Demystifying CapEx, OpEx and Capitalisation for Product Leaders

If you’ve led product or technology teams long enough, you’ll have heard:

• “We need more CapEx”
• “Can we capitalise this?”
• “That team is CapEx-funded”

Understanding CapEx, OpEx and capitalisation is foundational for any leader shaping a product or technology roadmap, because these concepts directly influence what gets built, how it gets funded, and how success is measured by the business and its shareholders.

Most organisations blur the line between how money is spent and how assets are accounted for. These are related, but they are not the same. When the distinction is misunderstood, it creates confusion in planning and investment decisions, distorts delivery behaviour, and almost always drives unhelpful tension between executives, finance teams and the people doing the work.

For leaders operating or transforming into a modern Product Operating Model, having a clear understanding of OpEx, CapEx and the act of capitalisation becomes even more important. Funding models quietly shape team behaviour. If you do not understand the financial mechanics, you do not fully control the roadmap.

As Marty Cagan consistently argues, empowered teams and outcome-led delivery depend on the freedom to invest in learning and iteration. His work surfaces a core tension inside many organisations: the same words are used both to govern authorisation to spend and to describe how spend is later treated in the accounts. When those meanings blur, teams start optimising delivery around financial mechanics instead of customer value.

This article exists to untangle that confusion. It separates the governance meaning of CapEx and OpEx from their accounting meaning, and then shows how capitalisation fits into the picture. Our goal here is not to teach accounting, but to give product and technology leaders a mental model to help unlock better investment decisions and delivery behaviour.

Full disclosure: this article is written to help product and technology leaders work more effectively with finance and the wider business. It is not intended to be accounting advice.

CapEx & OpEx are overloaded definitions

Depending on the context, CapEx and OpEx can mean very different things, but they broadly fall into two distinct categories, and most confusion comes from not being clear which one you are actually talking about.

CapEx & OpEx as investment governance and budget authority

In this interpretation, CapEx and OpEx are not accounting concepts at all. They are controls on financial risk, decision-making authority, and how much irreversible commitment the organisation is prepared to take on.

Here the distinction is not about how money is recorded, but about who is allowed to commit it, how much they can spend, and how hard that decision will be to undo.

• OpEx governs ongoing, reversible operating spend.
• CapEx governs long-lived, capacity-creating commitments.

When leaders say “we don’t have CapEx approval” or “this isn’t in the capital plan”, they are not making an accounting statement. They are talking about permission to commit the business to something durable and expensive.

CapEx exists here to control exposure, not to classify cost.

CapEx & OpEx as accounting classification and financial reporting

In this second interpretation, the question changes: it is no longer “are we allowed to spend this?”, rather it becomes “how should this appear in the accounts?”.

• OpEx is cost recognised immediately through the P&L
• CapEx is cost recognised as an asset and expensed gradually through depreciation or amortisation

CapEx, OpEx and capitalisation are not the same concept

CapEx, OpEx and capitalisation are three distinct ideas, but confusion arises because people conflate the act of capitalisation with the class of spend.

In practice, teams mix up:

How work is funded

With

How outcomes are accounted for

The result is product strategy becoming distorted by accounting mechanics instead of led by value.

What CapEx is

In the governing sense, CapEx (capital expenditure) describes a purchasing commitment where the outcome is a long-lived asset that is difficult to walk away from. It is fundamentally about irreversibility and risk. Building a factory is CapEx because once construction begins, the business is committed. You cannot easily unwind the decision without significant loss.

In this framing, CapEx is not an accounting concept, it is an investment control mechanism. It exists to govern:

• How much capital the organisation is willing to lock up
• How irreversible a decision is
• Who is authorised to place long-term bets on the business

In the accounting sense, CapEx is money spent to acquire or create something you now own: durable, identifiable assets that are expected to deliver value over a long period of time.

• Buying servers is CapEx spend
• Purchasing hardware is CapEx spend
• Owning software outright is CapEx spend (licensed software is not, that is OpEx)

In accounting, CapEx exists to put assets on the balance sheet and recognise their cost gradually through depreciation or amortisation, depending on the asset type. The asset remains visible on the balance sheet, and its cost is spread over its useful life or (partially or fully) recovered through disposal or sale.

What OpEx is

In the governing sense, OpEx (operational expenditure) describes spending that keeps the organisation running day-to-day. It is mostly reversible and repeatable. If CapEx is about long-term commitment, OpEx is about short term borrowing. You can scale it up (borrow more), reduce it (borrow less), or stop it entirely with relatively little consequence.

In this framing, OpEx is not an accounting concept, it is an operating control concept. It exists to govern:

• How much operational cost the business can sustainably carry
• How spend is allocated across teams and functions
• How flexible the organisation remains in changing direction

In the accounting sense, OpEx is money spent to operate the business where no durable asset is created. The value is consumed as it is purchased.

• Salaries are OpEx
• Suppliers and contractors are OpEx
• Software subscriptions are OpEx
• Cloud infrastructure is OpEx

In accounting, OpEx is recognised immediately through the profit and loss statement. There is no asset created, nothing is held on the balance sheet, and no value is spread over time.

What the act of capitalisation is

Capitalisation is the most commonly misunderstood concept in this entire discussion. It is often spoken about as if it were a form of funding, an approval mechanism, or even a budget category. It is none of those.

Capitalisation does not give permission to spend, it does not release money, and it does not control what teams are allowed to do.

Capitalisation is an accounting decision applied after money has already been spent, it is not concerned with authority or investment risk. It sits entirely on the reporting side of the house, where governance is about what the organisation is allowed to commit to, capitalisation is about how the result of that commitment is reflected in financial statements.

Capitalisation determines whether the outcome of spend is recognised immediately as a cost in the P&L, or recorded as an asset on the balance sheet and then expensed gradually over time. Nothing about capitalisation changes cash flow. Nothing about it alters delivery decisions. It simply changes the financial representation of what has already happened.

You never capitalise people. You never capitalise vendors. You may capitalise what their work produces, but only if the result qualifies as an asset under accounting rules. Capitalisation is not a steering mechanism for delivery. It is a reflection mechanism for value.

A useful rule of thumb

A finance mentor once gave me a rule of thumb that still cuts through most of the confusion, provided you’re clear about which lens you’re using at the time.

From an accounting perspective, the question is straightforward:

Does this spend result in a direct, identifiable asset with future economic benefit that will be expensed over time?

If the answer is yes, it leans toward capitalisation.

If not, it is OpEx.

This test works well when what you are trying to decide is how something should be treated in the accounts, because it forces you to look for ownership, durability and economic life rather than activity or effort.

In the governance sense, the same question translates slightly differently:

Is this a reversible operating cost, or is it a long-lived commitment that locks the organisation into a strategic decision?

If the spend increases flexibility, it behaves like OpEx and is simply part of operating the business. If it creates lock-in, it behaves like CapEx, which is usually a signal that the decision deserves more scrutiny and senior approval.

Why software makes this unusually messy

Accounting assumes a clean separation: first you build an asset, then you operate it. Software does not work like that. Products are never finished, platforms never stabilise, and performance, resilience and scalability never stop evolving. The idea that something is “built” and then merely “run” does not map cleanly onto how digital products actually grow.

From an accounting perspective, this creates a practical problem. Finance still needs a dividing line between work that becomes an asset and work that is treated as ongoing cost. So most organisations default to a blunt but workable convention: initial builds are capitalisable, everything after is operational. It is not conceptually pure, but when applied consistently, it gives finance a line they can defend and auditors a model they can test.

From a governance perspective, the messiness shows up differently. Software investment rarely arrives in a clean, one-time decision. It accrues gradually through hiring, tooling, migration, architecture choices and platform dependency. Teams quietly accumulate long-term commitments without ever triggering a single “capital decision”, and by the time leadership realises they are locked in, the investment has already been made.

In large organisations, capitalisation also plays a second role: it preserves the integrity of reported profitability by matching investment costs to the periods in which the associated benefits are realised. Without this separation, significant long-term investments would distort the P&L, making a healthy business appear unprofitable.

This matching effect has a secondary consequence: it smooths reported earnings and stabilises investor expectations. Costs never disappear; they are simply recognised over time rather than all at once. That does not make the practice wrong, but it does mean that product and technology leaders should understand the strategic and political weight behind these conversations.

A simple framing to keep you grounded

In the governance sense:

• CapEx is about committing the business to long-lived decisions
• OpEx is about funding ongoing activity

In the accounting sense, the framing is simpler:

• CapEx is buying assets.
• OpEx is renting effort and access.
• Capitalisation is accounting applied afterward.

Bringing it back to the Product Operating Model

A modern Product Operating Model only works when financial mechanics are understood.

Funding models shape how teams are formed, what gets prioritised, and what behaviour is implicitly rewarded. Whether leaders realise it or not, the way money flows through the organisation quietly dictates delivery culture.

When CapEx, OpEx and capitalisation are misunderstood, governance and accounting become entangled in destructive ways. Roadmaps drift toward large, “parallelisable” initiatives because they look like investments. Discovery starts to resemble waste because it doesn’t map neatly to assets. Teams are nudged toward work that appears capitalisable rather than work that actually reduces risk or delivers value. What should be an investment conversation becomes an exercise in financial optics.

When leaders are financially fluent, that distortion disappears. Investment governance becomes explicit, and accounting is treated as what it is: a reporting consequence, not a design constraint. Teams are funded to pursue outcomes, not to manufacture assets for the balance sheet. Discovery and iteration can be approved with confidence once capitalisation is understood as an accounting outcome rather than a delivery rule.

This separation restores intent. Governance is about deciding where to place bets. Accounting is about recording what happened. When each is allowed to do its job, artificial pressure to bundle work into large initiatives fades, incremental investment becomes easier to justify, and finance can account for value after it is created rather than steering how it is created.

The result is faster decisions, cleaner prioritisation, and delivery that optimises for impact, not optics.