We got a complaint about Capa. The customer couldn't sleep.

A few weeks ago I got a phone call from one of our customers. He had been testing Capa, our AI CFO, for a few weeks, and I was looking forward to hearing how it was going. "I have a problem with Capa," he said. I'll admit I felt it in my stomach. I use Capa a lot myself and hadn't run into any problems, but that doesn't mean others can't.

"Oh, what kind of problem?" I asked.

"I can't get myself to go to bed in the evening."

That wasn't the answer I was expecting. He explained that he had started using Capa to explore his clients' finances. He asked a question, got an answer that made him curious, and asked another. Then he discovered something else he wanted to look into, and before he knew it, several hours had gone by. "And I'm not the only one with a problem," he said. "My family isn't particularly keen on Capa either." I had to laugh. This wasn't the customer complaint I had prepared myself for.

When one question leads to five more

What made the conversation interesting wasn't that he had lost a few hours of sleep, but why. He had started looking at his clients' finances in a different way: not just checking that the accounts were correct or noting that the result had gone up or down, but exploring the relationships behind the numbers. Why had profitability changed? What was driving the costs? How would liquidity develop if revenue fell? And perhaps most importantly: What could the client do about it? Nobody had asked him to spend his evenings on professional development, and nobody had set up a course or given him a list of analyses. He did it because he wanted to. And that's exactly what got me thinking.

What if we've underestimated the most important effect of AI?

We talk a lot about artificial intelligence in the accounting industry: about automation, efficiency and time savings. That matters, and the potential for freeing up time from routine work is huge. But I wonder whether we're overlooking something that could become even more important.

What if the greatest value of AI isn't primarily the time we save, but the curiosity we spark?

For many years, the industry has said that accountants should move towards financial advisory work. Yet in the survey of member firms by Regnskap Norge, the Norwegian association for accounting firms, advisory accounted for only 4.7 percent of revenue in 2025. We can send people on courses, write strategies and set targets, and all of that has its place, but there's a long way from an ambition of more advisory work to advisory becoming a natural part of the working day. What happens when employees start exploring on their own? When they find something in the accounts they don't understand, become curious about why, and get an answer that makes them want to ask another question? Then something happens that's hard to achieve through a strategy alone.

Learning is no longer just something the organisation expects. It becomes something the employee wants.

From doing the job faster to getting better at the job

If an accountant uses AI to produce a report in five minutes instead of an hour, we've achieved an efficiency gain, and the time freed up has value. But if the same accountant starts looking into why a client's margins are falling, how working capital could be improved or what measures could increase profitability, we've achieved something more. Then we're using technology to develop the skills of the person working with it, and we're making it interesting to use those skills. For me, this is where much of the potential lies: Not in AI making the accountant redundant, but in it helping them become more analytical, more curious and better equipped to understand what's happening in their clients' businesses.

There's also a difference between getting a finished answer and being able to have a conversation about the numbers. When Capa knows the company's finances and the context behind the questions, it's easier to go further than the first analysis: why something has happened, what it means and what options the client has going forward.

Imagine what this could mean for an accounting firm

Imagine a firm with 20 employees where five start exploring their clients' finances this way. Maybe they discover things they wouldn't have noticed before, maybe they start discussing their findings with colleagues, and maybe they become more confident asking questions and suggesting measures. Over time, it might become more natural to pick up the phone to a client and say: "I've noticed something interesting in your numbers. Shall we have a chat?" Perhaps we've been a bit too focused on telling accountants that they should become advisers, and a bit too little focused on giving them tools that actually make them want to. Because advisory work is about being curious enough to ask the right questions and confident enough to discuss the answers with the client.

A phone call I'll remember

One conversation isn't statistics, and I don't know whether this applies to everyone. But the customer had been testing Capa for about four weeks, presented it to his management team, and they wanted to go further. That was good to hear, and it means a lot to a founder. Still, it's the first phone call I'll remember best, because it reminded me why we started Capassa. We didn't want to build yet another system that produces reports. We wanted to make financial insight more accessible, understandable and useful, and to give accountants better opportunities to develop and create value for their clients.

Of course I hope he and his family get a bit more sleep from now on. Because lack of sleep isn't the goal. Engagement is.

And the next time a customer calls and says he has a problem with Capa, I'll still take it seriously. But maybe I'll be a little less nervous before I ask what the problem is.