Lukáš Jac (ŠKODA AUTO): All employees can see real-time reports and how the production meets KPIs
05. 10. 2022
Reading time: 10 minutes
Adastra was one of the pioneers of data warehousing in the Czech Republic and has delivered many solutions to banks, telcos, operators, insurance companies and energy companies over two decades. Eventually it developed its unique warehouse, which it called FinReCo. "We are of great importance to companies with diverse sourcing systems, work with different currencies or are creating a new business of their own that they need to process with reporting uniformly as from other businesses. Often these are acquisitions and mergers," says Adastra business developer Oto Gücklhorn in the Adastra podcast.
Ivana Karhanová: Up to 65 percent of M&A deals fail to meet original financial targets. Two companies, two cultures, two financial and accounting management systems, and two management groups collide. However, they only need one view of the business. One truth on which to base management decisions. Where is the key to data and understanding the business? I'll be talking today with Oto Gücklhorn, a business developer at Adastra.
Oto Gücklhorn: Hi.
Ivana Karhanová: Oto, management reporting needs to be tied to the company's strategy, showing how the KPIs can be met and possibly where the KPIs are lagging. Adastra has developed its tool for this. Why own?
Oto Gücklhorn: Adastra has delivered many solutions to big companies like banks, telcos, operators, insurance companies, and even the energy sector. But the big banks had different specifics. The system was mostly customized. It's just that after a while - and we've been doing this for over twenty years - we found that the basic needs were very similar in terms of some financial data. And since IT people don't like to do something over and over and pretty much the same way, we just decided to develop a tool that would materialize our twenty years of experience and prepare some tool that would allow for quick deployment of the solution, with the data processing scenarios being pre-packaged and making the processing much faster and more efficient. So there's just the know-how that's materialized.
Ivana Karhanová: The financial reporting tool is called FinReCo. If we go to the bottom line, it's a data warehouse. What is unique about it? What makes it different from regular data warehouses? I don't know if there's a regular data warehouse, but I mean from the data warehouses that companies already have.
Oto Gücklhorn: I would say that the main uniqueness of FinReCo is that it materializes 20 years of experience of our staff and that Adastra was one of the pioneers of data warehousing in the Czech Republic. The data warehouse is set up in such a way that nothing ever surprises the customer or us. The data warehouse will always support their reporting whichever way their business moves. This is one of the main reasons why I think our data warehouse is unique.
Ivana Karhanová: Why aren't they able to deploy such a data warehouse just in the IT departments of banks, insurance companies, and companies we supply FinReCo to?
Oto Gücklhorn: Good question. They might be able to, but of course, it's about the fact that bank employees work and gain experience for maybe ten years in just one environment. We work on different projects and gain know-how in multiple banks and institutions, so we can materialize the basis of the solution, which is always the same but faster.
Ivana Karhanová: Often in Adastra, you are dealing with making that single database work, making the company work with one data. At first glance, it's a fairly simple thing. I have data. It flows somewhere. The data should be uniform. So why doesn't it work so well in practice?
Oto Gücklhorn: If we go back to banks, a bank, of course, has several systems, mostly independent and from different vendors. It's just a core system where there are movements on bank accounts. It has some CRM, lending systems, and probably a few other systems. Each system has its dials, its data, and data controls. And then, when you pull data from those systems, and you want to bring them together, it's not exactly a trivial task. You have to link the data together well. That's the basic task of a data warehouse. It's not rocket science. It just rearranges the data from the transactional systems into a form efficient for reporting.
Ivana Karhanová: And the know-how is what? In knowing which data to move to which pile?
Oto Gücklhorn: How do we set it up so that it's easy to pull the data out of it or to report quickly, for example? Because there are a lot of data warehouses that somebody has built - sometimes IT - that worked in the beginning, but maybe after two years, a customer comes to us and says they have problems. And it's usually about how just about every data warehouse works in the beginning. But we have a lot of customers who have been using our data warehouses for ten, fifteen, twenty years. After fifteen years, it's not as easy to make them work. So that's where I would say there's a fundamental difference. And it's just in the right design of data structures and data aggregation.
Ivana Karhanová: Once I have that, how easy is it for me to build another reporting tool, like Power BI or Qlik?
Oto Gücklhorn: Or Excel - that's important to say.
Ivana Karhanová: Or Excel, okay.
Oto Gücklhorn: Then that's just the cherry on top. It almost doesn't matter what tool you put on top of it. The most important thing is that the data is put together correctly. What is more important is that usually, there is a so-called OLAP cube at the end of the data warehouse, which prepares the data so that, basically, even a person who is not familiar with the structures of the source systems will not make a mistake in joining the data. In contrast, if he tried to pull data from different source systems, which technically may not be that complicated, he could make fatal mistakes in joining the data, resulting in wrong numbers.
Ivana Karhanová: Can you give an example of that? What can I link to that is completely wrong?
Oto Gücklhorn: The starkest example is to link crowns to dollars, which can happen in real life. A person who deploys, for example, Power BI directly on the source systems, in theory, it can work, although probably not at an optimal speed, is put in the role of an analytical architect because he has to know exactly the physical data model of the systems to connect the data correctly. And those source systems are also usually optimized to get the data right, not to report from it. And that's a pretty fundamental difference, too. The biggest problems are with the coupling of the physical data in the right physical data models because in that implementation, for example, when we're working on it, we're working with the source system vendor in local IT to translate the business data interpretations from the source systems perspective. And one would have to be able to do all of that, so one would have to be technically proficient and understand the physical data model of that source system.
Ivana Karhanová: I mean, what's falling into what piles?
Oto Gücklhorn: Yes, how the data is stored there and why.
Ivana Karhanová: Normally, data warehouses, accounting, and statutory reports are reported. How is management reporting different?
Oto Gücklhorn: In the latter, we adapt the reports to the needs of managers. I can then do result reports for my department, basically for an individual employee or product for one country. But of course, to get that data, there usually has to be some managerial off-balance sheet adjustments to the data to make it replicate the reality in the data versus the life in that company. So most of the time, I don't get that data directly from the financial statements or the accounting data.
Ivana Karhanová: Does that mean that when you implement FinReCo, you have to talk to the business itself, in addition to IT and the vendors, to tell you what are the key things that they need to see in the data or related to the data?
Oto Gücklhorn: Mainly, they also deal a lot with the so-called dimensions, which are centers, contracts, and so on, and they also have to be structured properly. This also applies to the analytical part of the accounts so that we know what they represent and can then interpret it correctly, for example, from the perspective of management reporting or changing the procedure a little bit. Especially when you have a holding structure, and the source systems are in different countries with different legislative frameworks, they have to be set up properly or somehow sort out the data and the dimensions, so they go into the warehouse correctly.
Ivana Karhanová: I started today's podcast with mergers and acquisitions because you use FinReCo a lot in mergers or acquisitions. Why in these transactions?
Oto Gücklhorn: Because these companies probably need us the most. Firstly, they have to resolve intercompany transactions, and secondly, it just so happens that their sourcing systems are, for example, completely different. Often, when you have a holding structure, for example, four big companies have their ERP system in the Czech Republic. You have another eight companies that already have a system from which you can't do a proper data export, and the data is pulled from Excel. The same is true in Romania, in Hungary, where the size of each subsidiary is different, and that corresponds to a different information system. They often still have their accounting charts, which correspond to the local legislation. Then the problem is to do the reporting in that. So our contribution at that moment is higher than, for example, a company with only one source system and accounts in Czech crowns. The company will probably handle the reporting within the ERP system and will not need our services. These customers do not need our services.
Ivana Karhanová: So where do they benefit the most - in mergers, in acquisitions?
Oto Gücklhorn: Or when they have a lot of different source systems and different currencies. And then I think we have a lot of value for companies with dynamic growth because they are creating their own new business that they need to handle with reporting consistently like from other businesses. Quite often, they are acquiring and entering abroad, or conversely, foreign companies are entering here, which is the biggest potential for our FinReCo solution. And again, if I were to define it the other way around, it's just that if there's a stable company that's been in business for a few years, they have one source system, they can handle their reporting from the ERP system and probably won't need our services.
Ivana Karhanová: Because of your 20 years of experience with the business, you have included period locking in your reporting. Why do companies need this?
Oto Gücklhorn: Every item or transaction in the data warehouse has two dates, one of which is accounting or transactional, and the other is reporting. And suppose you want to deal with some final things based on the data in the data warehouse, like bonus payments, which is an important thing and would not be reputationally welcome to return them. In that case, we ensure that the reporting doesn't move backward. So you lock in the reporting period. So the data that is then reported to the government, taxes, and so on, you move because of, for example, credit or other situations where you just adjust the data. The financial statements are made on an annual basis, so you can change the accounting within the year, but reporting it would be a big problem, because first of all, you would be changing the numbers in the previous months, and then, in turn, you would be multiplying back somewhere - what was written off last time would be added back in the next month. So you would be moving the results over time. That's why we've introduced period locking, where some responsible person, usually the head of controlling, locks the data now. And any further reporting transactions only go into the unlocked periods. This way, reporting cannot move backward.
Ivana Karhanová: Let's give a couple of other examples where management reporting is completely different from any accounting or other legal systems that a company reports in.
Oto Gücklhorn: In many companies, there are internal business agreements that if someone is credited for a contract, some part falls on his department or directly on him. And in such situations, the controlling staff must adjust the statutory accounts. They probably do it by off-balance sheet entries, but that can be done differently. They influence the management accounts to fit the situation. So we can then create income statements for each department or person.
Ivana Karhanová: So how does that correspond to the business realistically, what is written in the accounts, because invoices can also be very late or very large.
Oto Gücklhorn: Not only that, but you don't capture the share there. If you imagine, for example, that you have two departments behind you and then there is some final invoicing, you have to split it somewhere, what belongs to you and whoever is behind you.
Ivana Karhanová: From that management reporting, I want to see, for example, how does which department contribute to my business, what proportion?
Oto Gücklhorn: Or maybe to the profitability of the products, how specific departments or companies contributed to it, with the final return coming from sales. There is accounting and certain tools for it, but it is not always the easiest way to capture everything. Or conversely, there are various legislative constraints in terms of taxes. Like how you can value those products, or between products, semi-finished products, but realistically you want to do it differently because you feel that the benefit doesn't quite match the legislative requirements, so that's how you can adjust the accounting to then value those workers or whole departments correctly.
Ivana Karhanová: How long does it take you to understand the internal rules of the business when someone wants to deploy FinReCo?
Oto Gücklhorn: We approach it a little bit differently. We usually do some quick analysis and try to do a first implementation that meets 80 percent of the business requirements within three months. So we're going for what tends to be pretty much the same thing. We'll implement that and then work out any specialties, but at that point, the company is reaping the benefits of FinReCo.
Ivana Karhanová: How long does it take you to fine-tune the details or the specifics?
Oto Gücklhorn: I'm not able to say exactly, but we usually go through some further development, that we focus on some business besides accounting, and there is already prioritization. Quite often, it's planned out in the project, but when we deploy the first part, the priorities come from. The situation changes. If I were to generalize, we have three phases planned. The first one is a given, and it usually goes through.
Ivana Karhanová: That's the 80 percent.
Oto Gücklhorn: Yes, and the second one will take place maybe eighty percent as originally thought. The third one often doesn't go through at all because there are other priorities, where the customer pushes us to do something else first, or they find out that they didn't even need it. That happens, too, because FinReCo solves it for him differently. And even though he insisted on it before, and we explained to him that he was going to have it there anyway, maybe he insisted on it. But that's the normal life of IT projects.
Ivana Karhanová: So the first phase covers most things that are the same, and then the other two are exactly what?
Oto Gücklhorn: That's where the more complex things are finished, or things that need some more preparation.
Ivana Karhanová: They are already nuances, where I adjust the numbers as I need them.
Oto Gücklhorn: You import data from different countries and information systems. In the beginning, for example, it's done from Excel, and we say that in the second stage, we will deal with the software supplier to do some data exports so that he can prepare it for us. And there, it depends a little bit on the third party whether or not they prepare it for us, so it's done accordingly, and at the beginning, it's just going to be an accounting transaction in Excel. At that point, that's enough for us, and then it's more convenient, and it can be automated. With FinReCo, process automation is also important. What we're able to achieve is that there is some sort of business workflow in FinReCo that allows you to schedule the whole closing process, so how the month-end is going to go, and knowing what is needed to get that data ready.
Ivana Karhanová: I think automation is simpler than it is in practice. How many customers can achieve full automation with FinReCo?
Oto Gücklhorn: I think everybody. Even if, during the checking process, maybe some errors are found in FinReCo, there is the possibility to catch up with the correct data there, correct it in the system and send it back to FinReCo. The way it usually works is that the data is loaded overnight. Still, within the data warehouse's closing operations, they can only catch up on corrections, management adjustments for management reporting, or even other transactions. But that's a small file to pull in and quickly recalculate, for example, an OLAP cube. So they can do that round in maybe even ten minutes where they find an error in the data warehouse, correct that data, and get it back in there in ten minutes. That's not normally used because it's not needed, but just before the month-end deadline, they might do it ten times if they have errors.
Ivana Karhanová: And if I don't have this special data warehouse, the reporting automation doesn't work?
Oto Gücklhorn: It depends on how I deal with it. There are more options. Either I make an excel hell, so I combine different data from different source systems and have made different adjustments. I have special nomenclature for that, numbering, how to combine it, and what to put where. And that's the hell of it. This process is perfectly correct up to a point before it becomes that hell. I think that Excel is a great tool for a lot of things, specifically in the reporting area, but only as long as the complexity of the data and/or the complexity of the reporting is such that it allows you to be very comfortable with the Excel that you're familiar with. Then there's no reason to change it, but the moment controllers or other staff spend most of their time merging files, finding errors, and not doing their job to analyze the data and support the business based on it, then it's wrong.
Ivana Karhanová: I was also going to ask you when is the point where I know I've crossed the permissible line of working with Excel.
Oto Gücklhorn: When people start leaving. When they're just frustrated that they're just preparing the data and not working with it, they feel like they're just an efficient machine preparing the data for the closure. But, still, what they should be doing, analyzing the data, they're not doing that at all. And then we just help that company by helping them maybe ninety to a hundred percent with the data preparation, and then those auditors do their job.
Ivana Karhanová: So that's the point at which I should throw away Excel?
Oto Gücklhorn: Yes, that's the point where it's just spending most of the time preparing the data.
Ivana Karhanová: Are there any other moments when you know from experience that it's pointless to manually prepare this?
Oto Gücklhorn: When you find out in some meetings that you have different truths, that the sales director brings completely different documents than the financial ones, that's where I would also wonder what it is. And usually, it's just that somebody is pulling data based on knowing the physical data model from one place, and the other department is pulling it from another place. The result, of course, is different results in the data. I don't know who's more optimistic. I usually think of the business manager. But then, you can't run that business if everyone has different results. So that's another thing where people can't do it manually anymore.
Ivana Karhanová: So that's when you automate the whole reporting and move to a dedicated solution?
Oto Gücklhorn: Yes. It can even happen that we just don't have the truth completely accurate because we make some distortions there for the sake of speed and performance, for example. But the advantage is that the data is still the same from all perspectives. We have the same data. When the data partitioning is very complex, it can be done in a simplified way. It may not be completely accurate in crowns, but in millions, to billions, we don't mind. All the reports still come out with the same numbers when you aggregate them as you have. And that's the one truth.
Ivana Karhanová: So, does that mean I'm making decisions based on a single database?
Oto Gücklhorn: Yes.
Ivana Karhanová: Says Oto Gücklhorn. Thanks for coming into the studio and letting us talk about management reporting. Bye.
Oto Gücklhorn: Bye.