Before You Invest in AI, What Return Are You Actually Looking For?

I suspect one of the strongest forces driving AI investment today isn't strategy.

It's fear.

Not an irrational fear. AI is developing quickly, and its implications for business are significant.

But speak to enough business leaders and you hear a familiar concern.

What if our competitors move first?

What if our people are falling behind?

What if our business model becomes less relevant?

What if we're the last to understand what's happening?

Nobody wants to be the CEO who failed to recognise the next major transformation.

So the pressure builds to do something.

Introduce new tools. Train the team. Automate workflows. Build agents. Announce an AI strategy.

And increasingly, measure how much time and money the technology might save.

Those activities may be worthwhile.

But I wonder whether we're allowing the pressure to adopt AI to get ahead of a more important conversation.

What value are we actually trying to create?

The problem isn't AI. It's where we begin.

I'm currently studying AI Management at the University of Technology Sydney.

After more than twenty years building and leading a strategy-based business, I wanted to understand AI beyond the tools and headlines.

During a recent lecture, the discussion turned to how many AI initiatives begin without clearly defined scope or a convincing understanding of return on investment.

A figure of 95% was mentioned.

I haven't independently verified that figure or its precise definition, so I wouldn't present it as an established industry statistic.

But the underlying concern resonated with me.

Because I've encountered versions of this problem throughout my business career.

A business identifies a new technology or capability. The possibilities are exciting. Investment begins.

But the organisation hasn't properly established the problem it's trying to solve, the opportunity it's pursuing or what success would actually look like.

How can we know whether an investment is successful if we haven't defined what success means?

And with AI, I think the stakes are particularly important.

Because the technology may make the existing business more efficient while simultaneously changing the value of what that business sells.

Those are two very different commercial conversations.

We've been here before

I remember the transition to digital marketing.

For years, advertising had largely been a one-way message. Businesses developed campaigns, purchased media and spoke to their audiences.

Digital changed that.

Customers could respond, question, compare, share experiences and engage directly with businesses.

What had largely been a one-way message became a two-way conversation.

And with that came extraordinary amounts of data.

We could measure impressions, clicks, engagement, conversions and customer behaviour. Reports became more detailed. Dashboards became more sophisticated.

Yet meaningful insight remained surprisingly scarce.

We had more information about customers, but were we actually understanding them better?

And when we did learn something important, was that knowledge changing the decisions we made?

Many businesses initially approached digital as another collection of channels and technologies.

But the more fundamental transformation was happening in the relationship between businesses and their customers.

The opportunity wasn't simply to advertise more efficiently.

It was to understand customers differently.

I see a parallel with AI.

We're concentrating heavily on what the technology can produce, how quickly it can work and how much it can save.

But are we spending enough time understanding how it's changing customer expectations, competition and the economics of the businesses we operate?

Or are we simply trying to demonstrate that we're keeping up?

The disruption has already begun

I sometimes think we're approaching AI as though we're waiting for the decisive moment.

The next model. The next capability. The next breakthrough.

But I wonder whether we're missing something.

The fundamental change has already begun.

Think about the introduction of the first iPhone in 2007.

It wasn't simply a better mobile phone. It helped change what people expected a phone to do, how they accessed information and, ultimately, how businesses could reach and serve customers.

The technology continued to evolve. New applications, services and business models emerged. Some of those later developments were transformative in their own right.

But the original shift had already changed the direction of the market.

I think AI presents a similar moment.

It will become smarter, faster and more capable. There will be new tools, models and developments we haven't yet imagined.

But we don't need to wait for the next generation to recognise what's already changing.

Customers can access capabilities that once required specialist knowledge, considerable time or significant expense.

Businesses can analyse information, produce work and connect knowledge in ways that were previously impractical.

That changes expectations.

It changes competition.

And it changes the economics of value.

The question isn't whether AI will continue improving. It's whether we understand what its arrival already means for our businesses.

Are we measuring the right return?

When businesses discuss AI return on investment, the conversation often begins with efficiency.

Hours saved. Lower costs. Faster production. Fewer errors.

Those are legitimate returns.

But they're not the whole picture.

Consider a business that introduces AI and saves ten hours of administrative work each week.

That's useful.

But what actually happens to those ten hours?

Does the business reduce a cost?

Does it create additional capacity?

Does someone spend more time serving customers?

Does the quality of the work improve?

Or does the time simply disappear into other activities without producing a measurable benefit?

Saving time is not automatically the same as creating financial value.

And financial efficiency isn't necessarily the most important opportunity.

What if AI helps a business recognise a customer need it previously missed?

What if it helps preserve the context behind a client relationship so expectations are better understood?

What if it gives capable employees the knowledge they need to make decisions without everything returning to the founder?

What if it enables a service the business couldn't previously deliver economically?

Those outcomes may be harder to measure than hours saved.

But they can be commercially significant.

The challenge is to define the outcome, establish a baseline and determine what evidence would demonstrate progress.

ROI should begin with the value we're trying to create, not simply the activity we're trying to automate.

Research reinforces the need for commercial discipline

BCG's research on AI value creation has identified difficulty establishing return on investment and prioritising AI opportunities among the challenges organisations face.

In its research involving 1,000 organisations, 66% of respondents identified difficulty establishing ROI on AI opportunities as a challenge, while 59% identified difficulty prioritising opportunities.

These findings don't mean AI investments are inherently unsuccessful.

They point to a management problem: organisations need to become better at identifying where AI can create meaningful value and evaluating whether that value is being realised.

Source: BCG — AI research and insights.

And I think this is where the CEO's role becomes particularly important.

Choosing technology is not the same as choosing a commercial direction.

A business can successfully implement AI and still fail to improve its competitive position.

It can become more efficient at producing something customers are becoming less willing to pay for.

Or it can use AI to discover new sources of value, strengthen customer relationships and develop capabilities that make the organisation more competitive.

The technology may be similar.

The commercial outcome can be very different.

Listen. Learn. Then act.

If there's one lesson I'd carry forward from the digital transformation, it's this.

Listen. Learn. Then act.

Listen to what's changing outside your business.

How are customers using AI? What are they beginning to expect? What can they now do themselves? What are competitors offering differently?

Learn what those changes mean for your own business.

Which parts of your proposition remain valuable? Which are becoming less distinctive? What intelligence and capability already exist inside the organisation? Where are opportunities being lost between people and functions?

Then act.

Choose the opportunities that matter.

Define the expected value.

Test your assumptions.

Give people responsibility.

Measure what changes and adapt.

That doesn't mean waiting until every question has been answered.

Experimentation is part of learning.

But there is a difference between a small experiment designed to answer an important business question and a major investment driven primarily by the fear of falling behind.

One creates understanding.

The other can create activity without direction.

AI is too important to ignore. But it's also too important to approach without commercial judgement.

Don't redesign for the next technology cycle

There's another lesson I take from the disruptions I've experienced throughout my career.

The technology will change again.

It always does.

If we redesign our businesses around the capabilities of today's tools, we risk having to repeat the exercise when those tools evolve.

But if we begin with what customers value, how that value is changing and what our organisations need to become capable of doing, we build something more enduring.

The objective isn't simply to become proficient with today's AI.

It's to develop a business that can listen, learn, adapt and create value as the environment changes.

That requires understanding customers.

It requires connecting knowledge across the organisation.

It requires people who can question assumptions, exercise judgement and take responsibility.

And it requires leadership willing to reconsider how the business works.

AI can help us do all of those things.

But it shouldn't become the reason the business exists.

Know the value you're trying to create before you decide what to change.

Because AI won't be the last disruption your business faces.

Ride the wave of change

I've lived through the introduction of digital marketing, the Global Financial Crisis and COVID-19. Now we're experiencing the arrival of AI.

Each has challenged the way businesses operate, compete and create value.

And AI won't be the last.

That's why I believe the most important thing a CEO can do isn't simply develop a strategy around today's technology.

It's to understand the value their business creates.

Why do customers choose you?

What do they trust you to do?

What problems do you solve that really matter to them?

What would they miss if your business wasn't there tomorrow?

And how might those answers be changing?

When you understand those things, you have a stronger foundation for deciding what needs to change and what should remain.

You can recognise when technology creates an opportunity to serve customers differently.

You can see when a traditional service is becoming less valuable and begin developing something better.

You can connect the knowledge, relationships and capabilities already within your business to create value that wasn't previously possible.

And you can invest in AI with a clearer understanding of what you're trying to achieve.

The objective isn't to build a business around the technology of today.

It's to build a business capable of responding to the opportunities of tomorrow.

Know the value you create before you decide what to change. Because the technology will change again.

And perhaps that's the greatest opportunity AI presents.

Not simply to make our businesses more efficient, but to make us reconsider what truly makes them valuable.

If you understand what you mean to your customers, you're in a much stronger position to ride the wave of change — rather than simply react to it.

That's what I mean by growing better.

Joe Papadatos is the founder of Two Icons, helping established businesses discover commercial opportunities and develop the capability to realise them.

Grow Better.