What is Technology Capital Performance?

The definition

Technology Capital Performance is the measurement of whether a business’s technology spend is producing a proportionate, attributable return, treating that spend as deployed capital to be judged on what it produced rather than as an operating cost to be reduced.

OPTIMAZE

Last reviewed August 2026

7 minute read

On this page

Why it exists

Capital, not cost

What it requires

How it differs from FinOps

Why it is hard to measure

How OPTIMAZE measures it

Common questions

Book a demo

Why it exists

It exists because technology has become one of the largest lines on the P&L and the least likely to be measured the way the rest of the business is measured. Cloud, AI and platform spend get reported accurately and rarely get assessed for return. Technology Capital Performance is the discipline of closing that gap, and the figure that results from it.

Capital, not cost

The distinction is not semantic. A cost is something you minimise, and every instrument built for it is an efficiency instrument, which is why a practice organised around cost produces savings targets, utilisation rates and commitment coverage. Capital is something you deploy, and deployed capital is judged on what came back.

Two framings of the same dollar

Treated as cost

Something to minimise

drive it down

Savings targets, utilisation rates, commitment coverage. Every instrument built for it is an efficiency instrument.

Treated as capital

Something to deploy, and judge on what came back

activity

3.3×

return

A proportionate, attributable return. The same question the business asks of every other investment it makes.

Every instrument on the left is built to make the number smaller. Nothing on the left can tell you what the number bought.

Those two framings ask different questions of the same dollar, and a business gets the answer to whichever one it chose. Most estates chose cost without deciding to, because that is the shape the bill arrives in.

No board asks whether a manufacturing line was optimised without also asking what came off it. No finance function accepts a maturity assessment as an answer to that second question. Technology is the one line on the P&L where that substitution has been allowed to stand.

What Technology Capital Performance requires

Three things, in this order. The order is not optional, because each one depends on the one before it.

The dependency, bottom up

Budgets and goals

Is the ratio tracking to plan?

needs a ratio to track

Unit economics

Attributed spend over business activity

needs attributed spend as its numerator

Attribution

What caused the cost

What comes out

$11.60

cost to serve, per customer

A figure with a trend line

Remove the bottom layer and nothing above it is defensible. A ratio built on spend that cannot be traced is a number without a denominator anyone can check.

Attribution establishes what caused a cost. Spend is traced to the product, feature, service or team that incurred it, built from the workload upwards rather than the invoice downwards. A cost that cannot be traced cannot be assessed for return, which is why attribution comes first and why nothing downstream is defensible without it.

Unit economics expresses attributed spend against the business activity it supports: cost per customer, cost per transaction, cost per feature. This is where a cost becomes a ratio, and where technology spend starts behaving like any other unit cost a finance team already reports on.

Budgets and goals establish whether that ratio is tracking to plan. A target set against a unit economic turns the figure from an observation into something that can be governed, with movement visible in time to act on rather than after the quarter has closed.

What comes out of those three is not a report and not a maturity score. It is a number, with a trend line, that a CFO can put in a board paper and defend under questioning.

How it differs from FinOps

FinOps is an operational framework and cultural practice, and a good one. It measures how well a business manages its technology spend: capabilities adopted, personas engaged, maturity reached, cadences held. Those are measures of practice.

Two different scoreboards

FinOps measures practice

How well the business manages the spend

Capabilities adopted

Personas engaged

Cadences held

Working groups

Certifications

Crawl

Walk

Run

Many measures, all of them describing the practice.

Technology Capital Performance measures outcome

What the investment returned

3.3×

return on attributed technology spend

One measure, and it is the one the board asked for.

Both scoreboards can read well at once. Only the one on the right answers the question a board asked.

Technology Capital Performance measures outcome. It is not a FinOps capability, not a phase, not a maturity band. Different question, different owner, different artefact. A business can run a mature practice and still answer “what did our technology investment return last year” with an estimate, because process maturity and return are not the same claim and only one of them survives a board conversation.

The two are complementary. FinOps optimises the spend a business has already committed. Technology Capital Performance asks whether it should have been committed, and what came back when it was.

Why it is hard to measure

Not incompetence, structure. A technology bill arrives as a single figure covering dozens of teams, products and workloads, so working out what it produced means first working out whose it was. There is no obvious unit: a factory has one widget and one cost to produce it, a cloud bill has no equivalent.

Where the unit went missing

A factory

One widget, one cost to produce it

$4.20

per widget

The denominator is obvious and nobody has to invent it.

A technology bill

One figure, dozens of owners, no obvious unit

$2.1M

per what

Working out what it produced means first working out whose it was.

The factory did not have to invent its denominator. A technology bill has to be given one before any ratio exists.

The conventional answer has been to require a tagged estate before anything can be measured, which puts a data programme in front of the first defensible number. AI spend makes that worse, because it scales with usage rather than provisioning, and an inference call is an event rather than a resource that can carry a label.

How OPTIMAZE measures it

OPTIMAZE is built to produce this figure at the measurement layer rather than the process layer.

Spend is attributed to the products, features and teams that caused it, derived from how resources, accounts, services and metadata actually relate to one another rather than from a tagging programme that has to finish first. That means a defensible view exists on an untagged estate, on a partially tagged estate, and on an estate carrying incompatible conventions inherited from acquisitions.

The tag is an input, not the ceiling

Resources carrying a usable tag

30%

Spend attributed by OPTIMAZE

95%

Illustrative. On the same estate, attribution derived from relationships is not bounded by tag coverage, which is the gap that removes the data programme from the front of the process.

Unit economics sit on that attribution, tracked continuously rather than rebuilt each quarter. Targets are set against those ratios and update as the underlying cost and usage data changes. Attribution is maintained as rules over relationships rather than a static chart of accounts, so a reorganisation, a rename or an acquisition is absorbed rather than requiring the estate to be retagged and the previous year restated.

A business that can only report how well it manages technology cost is reporting on its own process. A business that can report what its technology investment returned is reporting on its own performance. That second figure, quantified and defensible, is Technology Capital Performance.