PegaWorld | 44:00
PegaWorld iNspire 2024: Cash is King - How Siemens Utilizes the Strength of Pega to Digitalize Collections
Transcript:
- My name is Fred Kuni. I'm the one Daniel mentioned on the stage today. So I made it virtually to the stage. So that was my pride of the day because I was the one to make the , so with the blueprint. So you've heard the word blueprint a couple of times. Try it out. Try it out. So I have the privilege to supervise a wonderful team, and this wonderful team is working with great customers and one of these great customers you've seen on the stage of Deutsche . The other great customer is the company Siemens, which I've been working actually for long, long years also prior to be at Pega. Today, I thought Elvis is a king, but Nico taught it better to me, cash is the king, and what we are talking today is about collections and how Siemens utilizes the strength of Pega to digitize collection process. So we've got Asrin Yaman, she's the application program manager, and Nico Mueller is the global process owner for order to cash. And I'm very happy to listen to the presentation. After the presentation, we will have a round of questions. So, and I would please beg, let's say ask you for the questions, go to the mic, go to the microphone, because this session would be broadcasted afterwards, and it's better when we get all the information. Thank you for your attention. And Asrin, Nico, the stage is yours. Rock the stage.
- Okay. So welcome everyone again. So just to give you a little bit of overview about our content today, we will be mentioning a little bit about Siemens and global business services, and then we will share some insights about our collection solution and we will share six lessons learned through our journey with Pega, and we will mention about our innovation roadmap. So Nico, can you tell us a little bit about Siemens?
- Yeah, I can do that. So Siemens is a leading technology provider, works in four main industrial domains. So in digital industries, infrastructure, mobility, and healthcare. So what we are doing is we are building and enabling resource effective factories. We are building smarter homes, we are taking care of comfortable transportation, on time transportation and advanced medical support. The two of us are working in a global business services division that are operating and supporting those businesses in different processes. So for example, in procure to pay process, so starting from the purchase order up to the invoice, the hire to retire process or in the order to cash world. And as this is one of the topics we are driving, I want to go a little bit deeper into there because then you also can see where Pega supports us in that journey. So for us, we don't call it order to cash, we call it opportunity to cash because we want to start at a little, wanted to start a little bit earlier in the process. So we start with the opportunity, so the wholesale cycle basically, right? So lead management, sales service desks, along those lines. And there we have a relatively young application that is helping us to work with leads and process leads in time as they're required. The second piece is the order management piece. So it's basically customer send us orders. We are processing those orders and help the cycle up to the billing. Once the bill is out, that's the handover to collections because from that moment we take over, in the collection space and handle it from there. And then you see a payment, you have an allocation process, and then the cycle starts again basically with after sales. And what we've been doing, starting on a transactional journey, right? Many, many years ago. But as shared service organization, you constantly have a drive to optimize, you have a drive to harmonize, you need to bring efficiencies. So this is some and along those lines we are moving more and more into the digital transformation for our operations and especially the harmonization of it. Good. And as we have seen the order to cash cycle right now and not Elvis is the king anymore, but cash is the king. I believe Asrin will help us or explain us or explain you a bit how we use Pega to transform really the collections process in Siemens on a global scale.
- To understand that, let's travel back in time to 2015. So Siemens management took the decision to hand over all the cash collection activities to GBS. And when we assessed the process, we realized we had scattered landscape of total. So collection was done by using SAP, by using Outlook, notepads, and maybe Excel files on desktop of a collector. And all the knowledge was in the brain of the collectors. So we didn't have transparency over the process across the globe. And the second, we had an overwhelmed workforce due to the manual extraction and the distribution of the correspondences and due to manual reporting and also undiversified collection approach because it was making it very tough to reach our collection targets. Again, cash is king. So what we wanted to achieve is basically a harmonized solution that can work globally and we can still fully integrate into our business processes with enough flexibility. And second, we wanted to basically remove IT dependencies from the process. We wanted to simplify the work for our collection managers and also the collectors. And last, we wanted to basically improve cash flow, gain productivity, increase transparency, and drive automation. So with all these targets, we split it collection into two different phases. So the first phase is preparation and the second phase is execution. After all those many years of collection experience, what we realized basically customer segmentation is key for collection. So you need to be able to segment your customers based on their payment behavior, based on their size, based on their volume. And then you need to be able to assign a collection strategy for those customer portfolios. So for example, low volume, good paying customers maybe don't require manual interaction, right? You can fully automate and have a collection strategy accordingly, but for the large accounts where you have more problematic, maybe you need more manual touch and that's where you need to maybe use your resources to focus on. And the third phase is basically creation of the tasks based on your customer portfolio and receivables based on your basically strategies, executing the tasks or creation and assigning of those tasks to the resources. And also in the execution phase, we wanted to have a very front end where we can see all the collection history, see the recent status, and also execute some actions right with just one click. Also, when we talk about collections, dispute is a big portion of it. So dispute management is also done in the tool where we have the problems maybe, then we can target it. And the reporting is the last phase, having adult reports for the operational base or the management level in order to drive collections. Moving on, when we assessed this, we had to choose, right? Either buy a solution off the shelf from the market or make our own solution. And there were several collection solutions in the market, but what none of them was able to fully integrate into our business processes and manage the complexity we have in Siemens. So that's why we gave the decision to make our own solution and we wanted to make it with Pega. And the reason behind it, we are able to integrate across multiple processes, remember the portfolio that Nico just showed us, right? So we need to interact between the process steps. The second topic, reuse of available components. So imagine a simple interface, how many times we can reuse it. And also the situational layer cake, being able to introduce localizations is done also with the support of situational layer cake. And part of our cybersecurity concerns was as already addressed by our central of excellence team in the Pega platform. And Pega also enabled us to have an oddity story also track the performance. Last, Pega has proven to scale that you can manage basically highly complex algorithms as well as a huge volume of data which we had back then. So with this, Nico, can you tell us the lessons that we learned?
- Yes, I can do so. So what we thought would make sense for you is to share our journey and the journey you've heard is a number of years old from evaluating, deciding, piloting, implementing, changing, and operating. And there's that nice quote of Mark Twain, right? "Good decisions come from experience and experience comes from bad decisions." So we will expose both, and you will say, see, that six lessons. So the number one is, I mean the application needs a purpose, right? And the main topic, what you need to do building those applications is you need to drive adoption, right? In a SaaS world, you see that, I'm not sure whether you're familiar with the LAER model. So land, adopt, expand, renew, right? So adoption is at the key because if you have an application which is not used and people pivot left and right, it doesn't bring you the value you intend to do. So what we've done is we said, well, the collector right now organizers day himself, so we want to organize the day for him. So we said, okay, those are the topics we prioritize. So, those are the tasks you need to do. You have different ways how to process it, but at the end you need to fulfill the targets, processing those kind of topics. And also easy access to disputes for example. So just to explain, when you look in the collection space, what you aim in the call or in a discussion with the customer, either want to hear, I pay you tomorrow or you want to hear there is a problem with the delivery because then you have a moment to start fixing it. So you need what we did in the past, we sent emails to somebody and didn't follow up. Right now at the glance we see when we started the dispute with whom and it's in full control. The second topic and ask was elaborating a bit on that, every collector had at least two screens because you had all the different applications and it was a big burden because it was a lot of data tourism from left to right and we said let's build one screen which has everything. So which has the history, which has the items, the open items which have the master data, which have kind of credit scoring to say, well, based on that I can do decisions. And then the second topic is if I am able to do decisions, I want to execute, I want to utilize actions in there. And one of those actions I want to show you, which is that was one of the big killer features that we had was in sending invoice copies. And maybe that's the thing you can learn when you're dealing with Siemens, the typical thing what happens at the very beginning when you call the customer first, oh I didn't receive the invoice, can you send me an invoice copy? And what it was in the past, it took a long time and then you couldn't follow up. And what we have done is to said, we take that as our killer feature, one of the main killer features to say how we are able to compress the time. And we have a small video here where we recorded before and after, and I think it's a quite impressive one. So on the left side you see a recording of one of the SAP system and different solutions that use, it's basically extracting the invoice, finding the contact, writing an email, making some notes, finishing a diary, sending the email, right? And you will see even my text is over, the story continues, right? So that was quite an, that was the killer feature which we had, which every collector loved from absolutely day one, right? And just for to get the numbers right, we are processing around 600, 650,000 invoice copies per year, right? Not all of them are manual, right? So this is showing in a talk process, of course we can integrate them in a strategy and send them out automatically if we know that the customer always ask for them, right? Just to make sure, send it with a courtesy letter. But as an ad hoc task, we are simplifying. So the only thing which remains with the collector is the decision. And as you can imagine that's maybe something going forward, and what you have learned the last two days, you might have an idea in what direction it goes in the future.
- Okay, this brings us to the second lessons learned. Bandi made the rollouts, we always had smaller teams, right? And if you don't want to burn down your team with the workload or with the service management effort, you need to hand over those tasks to the business. So what we found is basically wherever it makes a big business impact and basically risk of the burning down the application is low, we wanted to hand over those tasks to the business. And for us, those were basically templates, correspondences, and reports. Because the business team can basically modify themselves and have the changes immediately in the tool and use it and enjoy it. And I mean imagine if you want to change a template in SAP, how long do you need to spend to find an available resource? It's a lot, right? So this way we were avoiding those dependencies. The second area, what we did also where the business impact is high, but also risk is high for those areas we build a modular function, which is called strategies also for us. And in this scenario we enabled business to make their changes in a controlled way so that they don't burn down the application, but they can still modify it with certain accesses that maybe they can choose the type of the strategy, right? What kind of tasks that they wanted to be created or the frequency to reach to a customer or the date or the type of the attachment. So these kind of things basically was provided to the business to change. And the second topic is situational layer cake enabled us to have our global solution with the local flavors. So I will give my favorite example on this, dispute management. Dispute management is done very complex way in US for example, but whereas the rest of the Europe, you don't need that complexity. So situational layer cake enabled us to have the same application but still introduce those differences. Maybe it comes from government to certain countries or maybe the business dynamics. And this way we were able to drive the local flavors to the solution.
- Good. Coming to the third lessons learned, a couple of findings in the rollout. I think it starts already with the selection of the pilot and that was a bad decision of one of my previous programs because there I said if it works in Argentina, they're not just for you, it works everywhere, but it didn't work in Argentina but everywhere else. So from that perspective we said, we need a well balanced pilot with enough volume, but not a super extreme complexity. So picked the UK, I think it was a success story for us, I'm still their favorite German, at least at Lee, until Asrin gets us a German citizenship, right? So that was number one. Number two is how to cue the rollout properly. So what we've done was to say, okay, we looked at the productivity that we can expect based on applications that we are replaced, based on factor costs in the country and so on. And we queued it accordingly. That was working very well with the one downside that I had to spend the February in with minus 29 degrees in Toronto. But apart from that, I was working fine. The third topic is designed to roll out, and I think this is something where the situational layer cake as Asrin explained, helps you enormously. I'm a headquarter person so I say, well, you use, you define a standard and that's what you do everywhere, but it will not work. You always, and Fred, I think you called it the last mile, right? So there's always something, what you need to do in order to make it work, be it certain legal requirements, be it cultural topics, and especially on the collection side, culture is very, very different, right? So we need to build something that has, and follows a standard, but is configurable in a way that you're able to be flexible in the regional rollout. And the last positive decision, and that might be something which is related to Siemens. In Siemens we have tons of ERPs, all SAPs. So we are kind of standardized, but we are talking about, I mean we are coming from 400 many years back to 60 recently, and I think now we are luckily down to 30, but all with very different contacts, all with very different processes, and getting a change there is a heart, a job. So what we've done is in Siemens we have a data cloud where all the systems deliver data in real time. And we use this because we have full access across all the different systems and we have the possibility to modify fast. And especially when it's about reading data, it's one of the local flavors, ah, I need that field in addition, ah, I need this field in addition. When you do that in an ERP by ERP, you already struggle with the first implementation, with the first rollout. If you want change, you most probably see then the number of SAP system in extractor versions, right? So if you have the possibility to link to a data lake and that's reliable and available, that's definitely also a very good decision that we've taken for the project.
- By the way, we are live in Argentina, I dunno-
- Exactly. Argentina is working for CCMT now, and also for the other solutions. So, but it's not, was not the best starting point at that other program. So the next lesson learned is mind the scale, right? So when we started with it, there was a lot of sentiment around, ah, yeah, we are working with a hyperscaler, that scaling automatically, it will work out. And in fact it worked until a certain scale where we were working or rolling out in Europe and Canada as well, right? But the moment we hit the US, the volumes just exploded in a way. And then we ran into issues with performance, right? And luckily you can do a lot about performance, right? We would've wished to do that right from the start, but maybe a few hints, right? I mean you can do a lot about the optimizations of the tables and views, externalization of documents, right? Storing documents in S3 buckets, archiving, but one particular one is also decoupling of interfaces, right? So let's assume you perform an activity, it should write back into SAP. If you write back synchronously, you need to wait until the system responds. In some cases the system doesn't respond and the user looks at the sand clock, right? That's something what you avoid. So what you do is you decouple and monitor basically the queue to make sure that those kind of things are processed, right? Monitoring was a big topic, a big lessons for us as well to say how to make sure that the data is right. When it came overall to the performance, what we did, so we were in the challenge that the system performance deteriorated and we needed to save that. So what we did was we built that, well, it's a live picture and that doesn't look very fancy, but what we did, we said what are the 30, 40 steps that I need, that the collector needs to do on a day-to-day basis? And we stacked that. So measure the time we stacked that, and then we said, naturally, how long would you want to wait for a step, right? And my reference is my 10-year-old son, right? Who is trying to stream a 4K Netflix video and get super nervous if it takes longer than five seconds. So we came to the point to say, well, it's maybe two or three seconds. So in that, that's what you see, that's a thin red line there, which represents basically our target. And you see it was a very big step to get there, but that gave you those synthetic KPIs that in itself the number doesn't tell you anything. But in KPIs it's the trend is your friend, right? So you need to see how to bring it down, and the nature of the KPI allows you to work on very different elements in order to bring things forward. And the last thing, what we've learned, and I'm not sure whether you are aware of the watermelon effect, does anybody know the watermelon effect?
- I heard.
- Okay, so it's basically you look at the watermelon and it's perfectly green, right? As your KPIs, but if you dig in, you suddenly realize it's red, right? So what we try to do is in the management team, we said, what is success for us? And we built the KPIs, but we also said success is I call a random collector and if the random collector says this is fine for me, then we have another checkpoint and another pulse check to say, well, application performance is fine. We called it the Zibby Test because Zibby was our favorite collector here in the US who was very outspoken and always tells the truth. If he doesn't like it, you feel it very, very well. So he was a great, a measure point for us to say now we actually have achieved success.
- Okay, so the fifth lessons learned, when you have the right people on board, every challenge becomes an opportunity. So this is one thing we want to mention to choose your partners right, in a right way. So I mean all of us has very strong support from the Pega team, but also we were lucky to have an internal center of excellence team who were supporting us, also IT professionals in-house that understands ERP landscape and have integration really works. We had great solution partner, Celltech, and also very strong project management team who is basically linking all these skilled talents together. So this brings us to the last lessons learned, Nico.
- Yeah. And it might sound familiar, right? Did we mention adoption already? So I mean as long as we are not reaching the autonomous entity or the autonomous enterprise, you need to deal with people. And when we look at the success of the application, it was mainly by the usage and by the adoption of it, right? So we have a couple of points, it's a bit of a summary, but there are also some new points. Make sure that you make the life easier, life of the users and customers easier, right? So this is the main topic, right? If it doesn't bring a value add, and you just start capturing things in, it most probably doesn't get the adoption you want, right? The second topic is adaptability and flexibility, but at the same time you still want to have harmonization, right? It doesn't make sense if you implement 150 pilots, right? So you want to make sure it's that balance of on the one end side I'm adopting or I'm adjusting to a certain extent. But on the other hand side, the majority of the application is the same. And what you can change is configuration, right? The third thing, and I think everybody who's driving sales here, the best sell is always using references, right? So we are reference, our users that are happy with the solution, our managers that are happy with the solution, and they help us advocating the application. That's something if you hear it, and I think that was a statement today as well. If you hear that out of the voice of your customer promoting you, I think that's the best thing you can get.
- Remember Zibby.
- Exactly, remember Zibby. And then trust in data. This is a, I mean maybe the collection solution is a bit more data heavy than normal Pega solutions because at any given point in time the collector need to have access to all the open item in order to do conversations with the customer, right? But what you have in data, you have temporary differences. So I push the button right now to send out a dunning letter. A minute later somebody allocates the invoice, the invoice is gone. So you always have those kind of differences. What you want to avoid is that your user base is constantly checking, ah, what's the value in SAP? Just to be sure. What's the value here? Just to be sure. Because the reconciliation that kills you and it distracts adoption quite heavy. So you need to find ways how people trust into the data. And then it's actually a kind of a spiral up, to say people trust in your data, they check less, they find less, they trust more, right? Have in mind that it also can go the other way around, right? Last but not least, and I think this is something which is very, very close to my heart. We very much talk about the Amazon experience in business to business, business to business or customer, end-customer experience in the applications, performance is really key, right? So if you need to look at the window and it takes time to build up, I think this is something which is not helping in the adoption. Good. Finishing the lessons with good decisions and bad decisions. Asrin, maybe you can explain us a little bit on what was the impact of the application that we had and what's next.
- So looking at the impact, in different rollouts, we started with different stages, let's say, right? We had countries like Brazil where we had 20% overdue and we were able to reduce it down to 5%. We had countries like UK, they were, they didn't have any overdue problem because they were already investing so much manual effort to do collections and we were able to gain 40% for productivity with CCMT. And we had some other countries like US, they were already using another collection solution which we replaced with CCMT and we were able to gain 5% additional productivity, but also in the previous solution they weren't able to automate the tasks. So we were able to automate 70% of their day-to-day tasks with CCMT. So the impact is really Verizon, but it's considerable. And this is not the place where we want to stop. We want to continue to follow the business trends and the technological progress as many of you are doing also being here today, and maybe let's mention a little bit of our roadmap. What is next? So remember the video Nico showed you for the invoice copy, and we left giving the decision to send the invoice copy to the collector themselves. So what we want to do, we have already the muscle in place and we want to have the artificial brain to take also that decisioning from the user themselves. In the last years we automated so much outbound correspondences to our customers and as a result we received a lot of inbound communications back. It's millions of emails we are talking about. And also with the COVID, the interactions with the customer changed from more calls than to more emails. So what we want to do is basically having GenAI giving the decisions, also understanding if customer needs an invoice copy, if they're raising a dispute, if they need a balance statement, and then automatically performing it. So this is the next step we are working. The other one, calls are very important still for the collections, right? It's the personal touch we have with the customers. And when a collector today makes a call, they need to come back to Pega and key in all the data, what happened in that call manually. And the impact of Pega is minimal. So what we want to do is basically with the support of GenAI, automatically transcript, summarize those calls. And you can imagine, right? The picture can look completely different. And we were mentioning also, right? Earlier adoption is key. So for the call tasks, this is also the next step that we want to move. The third one, we want to leveraging B2C in daily practice of B2B. So how often do you call someone from Amazon? I personally never call someone from Amazon, right? Everything works well. So this is another topic that we want to bring into B2B world. We want to provide information to our customers as much as possible through customer portals where they can see all the information regarding their payment statuses. And this would enable them to have less manual interactions also with our collections team. So it's win-win for the accounts payable department of our customers as well as our collections team in Siemens. The other topic, some of our collection teams are aging and average age is about 55 and they retire, it's so hard to find the right talents to replace them. So we need to drive more automation and we need to leverage from GenAI more. So this is the next step also that's with all these projects we want to basically achieve. Last but not least, Siemens had a lot of strategical decisions to change its portfolio in the last years. And a lot of parts of the business is spinned off right now and we are providing CCMT now as a collection solution to help them so that they can perform their own collection activities. But also CCMT is used by Pega as collection solution and sometimes when we don't pay, they chase Nico with dunning letters that comes from CCMT. So we are basically chased by our own solution and this is a topic that we also want to do more in the next years. So in case you'd like to have more discussions with Nico and myself, we will be in the innovation hub with our solution partner. You can always reach us. So thank you very much today for listening to us, and Fred, I guess we can move to the questions.
- Yes, we can open the question session. So I learned one thing. So you love red, but there is one red thing you don't like. These are red KPIs, right?
- True.
- Okay, so you like red cars, red dresses, but no red KPIs. Very good. Okay, so do we have any questions in the room? So if you have a question, go to mic and address our moderators. I have questions and I will start with one question. So you told me you are a talented dancer. You told me you're a dancer, but I think you're a talented dancer, in dance, like in project it's all about doing the right thing at the right moment with the music, not in a project but at the right moment. So if you roll back everything and you would like to start over, to start again, what would you do differently?
- I think minding the scale would be the first thing I would do because we spent so much effort after having the scale, we looked back and then we spent so much effort to make it better. So if I would be able to go back, I would mind the scale or vision differently and design for the scale.
- [Fred] Okay.
- That would be my response. I don't know, you're not a good dancer, but maybe if you like to add.
- And you, Nico.
- Thanks for your offense. Well, out of my perspective, it would go in a, that would be my ultimate answer as well, right? Make sure that you have the right scale in place. Make sure that you build from the very beginning in the modular fashion, right? I mean having the knowledge of capabilities that Pega has today. Of course there are also topics, I mean, traveling back in time maybe would have enabled them at that time, but right now what you have available, there are a lot of cool features that you would be able to utilize today as well, which we incrementally build in right now over time.
- [Fred] Yeah.
- So we are not giving Argentina this as a response.
- That's okay. Very good. Argentina was not the thing you do here.
- I guess you enjoyed the travel.
- I've got another. Nico, you you saw the world, you really, so you've been in countries that let's say no tourist would set foot in. Okay, because you've been covering SAP or ERP implementations for Siemens in the whole world, in Africa, et cetera. So, and you do, you know very well that the problematic of managing local differences and harmonized systems. So, how do you handle that concretely in these projects? We touched that a little bit with the differences in the culture, et cetera. But I really would like to have some plastic examples. What is happening? Some anecdotes about that. So how do you do that?
- I think what you need to do is, I mean let's take this strategy. I mean we mentioned before, right? You need to segment customers, good payers, bad payers, big customers, small customers, and you handle them differently, right? And what we could have done, and this is by the way, how many collection solution on the market do they give it to developers to build those kind of strategies, right? And this is something which will slow you down, or result in a capacity problem, right? Or result into costs. So what we have done was basically say you configured a strategy, so you have elements, right? And those elements are modular and people with the right authorization, they are able to build that very, very flexible. So I think this is one of the topics. What you see as well in countries, you see more and more that countries start to certify invoices. Mexico, for example, Mexico CFDI is a good example. So instead of printing the invoice out of SAP, send it to the customer. You send a data stream to a governmental agency, they put the big, I mean, in China it would be , right? So it would be big red stamp on it and then you forward it, right? And the invoice that you have represented at the SAP system is simply not the one the customer expects, right? So if you want to send a copy, you go into that agency system and pull a copy from there, right? If you would have a rigid, okay, it needs to be like this or this or this, then you wouldn't be able to do this, right? But I think the most topic is, and that was for us, in the rollout, it was more a capacity constraint that nobody had time to take care of all those kind of local things. And I remember, and that's anecdotical value. So when you work with SAP and you do partial payments, right? And that's, it's an, well, it's ASAP nerd topic right now.
- [Fred] Yeah, it is. Definitely.
- What you do is, SAP creates a new document with a new number and stores the old number in a reference field. But if the template is built in a way that it shows the document number, you show a number that the customer doesn't know, right? So you need to modify that. What we had is we had those instances and this is, it has a high electronical value. I said it costs five digit number because impact analysis and it takes several months to do it. Let's do it, let's not print it in SAP anymore. But key it in with word, right? And this for me was the point where I said, this is what you can do different and we put as much as possible where the impact in the application is not significant into the hands of the operational entities, right? Because it's instant for, if you can do changes and the change is instantly there, it's an extreme value for the businesses because they say, I want to send a Christmas letter, right? They go in and raise a campaign to send other Christmas letter, right? You wouldn't do that. I mean, that's what wouldn't be my first document that I would build in the headquarter, right? And by the way, make sure that you're paying your debt beforehand. Yeah.
- Any questions from the audience? Yeah. Fernando, please go to the microphone.
- [Fernando] So thank you very much, Nico, Asrin, for the presentation. It has been super inspirational. My question is, this has been really such a success for you, your organization. It is a lighthouse. So how can we replicate this at Siemens, maybe first of all, within your own organizations?
- Well, I think it's a matter of making sure that you find a product that scales, right? And finding those kind of things that make a difference for the users to drive adoption, right? I think there are ways while you, how you can use Pega which are not optimal, right? So for example, you pull emails or read emails in the system, give them a case number, but don't automate them afterwards, right? It's most probably from a cost benefit perspective, the benefit, okay, you have a little bit more transparency and tracing, but the benefit is not high enough in order to justify the case costs you have, for example, right? So I think it's really thinking about doing that. And then the second thing is drive that harmonization and convince entities. And this is very hard, right? And you need showcases, you need to start with, we have seen it in two different fashions. We've seen it, we make a decision, we roll out, we've been really fast. And then we have also operations like our healthcare sector where every CFO decides on his own, do I want to implement it or not, right? So when you need a compelling story for that, right? You sell internally, it's possible. And it's called, I'm not sure whether you're familiar with the law of diffusion of innovation, right? It starts and then you have a certain critical mass and then it just rolls, right? And this is the point where you need to come through. And this is especially when you're not mandated and as an application, it's a challenge, right? But it's possible to do so.
- Asrin, Nico, I thank you very much for this presentation, and I wish you a very good rest for bigger world in the last stay. Thank you very much for your participation, and well, the next step is to have lunch.