The AI Investment Question Boards Should Be Asking

The conversation around AI investment is changing.

For the past few years, speed mattered. Organisations wanted to understand the technology, experiment with new tools and build enough experience to participate in what was clearly becoming a significant technological shift.

That phase has created something valuable: familiarity. AI is now present across software development, customer service, marketing, operations, research, analytics and knowledge work. Many organisations can point to successful pilots, growing adoption and measurable improvements in individual productivity.

For boards, the next stage requires a different level of scrutiny.

An employee completing a task 30% faster represents a productivity improvement. A team producing the same output in fewer hours has created additional capacity. The financial value appears when the organisation determines how that capacity will be used.

This distinction matters because AI can produce impressive productivity improvements long before those improvements become visible in business performance. As investment increases, boards need to understand what changes after the technology works: whether the organisation can serve more customers, accelerate product development, improve margins, strengthen decision-making, reduce risk or create new sources of revenue.

The AI investment conversation is therefore moving towards a more important measure of success: what becomes economically different in the business when the investment succeeds.

AI Activity, Productivity and Economic Value Are Different Things

AI activity is relatively easy to demonstrate. Organisations can measure licences, active users, pilots, automated tasks and the number of functions using AI.

These measures establish adoption.

Productivity goes further. A developer produces code faster. A customer service professional resolves an enquiry more quickly. A marketing team accelerates research and content development. A financial analyst completes part of an analysis in considerably less time.

Economic value begins when those improvements change the performance of the business.

Consider a team of 100 professionals who each save several hours a week through AI. Collectively, the organisation may create hundreds of hours of additional capacity. The significance of that capacity depends on what happens to it.

It could allow the company to handle greater customer demand without increasing headcount at the same rate. Product development could accelerate. Experienced employees could redirect time towards more complex work. Service quality could improve because people have greater capacity to focus on difficult cases.

In each example, productivity becomes economically meaningful because the organisation has decided where the additional capacity should go.

This creates an important distinction for boards: time saved is capacity created. Value depends on how that capacity is deployed.

AI Can Move the Constraint Before It Removes It

There is another consequence of AI productivity that receives less attention.

When one part of a business becomes significantly faster, the constraint often moves somewhere else.

Consider an engineering organisation where AI substantially accelerates coding. Developers can produce and modify software more quickly, but releases still depend on testing, security reviews, product decisions and deployment processes.

Engineering capacity has increased, while overall product delivery may improve much less because another stage has become the limiting factor.

The same pattern can appear throughout an organisation. AI can accelerate document preparation while approvals remain slow. Customer service teams can process routine enquiries faster while complex cases continue waiting for specialist attention. Commercial teams can generate more opportunities while qualification capacity remains unchanged. Analysis can become almost instantaneous while important decisions still move through lengthy approval structures.

The investment has improved part of the system. The economic opportunity lies in understanding the whole system.

This means the next constraint can become just as important as the activity AI has accelerated. Organisations that recognise it can direct investment towards the points that determine overall performance, rather than continuing to optimise individual tasks.

The Greater Opportunity Is in Redesigning How Work Happens

Many early AI implementations have followed a logical approach: take an existing task and make it faster.

That creates value, but it leaves the surrounding process largely unchanged.

A more significant opportunity appears when organisations reconsider how the work would be designed if AI capability had existed from the beginning.

Take a customer request that currently passes through several stages. One employee reviews the information, another team completes checks, a manager approves the outcome and someone communicates the decision.

AI could make each of those individual stages faster.

A redesigned process could operate very differently. Information could be interpreted immediately, routine cases handled automatically, unusual cases identified earlier and specialists involved precisely where their judgement adds value.

The organisation has moved from accelerating individual work to redesigning how work happens.

That shift has broader economic consequences. It can reduce handoffs, increase the volume an organisation can manage, accelerate decisions and concentrate human attention on the activities where expertise matters most.

It can also change organisational design itself.

When technology becomes capable of performing meaningful parts of a workflow, businesses gain an opportunity to reconsider how responsibilities are distributed between technology and people. The result can be a different operating model, rather than a faster version of the existing one.

Productivity Changes Where Human Expertise Creates Value

This is where AI investment becomes closely connected to workforce strategy.

As technology changes the amount and type of work required within a process, the relative value of different human capabilities changes with it.

Routine execution may consume less time. Judgement, exception handling, architecture, customer understanding, quality control and complex decision-making can become more important.

The outcome is more nuanced than a simple reduction in headcount.

Some organisations may be able to grow without expanding teams at the same rate. Other functions may require new expertise. Certain roles may broaden because professionals can manage significantly more activity, while others may become more specialised because the remaining human contribution involves increasingly complex decisions.

Software engineering illustrates this clearly.

If AI allows experienced engineers to spend less time producing routine code, the opportunity extends well beyond generating greater volumes of software. Their time can move towards architecture, system design, security, technical strategy, mentoring and complex product problems.

The organisation has effectively increased the leverage of its experienced people.

Capturing that value requires deliberate workforce decisions. Roles, responsibilities and expectations need to evolve alongside the technology.

At iTechScope, we see this becoming an increasingly important part of the technology talent conversation. Organisations are moving from broad demand for AI skills towards more specific combinations of expertise that reflect how work itself is changing.

Technical depth still matters. Increasingly, so does the ability to connect technology with architecture, products, commercial priorities, customer needs and complex organisational decisions.

The talent opportunity created by AI therefore extends much further than hiring AI specialists. It involves understanding where human expertise becomes more valuable because AI is present.

The Real AI Investment Includes the Organisation Around the Technology

Technology represents only one part of the investment required to create enterprise value from AI.

Data may need to become more accessible and reliable. Existing systems may require integration. Governance and security capabilities need to develop. Employees need to understand new responsibilities. Managers may require different ways to allocate work and measure performance. Some organisations will need specialist expertise that has never previously existed internally.

This helps explain the significant distance that can exist between a successful AI pilot and a successful organisational capability.

A prototype demonstrates that something can work.

An enterprise capability must work reliably within the organisation's technology, data, security, regulatory and commercial environment. It also needs to become part of the everyday work of the people whose behaviour ultimately determines whether the investment creates value.

The full business case for AI therefore includes the organisational capability required around the technology.

For boards, this creates a more realistic investment picture. The relevant cost extends across technology, integration, data, talent, process redesign and organisational change. Understanding those dependencies early makes it easier to assess both the size of the opportunity and the resources required to capture it.

The Value of the Problem Should Determine the Investment

As AI becomes easier to experiment with, organisations face another challenge: abundance.

Almost every function can identify potential applications. Teams can create prototypes quickly. Vendors can demonstrate compelling capabilities. The result can be dozens of initiatives competing for capital, technology resources and senior attention.

The strongest AI portfolio does not necessarily contain the greatest number of initiatives.

A relatively small productivity improvement in an infrequent administrative process may create useful value. An investment that changes the economics of a high-volume customer interaction, accelerates a critical engineering process or improves a commercially important decision can create something much larger.

The difference is the value of the underlying problem.

This provides boards with a useful investment discipline. AI initiatives can be prioritised according to the business outcome they have the potential to change, rather than the sophistication of the technology involved.

That approach may sometimes lead to less visible investments. Improving data quality, modernising infrastructure or introducing specialist expertise could create greater long-term value than another highly visible AI application because those investments enable multiple capabilities to develop.

AI therefore needs to compete for resources in the same way as other strategic investments: through the economic importance of the opportunity and the organisation's ability to capture it.

Strong AI Investments Create Capability as Well as Return

Some AI investments can create another form of value alongside their immediate financial outcome: organisational capability.

A company that successfully redesigns an important workflow around AI learns how to integrate technology, data, people and decision-making in a new way.

A team that deploys AI successfully within a regulated environment develops knowledge that can accelerate future initiatives. An organisation that improves its data foundations for one strategic application may create infrastructure that supports several others. Technical teams that learn where AI performs effectively, where human judgement adds value and how systems behave in production accumulate experience that becomes useful across the organisation.

These capabilities can reduce the cost and increase the speed of future development.

An AI investment can therefore produce two forms of return: the direct business outcome and the capability the organisation develops while creating it.

The second matters because it can compound.

Experience gained from one implementation improves the next. Better infrastructure supports additional applications. Stronger internal expertise improves technology decisions. Successful workflow redesign creates knowledge that can be transferred into other parts of the business.

Over time, the organisation becomes better at converting technological possibility into business performance.

From AI Productivity to AI Economics

The next stage of enterprise AI will be defined by the ability to translate what the technology enables into measurable business performance.

Adoption tells an organisation that people are using AI.

Productivity shows that work is becoming faster or easier.

Economic impact appears when something meaningful changes in the business: capacity, cost, revenue, quality, speed, risk, decision-making or the ability to execute strategy.

For boards, that is where the AI investment conversation increasingly belongs.

It also creates an important implication for workforce strategy. When AI changes the economics of work, organisations need to reconsider where expertise creates the greatest value, which capabilities should develop internally and where new experience could accelerate progress.

For iTechScope, this is becoming an important dimension of understanding the technology talent market. AI will influence much more than demand for AI specialists. It will affect the shape of technical teams, increase the leverage of experienced professionals and create demand for new combinations of technical, commercial and organisational expertise.

The strongest AI investments connect three decisions that are often considered separately: where technology creates capacity, how the business converts that capacity into value and which human expertise is required to make that conversion possible.

Boards therefore need something more useful than a measure of how much AI the organisation has adopted.

They need clarity about economics.

What becomes meaningfully different in the business when the investment succeeds?

A strong answer should eventually appear somewhere the organisation can recognise: in greater capacity, stronger margins, faster growth, better decisions, lower risk or a capability that expands what the business can do next

By Konstantina Thoma, Digital Office Associate, iTechScope, 28/09/2026