If the AI value doesn't appear on the bottom line, we haven't made enough progress

How many AI pilots do you have? I’m not sure that’s the right question. Instead, ask what AI has done for productivity, risk, revenue, or margins. AI is making its way onto the executive agenda, and so we must also start placing far tougher demands on the value the technology creates.

I recently had the pleasure of participating in EnergiRike, where I spoke about how AI is moving from pilot projects and experimentation to something much bigger: strategy, leadership, and real business value. For me, this is an important shift.
We’ve spent a lot of time talking about what AI can do. That was necessary. We’ve had to explore, test, learn, and sometimes fail. But now we need to move forward. Because when AI moves onto the leadership agenda, the questions change as well. Then it’s no longer primarily about how many pilot projects we’ve launched or how many employees are using the technology. It’s about what difference it actually makes.

Three reflections are particularly on my mind right now.

People will follow people

There’s a lot of talk about what AI can do faster and better than humans. I find the opposite question almost more interesting:
What can AI free us humans up to do more of?
For me as a leader, this is perhaps one of the most exciting aspects of this development. Much of our workday is spent finding information, analyzing, coordinating, preparing, summarizing, and managing. If AI can help us with some of this, we should use the time we gain to focus on what we’re actually meant to do: Set the direction. Develop people and teams. Stay close to customers. Make better decisions. And ensure results.
I don’t think AI makes leadership any less important. I think the opposite may be true.
The more technology that enters our organizations, the more important the ability to build trust, create meaning, and provide direction becomes.
People will follow people. AI doesn’t replace leadership. But used correctly, technology can give us more time to actually lead.

Technology can’t be introduced after the strategy has been set

We’ve long talked about technology as an enabler of strategy. First, we decide where the business is headed. Then we find the technology that helps us get there. I think we need to challenge that way of thinking.
AI and other exponential technologies are now so powerful that they don’t just influence how we execute our strategy. They influence what is actually possible. Technology can open the door to new products and services, transform customer experiences, fundamentally change processes, and create entirely new business models.
This means that technology cannot simply be something we discuss after the strategy has been decided. Technology must help shape the strategy.
To me, that doesn’t mean that all leaders have to become technology experts. But the leadership team must understand technology well enough to see the opportunities, challenge established assumptions, and ask the right questions. If technology changes the very scope of our opportunities, it belongs in the strategy discussion.

Image from Energirike 2026—panel discussion featuring Sintef, NSM, atNorth, and Cegal—moderated by Fredrik Solvang.

Then comes the question we need to tackle much more rigorously:  What business value is AI actually creating?

Here, I think both we in the IT industry and internal IT organizations still have work to do. We’re good at talking about technology. We like to count pilots, users, implementations, and projects. And those can be relevant metrics along the way. But they don’t necessarily tell us whether the technology has actually improved the business.
I’d rather know:
Have we become more productive? Have we reduced risk? Have we increased revenue? Have we improved margins? Can we do something today that we couldn’t do before?
That’s why I keep coming back to one sentence:
If it’s not visible in the P&L, it doesn’t count

Of course, this doesn’t mean that every effect of AI has to show up as a new line item on the financial statements. But we must be able to demonstrate the connection between the technology we invest in and the business value we expect to get in return. Because if we can’t do that, something quite predictable happens. The investments become harder to justify. Management’s attention shifts elsewhere. And initiatives that might have great potential lose momentum.

A lack of technology isn’t necessarily what’s holding back AI development. An inability to document the value can be just as critical. The fastest path to value doesn’t necessarily start with AI.

This brings me to something I believe is absolutely central. The most interesting opportunities arise when two worlds collide:
- Those who know the business, the customers, the processes, and the actual value drivers.
- And those who know the technology, understand the possibilities, and know what can actually be achieved.

Deep domain expertise combined with strong technological expertise is a powerful combination.

If we only have the technology, we risk creating impressive solutions to problems that don’t really matter. If we only have domain knowledge, we risk improving the current way of working without realizing that technology has opened up entirely new possibilities. But when these two groups sit down together, something happens. That’s when we start asking better questions. And perhaps the most important thing is to stop starting with: Where can we use AI? Instead, start with: Where do we create value in our business, and how can AI help us create significantly more? It may sound like a small difference. I believe it’s absolutely crucial.

Now the real AI work begins

I don’t mean that we should stop experimenting. On the contrary. We need that curiosity. We have to test. We have to learn quickly. And some initiatives will, of course, fail. But we can’t stay stuck in the pilot phase.
AI is becoming a leadership issue. A strategic issue. And a matter of competitiveness. For me, that means we have to juggle multiple tasks at once. We must use AI to free people up to do what people do best. We must stop treating technology merely as a tool for executing strategy and let it become part of how the strategy is shaped. And we must become far better at documenting the business value we actually create. Because there’s plenty of AI potential.

Now it’s the results that count. If it’s not visible on the P&L, it doesn’t count.