Rethink! Corporate Finance 2019: Interview mit Marco Nix, CFO, 50Hertz Transmission

Yes, perhaps if you... I myself am from the company 50 Hertz, which is a transmission system operator. We operate about 10,000 kilometers of overhead lines, very high masts over long distances. We operate the electrical system as a whole, ensuring that generation and consumption are always balanced, and since electricity cannot yet be stored at this scale, we have tasks that are particularly driven by the energy transition. On one side, renewable energy in the north, and consumption still in the south, which strongly drives us. Also, due to the many decentralized elements like home storage, photovoltaic systems on roofs, and the digitalization of measurement and control data, we are affected by a strong digitalization trend. Personally, I have been with the company for ten years, now for four years as... in these four years I have also been responsible for IT, which will be handed over this year. We see that this function needs its own space, and we will strive to create a digital office that will take care of all these digitalization systems. We will have a workshop tomorrow about the classification of the finance function. So, less clarity, probably a more controversial debate. Overall, how to approach the topic today at the conference, I've already listened to how to structure it. A clear strategy, the builder, and the truth probably lies somewhere in the middle. And above all, taking the participants' viewpoints and experiences is what interests me, and that is what I want to give the group.

Well, both things are naturally incomplete. A strategy tends towards arbitrariness, where you write down big words like artificial intelligence, robotics, without saying exactly what that means. On the other hand, there is the danger of working too quickly and in a too fragmented manner, and then a large block in the middle is missing. But at 50 Hertz, we approach concrete topics and try to develop a feeling for what scaling, benefits, and implementation duration mean through various use cases that we try to implement. Especially for companies with larger lives, like ours with 100 employees, I think they find it extremely difficult to design three-year transformation processes. So the question is whether evolution rather than revolution is the better path. Clearly, SAP has many modules, and you can set it to a new level with a new core and modern components around it. But then it is likely that the finance organization will be busy for three years, and all the freedom you try to create may already be exhausted because time is running very fast, and especially the value creation from analytical activities and controlling, for example, cannot be done. So I believe that we must allocate our resources, have an overview of all systems, but definitely approach them one after another. Especially since we don't have an idea yet of what the result will look like. It should not be a rash action but a continuous process. Our cases are much better that way. I can't say if that's the case everywhere. There are advantages if I say I'll do a business project and go through it, but my fear is that it takes a very long time and is very tiring, and one tends not to do things, and if you are unlucky, the technology is already outdated before it is fully rolled out. And that would actually be a waste.

We have already made some attempts. We have several robotics solutions in use. For example, we have done the homework on the most boring and unsexy thing, namely structuring our data. We tried to automatically assign incoming goods receipts via scanning, but because our text recognition software is not yet mature enough, the system only processed 50 percent. So with those experience values, we have now incorporated further robotics solutions, so that we previously sorted our data structure and thus gained direct benefit. How do we identify processes to be digitized? We set priorities. At the moment, we are still very much in discussion with employees. We did an inventory of processes and came up with 50 processes. In various workshops, we discussed where the benefit could lie in automating or digitalizing them, also at the interface to controlling, for example. From that, some use cases emerged. For example, for Opel, it's paper-based. We have a robotics solution on the other side. Our energy business ensures that generation and consumption are balanced, and electricity transmission usually has losses. We procure these losses step by step. In this area, driven by the energy business, we have built a machine learning system with a neural network to improve the forecast. That has a direct benefit, and this direct usability leads to higher acceptance. The thesis is that this also creates scalability. Because if that works, I also dare to tackle larger and more important projects.

So very much at the moment still in discussion with employees. We did an inventory of processes and came up with 50 processes. In various workshops, we discussed where the benefit could lie in automating or digitalizing them, also at the interface to controlling, for example. From that, some use cases emerged. For example, for Opel, it's paper-based. We have a robotics solution on the other side. Our energy business ensures that generation and consumption are balanced, and electricity transmission usually has losses. We procure these losses step by step. In this area, driven by the energy business, we have built a machine learning system with a neural network to improve the forecast. That has a direct benefit, and this direct usability leads to higher acceptance. The thesis is that this also creates scalability. Because if that works, I also dare to tackle larger and more important projects.

That is quite visible. But it is less driven from top down, but from the bottom up. Employees demand more freedom and design freedom in technological matters, and they bring in their own know-how. When we recruit university graduates today, our colleagues sometimes have extremely high skills in scripting and open-source software, which the long-term staff do not have. This creates new constellations that foster such development.

The first topic, which is perhaps less a topic for the conference but is a fundamental homework, is to restructure one's data structure. That is a key success factor for everything that sits on top of it in terms of artificial intelligence and robotics. That's the top priority, clearly. The second is, once I have done that, to make it usable, to find many use cases that go beyond what can be done with Excel and other applications in Germany, and to actually derive value from them. The third point, which we already touched on briefly, is how to drive this within the company. Does it come from the workforce, but it's difficult to get people to think about it in certain formats, or do I bring in a consultant in the classic way? I think that is a totally exciting question that many of my colleagues are grappling with. And I think many still underestimate the effort involved in data structuring. So the third point is: should you build know-how in your own team or continue to hire consultants?

In our company, we generally have a relatively high outsourcing rate. Roughly estimated, we have a 50-50 ratio for our core activities. I think there is a good reason for that. A certain degree of flexibility is maintained. Basically, we have always had ministers? Actually, we have no competence to judge whether the consultant is doing a good job or not. I think that is absolutely necessary, and you can sometimes go beyond that ratio. It is not a dogma, but an indicator of how much you have acted yourself.