps (Cloud Financial Management) become necessary—you must actively govern and optimise usage. This requires skilled employees.
2. Egress fees and data transfer: While storage is cheap, moving large volumes of data out of a cloud provider's network (egress) can incur substantial, unexpected fees.
3. Labour and skills: Organisations often underestimate the cost of specialised cloud expertise required to manage, secure, and optimise complex cloud systems. These are skills that traditional IT teams typically do not possess.

The key lesson here is that moving to the cloud replaces a CapEx problem (large initial spend, slow cycles) with an OpEx governance problem (continuous cost optimisation). You can't take your eye off the ball.

No depreciation

Few understand the financial and strategic implications of not having large IT investments on balance sheets. Tax laws mean companies depreciate installed systems over several years. Traditional technology is capital expenditure.

Cloud subscriptions are service payments. You can write them off as operating costs.

A switch to the cloud frees capital for investment elsewhere. It also changes how people, especially at the top of businesses, think about technology. The implications of this aren’t yet fully clear, but most likely companies will become more nimble in their thinking.

There will be less sunk cost thinking – that’s where people argue “we’ve paid for this stuff we need to get a return on what we spent” – and more “what technology makes the best sense now” thinking.

This will give business owners more agility – they can better respond to changing conditions. This works just as well when expanding a business as when cutting costs.

Which could mean turbulence. That’s not always a bad thing, change means opportunity, but it can also mean greater business uncertainty and more risk-taking.

FinOps is the way to stay ontop of OpEx

If the shift to cloud eliminates capital depreciation concerns, it introduces a new challenge: continuous cost governance. This is why many organisations experience "sticker shock"—the lack of upfront friction to spin up resources means costs can spiral quickly if left unchecked.

The business world has responded by creating a new operational culture: FinOps (Cloud Financial Management).

FinOps is not just a tool or a team; it’s a set of practices that bring financial accountability to the variable spend model of the cloud. It aims to achieve three primary goals:

1. Inform: Providing clarity and visibility into who is spending money, on what resources, and why.
2. Optimize: Right-sizing cloud resources (e.g., downsizing over-provisioned virtual machines or deleting "zombie instances" that are running but unused) and leveraging discount models (like reserved instances or savings plans).
3. Operate: Integrating financial management into the entire development cycle, ensuring engineers are financially accountable for the architecture they build.

By mastering FinOps, businesses fully realise the promise of OpEx: they gain maximum agility while maintaining control over the monthly burn rate.

Technology as an elastic service

Technology is no longer a physical asset to be owned and depreciated, but an elastic service to be consumed and optimised. This fundamental CapEx-to-OpEx shift has moved the strategic challenge from:

• Then: How do we justify the massive capital investment?
• Now: How do we continuously optimize and govern our monthly consumption for maximum agility and minimal waste?

The new reality demands that technology leaders are also financial managers, ensuring that every dollar spent in the cloud delivers maximum, measurable business value. It's not like your granddad's enterprise IT, but done well, it can be far more efficient.