One Year, Two C-Level leaders, One Vision: Simon Paris—who joined UNIT4 a year ago—and Claus Jepsen combine a fresh perspective with deep expertise. In an interview with Prof. Dr. Norbert Gronau, they discuss the future of the multinational software provider.
Simon, you are the CEO, so you have to have an overview of next to everything. So how important is Germany as a market for UNIT4?
Simon Paris: I’m the CEO and have been here just over one year, compared to Claus, who has been here over twelve years. So between us, we have a fresh perspective and a very profound perspective. Since my arrival together with Claus and the management team, we have really narrowed the focus of our ideal customer profile. We focus on four verticals in the ERP space: people-centric industries, where people are the business model. This means the public sector, higher education, not-for-profit, and professional services. We’re not into construction or warehousing or logistics or supply chain. It’s that people-centric triangle of people, finance, and projects.

Claus Jepsen: In terms of geographies, we focus on five geographic clusters. The largest cluster by revenue today is the UK and Ireland; the largest cluster by logo count is the Nordics, because UNIT4 was actually fully called UNIT4 Agresso, and it merged with Agresso of Norway. The third largest is then Benelux, which is the historical origin of UNIT4, then North America, which is Canada and the United States, but not Mexico, and then France and DACH. So France and DACH are, of course, the largest economies; they are the largest battlegrounds for ERP vendors. So it’s an absolute priority for me and the team.
We focus very much on the mid-market now. We currently define it as around 200 employees up to 5,000 employees. We can go much higher than that. For example, in Norway we go to 20,000-30,000 employees. We want to be slightly below the radar of an SAP, Workday, Microsoft, and Oracle, but we absolutely intend to be on the radar of a national champion when their customers are becoming international. We are perfect when it becomes “multi-to-the-N”, meaning multi-jurisdiction, multi-language, multi-time zone, multi-currency, and multi-entity.
Simon Paris: Since I arrived, we’ve reorganized UNIT4 into kinds of three swim lanes. The largest swim lane is our ERP customer base as well as the enhancement solutions where we have acquired some businesses to sell and cross-sell to our ERP customer base. For example, we have acquired Scanmarket, a „source to contract“ vendor originally fromDenmark, and we acquired the FP&A business of Prevero in Germany. We then have the CODA swim lane. And CODA, you can think about it as being quite different. CODA is a unified single ledger, and is very much focused on high transaction volume verticals. So think about financial services, manufacturing, transportation, and logistics, and there are around 550 customers, including, obviously, in Germany. And then the third swim lane is really those other assets that we refer to as Cornerstone some of these assets may be more valuable in somebody else’s hands, and we may divest them over the coming two or three years.
Do you think the German market or the DACH market is different from. The Nordics or from the UK? When it comes to a decision about a new ERP system?
Simon Paris: I think we’re very respectful of the idiosyncrasies of all geographies, and Claus, being Danish, can talk to this as well. We do see, for example, in the Nordic region that there is a broader consensus required to execute a transformational project. You may be speaking to and selling to the CFO, but there’ll be much more inclusion of the COO and the CHRO and business process champions and so on. In the UK, it can be a little bit more hierarchical, so it might be a smaller team that makes the decision. It might be a faster decision process. So I think there are idiosyncrasies. Obviously, the German economy is slightly differently constructed than, say, the UK, but it is quite similar to Sweden, for example.
Pioneers in Speech-Based Input
How does UNIT4 intend to succeed in this cutthroat competitive environment? You were one of the very first ERP vendors with a voice-based input and output assistant long before. That was 2017. Is it possible that UNIT4 was too early on the market with that?
Claus Jepsen: With that, we were. This notion of having an entirely different experience with ERP is driven by a conversation. And then we were testing out the voice enablement as well, right? So yes, that was a little too early. But again, the good thing was that we had a lot of our underpinnings in place from a technology perspective that allowed us to pivot pretty quickly to what we have today, which is Ava, our Advanced Virtual Agent. So it was worth the design effort and learning.
Which specific AI use cases are currently running productively in your companies, and which ones will be the next you make available for your customers? Maybe either in the ERP swim lane or in the financial swim lane?
Claus Jepsen: Oh, I think that’s a great question. Let’s talk about the ERP. We have a number of targeted use cases, like helping with timesheet approvals and tasks. The data here is key. The data itself is useless if you don’t have meaning around it. So, the approach we have taken, which is slightly different from anyone else, is that we built a meaning layer such that we can explain what the data means.
And you have the transactional data and the functionality. In between sits the meaning layer. The meaning layer is translating what the user asks for to the actual data points without having to go back to the database. It’s trying to bridge it with what the user is asking for. So, we build that meaning layer that allows you to ask it any question, and it goes to resolve it and comes back with what you need.
This will be going out for early adopters now, so we can roll it out later this year. The next step will then be to do execution. Users ask the system to execute tasks without being constrained about APIs, and without having to build specialized agents. We build a new intelligence stack that goes from system of record to a system of reason to a system of action.
“Double-Key Solution” for Greater Data Sovereignty
What about GDPR and AI? How do you solve this? Because, especially in some of your verticals, there is a special need for non-cloud or local cloud or special cloud services, so you can’t put it all on Amazon or IBM. So what are you doing in that case?
Claus Jepsen: We call it the sovereignty conundrum. There are data residency rules, and there is at the top-level, cloud sovereignty. So, the way we’re addressing that is that we call cloud mobility, the ability to serve our customers from various hyperscalers. We also have what we call the double key solution. The customer brings their own key. We store the key in our private cloud, so it’s not with Microsoft or whomever. On top of that, we have our own keys to get into our estate. UNIT4 has an estate that’s key-protected on Azure, and within that key protection, each customer has their own key as well.
Simon, as a CEO, what drives you in your role?
Simon Paris: I’m very passionate about purpose, and I like the idea of us being a very purpose-driven company. If you look at our background, we talk about giving time back, so we think about what allows us to give time back. So, for example, if you’re working in a nonprofit, giving time back for the volunteers on their mission, or if you’re in the public sector, giving time back for civil service when you’re working in a local municipality, or if you’re in consulting and professional services or an architecture firm, giving time back to serve your clients or be with your family. So we’re very excited about the potential of technology to “give time back for what matters most”. By being very thoughtful about the cognitive load that you give from the human to the machine and then how the machine obviously interacts and consults with the human.
Claus, where do you personally draw the line between meaningful automation through AI and areas where humans must deliberately remain in the driver’s seat?
Claus Jepsen: I don’t know if there is one single answer to this. I think that when you have these systems as we are starting to roll them out, I think you’re going to see a high involvement of humans in the loop to make sure of human judgment to trust. Over time the trust in the systems will increase, and then the human’s involvement will drop off. I don’t know if there is a clear point where we can say this requires humans, this doesn’t require humans. Anything that involves sensitive information about employees, I’ll probably keep strictly human, right? And I think people do that. But if you think about invoice accounting exercises or absence approvals and expense reports, all these elements will be fully automated down the road. That’s not just AI; it’s a combination of new algorithms together with the fact that more and more data is electronically shared.
Professional Expertise at the Leadership Level
What is the one skill that every future leader in the digital age must absolutely master?
Simon Paris: The problem, the challenge, the opportunity is always people. So, I’d always go to the people side of the technology equation, and I think, if I had to choose one thing, I think it would be domain skill. I think that as the capability of technology continues to increase, the context, which is in this case vertical domain expertise, becomes ever more relevant. So, I would always naturally, through instinct, go to the people side of the equation. The capability, the organization, the operating model, the talent skillset, the aptitude, and the attitude. Claus and I spend a lot of time on attitude and aptitude and much less time on pure technology choices.
Aren’t technical skills too much missing on the C-level
Simon Paris: I think a highly functioning team needs a diverse range of experiences and skills, of which technical skills must be at the table, absolutely. I’m very deliberate in the design of my leadership team, listening to the voice of the Chief Technology Officer and the Chief Product Officer. We have three computer scientists on the leadership team. In addition, I also like to hear different perspectives from legal, people, or markets perspectives. So the team must be diverse.
I strongly believe that badly designed processes do not get better because of AI. Inadequate data, architecture, structure, and governance do not get better because of AI. AI works best when it has really reasoned, clean data available to it with good rules that are on the ERP system.
If you have to explain your corporate strategy in one sentence, what would that sentence be?
Simon Paris: European software champion focused on people-centric industries. 85 percent of our revenue is in Europe. So being a European software champion is a very significant posture.
Deepening of the European presence
What three strategic priorities have you set for the next five years? And how do ERP, modernization, and AI specifically contribute to these three strategic priorities?
Simon Paris: Priority number one is to continue deepening our vertical moat, and that will be both organically and inorganically, and it will be through the reasoning layer and the semantic layer through vertical-specific domain understanding in our four or five verticals. Number two strategic priority is deepening our European footprint. Primarily now with the renewed focus on Germany and France. Number three would be getting to our goal scale, moving from where we are today, which is around half a billion euros, to about a billion euros by 2030.
Claus, do you want to add something from the technology side of the strategy?
Claus Jepsen: The technology totally underpins this vision. With all the new technologies coming out, solving some of the regulatory and compliance challenges that typically go into a new market. This will help speed that process up as well, right? So that would allow us to cement our position in Europe in terms of vertical functionality for our core areas. And that is what we constructed the technology around to deliver.
Mehr erfahren über