Deutsche Post AG (DHL) Earnings Call Transcript & Summary

It's Friday 1,500 bond time, so it must be dizzy Friday time again. My name is Martin Ziegenbalg. I'm Head of IR here at Deutsche Post DHL Group, and I welcome you to our third edition of our little summer series trying to familiarize you with the various aspects of digitalization how it's finding its way into our day-to-day life in operations and different other aspects. We started 2 weeks ago, we looked at how ESG is supporting how digitalization is supporting our ESG efforts. We have a deep dive on automation last week. And today, it's all about e-commerce and how digitalization can help. You see here the name of colleagues who are joining us today and taking us through the overall topic and Constantin, you're going to sort of lay out the overall background. And then we're going to have very specific examples deep diving into it. So before I hand over to Constantin for your little key note. Just a reminder to you out there, there is this little text box on your screen for Q&A, so for [ Q ] basically. And please make use of that. It worked pretty well last time around. And without further ado, over to you, Constantin.

Thank you so much, Martin. Good afternoon or good morning or good evening depending on where you are tuned in. My name is Constantin Peters, and I look forward to take you into how we unlock e-commerce potential via our digitalization efforts. In my day job, I had a team of 50 digital experts in DHL customer solutions and innovations that has been built in 2018. That was when the group decided to double down on digitalization with regards to customer-facing touch points. And ever since, we have been laser-focused on delivering initiatives that achieve happy customers in digital interactions as well as supporting profitable growth for the group. As we do this at group level, my team is a key pillar in DHL Customer Solutions and innovations, which my phenomenal boss, Katja Busch will introduce more next week. I trust the 4 mega trends reflected in our corporate strategy are well known to you. Today, the focus is really on how our digitalization capacity is actually proven instrumental to the success in e-commerce of our customers. And this is true both for the B2B arena where brands take back control of the customer experience as well as in the B2B arena where digital sales gained traction quite significantly. The exceptional challenges of the pandemic led to a tremendous surge in e-commerce demand 3 years ago. And at demand is now obviously normalizing. At the same time, we observed with our customers that there's a whole new level of awareness about bottlenecks and complexities in e-commerce. DHL has long understood that the individual customer needs in e-commerce evolve as they grow, and they have to reflect the local market characteristics. So while every market develops its own pace, there are 3 commonalities in place. First, e-commerce typically is not just a one-to-one affair between a provider and a customer, us and one customer, but it's much rather an ecosystem play with many different parties needing to be coordinated and therefore, jointly successful or not. Second, e-commerce is structurally growing. That growth continues. And third, the logistics excellence has a huge impact on the overall success because it's part of the buying experience. A recent data point from Octane confirms again that 4 out of 5 consumers will not likely buy again from a brand after they had to accept a negative delivery experience. And this is why it's so important. As you take a look at where dhl.com touches -- DHL sorry, touches the e-commerce value chain. That happens all the way from inbound to returns. To give you just one example, I recently had the chance to visit one of our operations in India, where we run omnichannel warehousing for one of the world's leading sports brands. From one warehouse in Bangalore, every e-commerce move to business and shops is conducted as well as all the fulfillment to the direct-to-consumer activities for the world's largest population. When I was there, what was most impressive is that we already earmark a huge space for further expansion because the brand is going to launch their own web shop in India, and therefore, are confident to see a lot more growth. But let's take a look at how digitalization practical supports that offering. And as you can imagine, I can talk about this topic for hours. Now I was told to keep it brief today, and this is why I would really like to look at 3 specific examples. Let's start off with delivery. And let's imagine a situation where it's Monday, you order a product that you really want to have available and at your house next week. Maybe it's some camping equipment for an outing with a family, maybe it's a sound bar because you look forward to the newest season of a TV show that you look forward to watching. However, you also know that you will be traveling. That the next 2 days, you'll be out of town and then Thursday and Friday will be filled with other errands. So basically, the likelihood that you will be home when our driver rings at your doorbell is not looking too promising. If that really happens, that we pass by and can't deliver because you're not at home, it's the perfect nightmare situation. It's actually a lose-lose-lose situation for the merchant on the one hand. Because they don't have a happy customer they hope to get. You might return the product if you don't have it, you definitely need to adjust your plans for the weekend, and we need to take a second delivery attempt, which is adding load to our operation system. In order to prevent this, we introduced on-demand delivery. Which gives the recipient influence to manipulate the shipment flow right up to their doorstep. You can pick a time window that suits you on that day that you're at home, you can reschedule to one of those days that you're at home or you can accept that the shipment is delivered to your neighbor and they can take possession and hand it over at your convenience. On-demand delivery is supported with digital communication, and it's available via an app. It's available via the web. So it's really convenient. Second example, is the duty and tax calculation. So I'll take you for a second into the intricacies of cross-border e-commerce. Let's assume that there's a product that you want to buy from another country, either a very cool gadget that is only available in the U.S. or maybe a high-end sneaker that you found in a shop and really want to have, or a luxury product that you certainly want to buy right from the source because you're concerned about the counter-fight potential. What you then do is you go to a web shop and you easily pay with your credit card just like you would do with a domestic transaction. And as of then, you eagerly await the delivery of that good. Instead, though, you are hit with a message by the delivery company saying that unless you're willing to cough up, let's say, EUR 80 more for import sales tax, delivery won't take place. Well, that's not a very good experience. And this is why leading e-commerce merchants try to prevent that from happening with upfront visibility during the checkout, which is a very sensible step in that journey. Basically knowing the origin from which it's going to be fulfilled as well as the destination, you want the good shipped to and the specific item in real time, they can match the right HS code and anticipate the import sales load that the customer needs to bear. You can pay it right on in there. And with that, we create a win-win-win situation. Because if that doesn't happen and the customer does not expect that charge to appear, quite often shipments are denied and returned, and that is not helping anyone. The third example then relates to the sheer number of tracking requests that comes with e-commerce. On our website, dhl.com, we saw the number of tracking requests skyrocket between 2019 and today. It grew to the tune of 800%. And this is why it's really important for us that our customers and users can get meaningful shipment status no matter what DHL service they look like. We address this via the universal tracking API or short UTAPI. That UTAPI is capable of really harvesting that shipment status from any of the operation systems that we apply across the globe and then play that back to the customer who doesn't care which DHL service in detail was being picked for him, for her. Now I casually use the term API, and I probably should explain that a little in more detail. API stands for application programming interfaces, and in a nutshell, APIs are the hidden hero's in e-commerce. APIs allow the quick and seamless transfer of data in real time, and this is how they connect multiple players whenever needed and in an automated fashion. I will try to illustrate this with 3 specific use cases. Let's start off with a customer of ours. Let's say, they are waiting for an ocean freight shipment to replenish the warehousing stock in their e-commerce warehouse. It's really important because otherwise, they run out of stock. They can trigger such a shipment and track such a shipment in our state-of-the-art customer portal for forwarding, myDHLi. myDHLi was launched in 2019, and it serves up all kinds of relevant information from the pricing of a shipment to the location of a shipment to the carbon footprint of a shipment. As well as to the progress of time stamps with regards to customs, a very sensitive element of these ocean freight moves. All of that data is found and surfaced via APIs. Let's take a second example with what we call partners. Many brands take advantage of post purchase specialists to really keep their customers in an on-brand experience as long as possible. As far as we talk about the digital realm. So that means these post-purchase players, though invisible because it's all white labeled, they need to surface and have access to data like tracking for shipments possibly locations where returns can be handed in or even return codes, paperless returns, QR codes. And this is why we need to provide them to them, even though they are not our customers themselves. We do that by APIs. And the third example then is looking at all those recipients. I don't know whether you're aware, but we have a state-of-the-art mobile app, for instance, in the German market, also in the Dutch market. And on these touch points, individual consumers, millions of them can track their shipment status, can follow individualized information and all of that is made possible via APIs. But let's bring it back to the customer. Every specific customer will have their unique footprint, you can call it a fingerprint in a way. So a customer will very likely operate their web shops using a web shop software. You may have heard of Adobe Commerce or Shopify, just to give you 2 examples. Then they might use a purchase specialist like Nava or [ parcel led ] for the other players. And then particularly, if they are sizable, they will already have a warehouse management system in place like Manhattan or Blue Yonder, and the list of services continues, including the logistics provider, hopefully, DHL. In order to make this overall system more performing than before, we need to use APIs to blend in and then make sure that basically, all these operations run seamlessly. In summary, e-commerce is here to stay and will continue to be an accelerant of logistics growth. And second, our digitalization efforts proved instrumental to driving the success of our customers. My colleagues will show you a couple of more examples how we do this specifically, and I would very much like to thank you for your attention. And with this, I hand over to my colleague, Ralf, from e-commerce.

Thank you, Constantin. A very warm welcome from my side as well. Good morning, good afternoon, good evening. My name is Ralf Pankotsch, I'm heading an IT department within the global head office of e-commerce solutions. And before I head into the IT part, just give me a click -- quick run through our business. Just a very, very basic fact. So we operate in the B2B and B2C space. Our sweet spot probably somewhere between, let's say, 1 and 3 kilogram. So in the -- let's say, if we compare ourselves to express, we would rather ship the iPad cover and not the iPad as such. We have millions of shipments every day. We operate in -- mainly in the U.S., in Europe, in the Americas, in Europe and in Asia, which is primarily in India, where we have with -- blot out of big, big player in the market. And also some Southeast Asia with Thailand, Malaysia and within Australia. So you all know the facts. So we are operating at a roughly 6 billion revenue plus. You know that we are in the ballpark EUR 350 million of EBIT last year. So very, very solid financial figures and platform. And we grow. And in order to grow, digitalization is one of the key levers. So without digitalization, there is no package, no single shipments that we can operate globally. We integrate multiple partners worldwide. We have a lot of technology as a baseline. And we have a lot of experts in our teams. Just in ecommerce solution as such, we have a team of roughly 650 IT specialists, and we run in an average about 100 to 120 IT projects, specific IT projects every year. And for today, I had picked 4 of those, just as an example, and it can only be on the surface. So I will run through these 4. And if there are any detailed questions later on, we have a Q&A and please reach out to myself later on if there are more details needed. So with this, let's go to the first, let's say, I think, the most important topic that we have to cover. This is our data. I totally believe that only companies who understand their data and use their data in order to create products in order to support our customers will be or will survive in the future business environment. From a technical perspective, we are, let's say, more common -- as common as everyone else. So we built a huge data lake, where we have all our data consolidated technology-wise, it's the Microsoft Azure database. We use data bricks for the data processing. We have primarily Power BI, the Power BI suite as our reporting platform. But more on the advanced analytics, you might know the company data robot, where we do some sort of, let's say, detailed analytics and detailed data science. And then the last topic here, and that is something we currently explore is everything that has to do with generative AI. Everything that is ChatGPT or the Microsoft fabric that will be a -- in very, very short time being introduced as part of the core Microsoft products. I know that Thorsten will speak later, and he already had last week, touched base on, let's say, the AI hub that we call Gaia in our, let's say, overall environment, which then opens the door to multiple of these generative AI tools. We will use them in our environment as well. So a couple of dashboards and reports are examples to this one, what we do with this platform. So #1 here, customer carbon reporting, oh, that's very important for our customers to know what is their footprint. We also have a global quality reporting. So we always want to know where we're at. So what is our current quality? Are there any issues in operations? Do we see any big customers who are affected by this. So there are forecasting alerting platforms in place, also, let's say, reporting dashboards in place where we can in the aftermath analyze what went good and what went wrong. And of course, volume reporting and volume forecasting is also something that is part of that platform. That all runs globally. And again, data is the foundation for all we do. So let me get to the second one. That is the question, how do we find the right prices for our product? How do we make sure that our customer pay what they are supposed to pay? What's fair in the market and what's fair basically a fair for us as a company. So we just recently introduced a advanced pricing tool. So this is a software-as-a-service platform based on a company Open Pricer. It's an e-cloud based, so it's scalable or there are more customer -- when we have more customers more -- let's say, billing runs that we have to do that scan - can scale easily. It has a very, very intense integration in our core platforms, be it salesforce.com for all the sales activities and customer service activities on one hand side and, of course, on the SAP or let's say, our finance environment in general on the back end. Constantin just picked on the API, and this is not different -- any different here. There are APIs and micro services in the back end. So it's a very modern and scalable architecture here. Some of the core features, I said already, pricing needs to be fair. It needs to be fair for, one for the customer, and it needs to be fair for us as a company. So this based on machine learning based on also artificial intelligence, we make sure that this is in balance. We run profitability analysis. So that's also very important for us as a company to make sure that whatever we do has at least a margin that we are happy with, but also, again, on the other hand, not basically compromising our customers here. Customer segment analysis, that's also something that can very well be supported by artificial intelligence, and that's what we do. Sales KPI monitoring, that's the standard features here. The third example that I brought is more on the operations side. So we have -- as Constantin also mentioned, we have a state-of-the-art customer portals globally. So our customer can interact with us using a portal, they can integrate with us using APIs. They can also have run a number of self-services where -- that we basically provide to the customers where they can do autonomously get service, get a shipment, get a transportation managed and all of that. And also on the other side, when it comes to delivery, we have multiple delivery apps globally. In any bigger market, be it, for example, in India or in Europe, in the Netherlands, there are also state-of-the-art delivery apps in place. And especially the Netherlands, the Dutch guys. They, in the last year, received a lot of, let's say, trophies or recognition in the market by technical and by marketing, let's say, companies who basically check what's there. And they were very often being recognized as the best delivery app or customer-facing app in the market. Some core features. You can read the tech stack here yourself. So on the core features, one is outstanding, I think. So when you look at the upper picture on the upper right side, it's called the last drop off. That is our whats revert integration. So we have in the U.K. integrated with what reverts or whatever, if a customer has these, let's say, combination of 3 words used as his address. We reach out to what 3 words, get back the geo coding data, and then the delivery app can automatically assign the address and we can deliver the parcel. So there is a very, very funny and nice video available on YouTube. So if you want to check that out, it's called The Last Drop Off. Some other features that you see here, delivery time prediction, pickup ordering, shipment tracking, all of that run seamlessly with all the APIs Constantin was speaking of in the back end. And last, that's my favorite example for today. It is -- everything that we -- that I said -- talked before was more like supporting business with technology. Supporting business with digitalization. Here, we have an example that digitalization in itself creates a complete new product. And I have a next slide, a detailed run through how the process looks like. But in general, what we do in India with our Blue Dart, we transport the exchange of high-tech mobile devices from today in store to our delivery persons in the field. So we can deliver and exchange these mobile devices. We have already multiple brands integrated. I have some more details later. And I will walk through how it works in detail. So you see the technology stack here, so it's a FireEye delivery app. And this FireEye delivery app is integrated or integrates with back end from some of the big brands in the, let's say, mobile space. So it's integrated with iOS diagnostics. It's integrated with Android. So when it comes to companies, it's Apple, Samsung, it's Huawei and some other companies that we are integrated with. We have the payment is being done with Razorpay gateway so also integrated, so then all run seamlessly in the back. So how does this process work? What's basically so I think so great in this product. So if you imagine a customer who has an old phone, whatever it is, an Apple, Samsung, whatnot. And he wants to trade that into a new one. So many of these companies offer a trade-in service. So what you typically do, you walk into a store, put your phone on the desk and ask for code. So they check that. Technically, they look at the phone and so on and they give a quote and then you buy the new phone and the quote that you get will be deducted from the payment. So that's about it. And then you leave the store with a new phone. So here, we have pretty much the very, very same service. The customer buys the device, but now he buys it online and he asked for the trade-in option. The company provides a preliminary quote and says, your phone is worth this or that, let's say, USD 200. Then the company ships the new device. It shows up in the Blue Dart delivery center and customer service at that point in time, call the customer and say, hey, your phone is here. So it comes in tomorrow in the morning, be ready, have your old phone at hand in order to hand it over to the delivery person. So at the doorstep, our guy, our Blue Dart delivery guy does 2 things. So first, he does a visual check, and then he runs a diagnostic software on that device. This is automatically integrated with the back end of these big companies. And then it's the company either validates or gives a new quote an update to the quote that has been given before. And then at the doorstep, the customer can accept this offer or leave it if they say he accepts it. then our delivery person hands over the new phone. He takes back the old one. He handles the payment, and we in the back, handle the return of the old device. So with this, this was my last slide. I'm more than happy to answer detailed question, but I'm also more than happy to hand over to Nabil. He cannot be on site with us here today. But will join us from sunny Florida. So Nabil, over to you.

Thank you very much,Ralph. It's a pleasure to be here. My name is Nabil Malouli, and I lead our e-commerce and return solutions at DHL Supply chain globally. DHL supply chain is our unit -- business units that look after the contract logistics business. And I'm very happy to share with you a couple of updates on this topic because I think the topic of returns and recommerce is such an exciting topic. It's such an exciting topic because of the fact that you can really come by a great product development and new solution that is good for the planet as well as also helping our customers in difficult times towards making their business more profitable and more sustainable. So if we go into the market situation, what we see in the following slide is that the recommerce market has been booming. And consequently, the activities -- the logistic activities that are behind that are also increasing. So it's projecting that the recommerce market will be reaching 350 billion by 2027. And we see many companies that are joining that trend. We see companies -- not only the large ones like such a [indiscernible] see there around Walmart or Nike or Carhartt, Lululemon as an example, but you also see some of the medium-sized companies joining that trend and we start to see that becoming critical for organization and also creating a huge logistics opportunity behind which we estimate to be 24 billion of solutions that are going to be required on the logistics to be powering that need. If you're going to -- what it means for the different clients in the following slide is more opportunities for companies. It's quite interesting because as I mentioned, we have a direct impact in terms of creating solutions that are more sustainable to basically creating secularity across the value chain by receiving products processing them when a consumer has not been satisfied with the product or that the product has any type of defect and being able to reinsert that product maybe towards the different channels, but we basically recreate a second life to that product, which is obviously much better for the environment, but it's also much better for the company when we're talking about sustainability. We've seen more than 88 companies that have large resell program just in the last 12 months. this is not only increasing, but it's really taking place where everybody is joining this trend. What also an element that is critical is that we [ put overs ] the consumer of our customers at the center because we know that's critical for all the companies in order to retain customers and also to grow the business. And so returns are the directly link in terms of satisfaction because obviously, we all ask consumers when we send back something, we're expecting our refund, and that is something that's also triggered directly through that process. Another element is based on the current economic situation and the slowdown of e-commerce growth, where we see that many companies are trying to also optimize their business. And basically, through this process, you can really help them to increase margin and also avoid destroying products a -- through better solution and digital solutions. Another element that is critical is that this is an area of business that has not been prioritized historically for different reasons, but we see now that companies are still trying to serve this problem of how do you process return? How do you resell these products in the market and the supply chain behind is actually a critical enabler to make this viable. And last but not least, there is obviously a high cost entitled to that, to the entire process. I mean we can all -- I would say, see that when we send back a product, there is obviously shipping, there is processing and so on. And that is something that also will be key for companies in order to reduce these costs and make that more valuable. When you look at the entire reverse logistics, I matching a couple of the elements -- of elements and you look at the entire banking chain in the following slide, what you're going to see is that there's really like 4 pillars and that they are each one of them critical in order to make this process of reselling items that have been returned by ever not only from an operational perspective but also from the financial perspective. And that is starting from the retail management side, which is really the experience that we will all leave whenever we return a product through the return of that item, the shipping of the product back to a warehouse to a store, as an example, the process of receiving that item into a fulfillment center when we will receive items, we regrade them. We'll evaluate what is the best channel based on the different policies, the different product categories, the different margin levels and then reinsert that product through our e-commerce experience and through recommerce channel, that could be the website of the company. It could be an alternative website for secondhand products. It could be on marketplaces. The channels are quite diverse, and that is really depending on the type of companies we serve, retailers, brands and also some organizations preference around what is best for that product to basically be resold. When we look at the type of solutions that we are offering and how DHL is helping customers. We really want to focus today on 2 dimensions. One is on the software side, what are we doing to digitalize our processes and also to optimize them. And one of the elements that we are -- have started to deploy at scale and is already operating is what we call purpose build retail management systems which are basically software that have been a customized and that have been reenhanced to take into consideration all the elements of a retail product. For example, trustability and visibility is one concrete example of, well, when you return an item to a brand or a retailer, having that visibility or how many items. What are the reasons why the products are being returned? Where are they in their journey back to the warehouse or to the store is helping you to plan better the labor, you're going to need to process also if there has been a defect on the product and you might have to do a recall, so that trustability and visibility of the entire experience is key. The second element is a state-of-the-art return functionalities. And here, we are really talking about the functions and features that you have in a retail experience. And every clients and even for our clients that are multi-brand clients have different rules on how they will process a return. A company could receive a T-shirt and because it doesn't have the label, consider that as a product that needs to be sent into another process. And another one might tell you, well, you can just put back a label and that's okay and then we can resell it into our premium channel. If that's the only thing that is happening with that product on the time where we inspect the item. And so all these rules and all these decision trees are quite complex. They are product dependent. They are brand dependent. And so having the right type of technologies, to do that, really allow you to do this much faster, much precise. There's no precise in terms of process and accuracy and also you will have data and right we are just talking about data and give also data back to the clients to be able to help them to improve their business holistically. And last but not least, the data ecosystem that we are building also around integration with the different marketplaces, websites, platform in order to help clients to resell this product is key. And so we see proven benefit around the ability to manage large-scale volume increase of 30% of productivity through this type of software deployment and also the ability to help our customers sell more to better channel and reduce the amount of digital product. If we go into the second example of how we help customers. Here, we are more looking at the execution within the warehouse and how do we deploy robotic and automation to also digitalize the call process. I was talking a bit about the feature and functionalities on software. But here, we're talking also about all the type of [indiscernible] autonomous vehicles that we are using in order to transfer the goods from an inbound process into a packing station where we're going to be looking at the products receiving. The conveyor system that we are using in order to make that process for B2B experience as an example of our B2B return channel versus a B2C. And here again, we see that this type of automation and robotic solutions are helping us to significantly improve efficiency, reduced cost by 25%, but also shorten the time that it takes us to process these items, which is critical when we start to look at the overall benefit that we can provide, which is how do I make the products available to sell again and in a shelf in a warehouse or in a store, the fastest possible because that is reducing inventory carrying cost. And that is one of the major area that customers are looking to improve when they look at the entire return process. So to summarize, the benefit that we see here across how do we leverage technologies to help our customers and unlock their e-commerce channel, it's really about speed and efficiencies to shorten that time between the moment the product is sent back to warehouse and then we can process that within 24 hours from what we call dock to stock in order to shorten that carrying inventory costs. The second element is really this ability to operate at scale. We are uniquely positioned because we can actually really provide that entire value chain to this organization in one end-to-end, a supply chain solution. And last but not least, the ability to manage B2C returns and process that to the entire value chain to the recent channel and B2B within an omnichannel environment within an existing facility infrastructure really help to create synergies and improve the overall digital experience for the customers and consequently for their, our final consumers. So that is all my part. I'd be happy to answer question as well as we go into the Q&A session, and then I'm happy to hand it over now to my colleague, Thorsten who will speak about data science.

Yes. Thank you for the handover. I'm happy to be here. My name is Thorsten Kranz. I'm Head of Data Science implementation, and I'm looking after large-scale AI solutions, which we are developing and deploying to support our business. If you have tuned in last week already, you might have seen me already there talking about automation. Actually, you might also have seen me twice already due to my short cameo a couple of minutes ago, sorry for that. But last figure, I was officially here talking about automation and how AI already supports us in that area already now. I already teased that one of our solutions, which we are running, which is called product classification tool or short PCT. And today, we have a deep dive on exactly that topic. So Constantin in the beginning mentioned the megatrends, which we -- which are determining our business. He mentioned 2, which are covered today, e-commerce and digitalization. I can expand it even to 1/3 of those megatrends with PCT, which is the globalization. And this sets the frame for what we are trying to tackle with PCT and doing it quite successfully, actually. So our main purpose, connecting people, improving lives holds totally true for us, and we want to do this across border. In our global network, we are just constantly crossing borders with our services. And yes, you have to be aware that we are really living in a connected world, but there are barriers. And in the recent years, actually, some of them have appeared a new, for example, just think of the Brexit. Think of the VAT de minimis removal in the European Union or ICS2, just to mention a few of examples where new regulations came up which make it more difficult to do global trade. Whenever we go across borders, we have to take care of customs processes. And these processes are expensive. They have implied a lot of manual efforts in order to make things happen and to facilitate our shipments to cross these borders and arrive on time at our -- at the recipients. In order to support this, we have developed a best-in-class internal solution to solve a very specific problem, which is based on the product description, predict the so-called customs code or commodity code [ IHS ] code, you might have heard this before, which is the kind of standardized representation of each product that is intended to be shipped across borders. How we are doing this? I will share more details with you now. So we have implemented PCT already some years ago, and it supports us really on a variety of different use cases. Most important, probably, it helps us in the acceleration of processes. Steps in the process, which had to be done manually by experts in the field of customs can now be fully automatically done by an AI model based on historical data. In addition, we can support customers audit processes to improve the efficiency of our customers' processes and also master data quality can be improved for enriching product catalogs, for example, to enable fully landed cost calculators and full visibility transparency for recipients of e-commerce shipments. So with this extremely detailed architecture slide, I want to briefly explain to you what PCT does, how it works. You can imagine, due to our history, due to our network, we have a whole bunch, billions of examples of historical shipments, which have been successfully cleared at customs. Which have crossed the border and they consist -- contain beyond many other aspects, a product description and the correlating customs code, which is crucial for making customs processes efficient. We are leveraging all those data sources we have across the company. We are combining them into a machine learning model, leveraging late state-of-the-art deep learning technologies which then can be used to automate the process of predicting a customs code. This tool, PCT is then integrated via APIs into the operational systems of our business units enabling them to fully automating this crucial step of the customs clearance process and having major benefits for all our business units basically. So we have deployed PCT across the group in many, many of our operations. And we do not only support the global standard. Which is yes, 6-digit customs code. No need to go to details here. but we have even extended it to regional European Union, for example, or even local country-specific standards, all of them are supported here. We have the ability to automate really at scale due to the design of the systems in an extensible API approach, we are able to process millions of shipments every week. And the average response time of this model is around 50 milliseconds. So this is just an enormous speed, enormous potential for scalable automation. And we, of course, want to make sure that we are delivering best-in-class solution. There are external providers. You might be aware of the offer a similar service and we are constantly benchmarking against those external providers. We have a regular process of reviewing if we are still better than external providers. And yes, we are. We are clearly by far better. And yes, maybe not because we are so super smart, but we have a data asset. We have our core network producing continuously new data for the asset that is then being fed into the model. And this is the core why our solution is best in class and can provide the best accuracy at the highest automation rate. So I already mentioned, we have deployed it across our business. For example, for Express, we are supporting an automated classification of import shipments, also in exports for ICS2, this is needed, global or local standards. Altogether, we have implemented it for e-commerce solutions, also automation Post and Parcel Germany, especially for the import into Germany, it is being applied. And for DHL Global Forwarding, this is one of the cases where we apply it for customs audit processes and really create major benefits in that area. But it's not just for our core business, where we are supporting with PCT. There is another area which I would like to quickly mention. So maybe you have heard about GoTrade, the youngest member of our family of go initiatives, you might have heard of Gohelp or GoTeach, for example. Go Trade has the purpose to enable to support SMEs in emerging economies to do international trade to engage in the exciting opportunities of e-commerce. And there are a whole lot of activities where Gotrade is supporting in that area, help grow those SMEs those economies. And in the end, of course, this is a win-win situation for the companies, but of course, also for us, when they find us as a valuable partner in the actual delivery process of their e-commerce. We have provided PCT, which is an API. I explained it before, but we added an easily usable front-end, a graphical user interface to PCT which can be used by those SMEs who are part of GoTrade, very easily just by going into the browser and going to the specific website, which has been provided to them. So they just enter the product details in the sleek user interface. The AI provides the recommendations for the most likely HS codes that will fulfill the criteria of the product that is being put there, they are guided through the step, the multiple level of hierarchy that are there for customs codes. And in the end, they can select the final code without having to involve any brokers, any expert knowledge or whatever, which is a massive simplification of rather a specialized task that's usually an SME would not be able to cover themselves. So to summarize briefly, PCTE implies enormous benefits for our company and beyond, as shown on the last slide. So just to mention a few, we see significant reduction in costs we're talking about millions and millions of euros every year. We see increased customer satisfaction because if the clearance process is efficient, if it is fast, we don't have any delay in the delivery process due to customs. And we have additional benefits around CO2 footprint, yes, so even sustainability is represented here. And the way that we implemented it in a very modern way, AI model, machine learning model, encapsulated in a API makes it so easy. And is one of the reasons why we were able to scale it across the company, implement in multiple ecosystems across different business units and provide a reliable service and better service in the end to our customers. And I'm very happy to accept any questions on this. And this is now the transition to the Q&A session. I will move a bit over. And hand IT over to Martin again.

Exactly. Thank you, Thorsten. Thank you all three for these very interesting insights and how -- and these are only a few examples we could share with very, very interesting ones. We come to the Q&A. And I think we've got the first question that came in via Slide Andy Chu from Deutsche Bank asking probably something for you, Constantin to address. Do you have any quantification? How do you measure whether -- whatever you're doing there has an impact versus you doing nothing.

So I think if I read it correctly, Martin, and the question is specific to on-demand delivery. And it's very important to say that there is no standard answer available by market, it will look different. What we learned in on-demand delivery is that, let's say, the preferred options to be used. These differ by -- in the U.S., for instance, safe place delivery plays a key role. It's very attractive. In other countries, no sictuture is preferred. Overall, we can say that 10% to 15% receive instructions and that is a very good indication that there's a significant impact on the number of successful deliveries.

Yes. I think that's a very important point. I mean doing business in 220 countries means that you are acting in 20 different market environments with different consumer behavior experience and therefore, demand. So being flexible is key, get that. All right. There is another question here that came in anonymous with regards to cybersecurity. No, obviously, I could ask either one of you, but [indiscernible] how you deal with it?

That's a very broad question. So just some general notes. So #1 is by policies. So we have very strict IT security policies on data privacy policies in place following all the rules and regulations that are out there. That's #1. And any IT system any IT platform and any IT team has to follow those, let's say, guidelines, rules and regulations. So it's part of the build process already when it comes to providing new IT platforms. It's also part of the testing. And we are very often audited. And if any auditor comes in, that's the #1 question we see.

Good. Makes an awful lot of sense.

And maybe just to add to that, that's again one of the benefits of an API with an API stay in control of your data rather than handing it over. So if there's a reason to block access to the data flow, then you can actually manipulate that. If influence.

Okay. Very good. Another question coming in and probably something for, again, starting off with you, but generally, it's touching all. We had this in the earlier sessions already. What's your approach towards finding partnersn you prefer a long-standing partnership with providers, technology providers as valuable or you want to rather stay on your toes and see who's got the cutting-edge stuff.

I think this is a combination of both, Martin, because the entire arena is evolving so dynamically that you have to stay on your toes, and you have to stay in touch. At the same time, on some of the more complex endeavors, it actually proves helpful to work in a long-term relationship. The key thing is always on how can we possibly join forces to solve customer problems better than on our own.

All right. Well, there's one question over to Nabil, maybe. Interesting example on the delivery option and the example that you've described there. To what extent do you think that is scalable.

Well, I think the delivery options that we want to be providing are infinite. I think we -- if we look at just like 7, 8 years ago, the delivery process or the return process was a straightforward process to us was basically on from the warehouse to a consumer door. And now you see, for example, what Constantin was mentioning around on-demand delivery options, the ability to hold the products, to reroute to drop at neighbors, to bring it to your office, to hold for a certain period of time. And we see the same thing happening in returns, right, return being picked up at your home, pick up and drop of location. And so I think we will continue to evolve these models. And I think it's actually scalable. We see it's just not scalable in the same way every market in the same manner because at the end of the day, the consumers are the one driving the preferences and then you have to build the right technologies to serve that, but I think it's absolutely scalable in terms of whatever the best solution is in each market to deploy that, and we're doing this, and we continue to do that.

Martin, let me add a few words to that because question is particular towards the Blue Dart, which is in the integration with the big tax. So yes, absolutely. That's a fantastic question. I hear that very often. So it's scalable and what I mentioned, I would say. So #1 is adding more, let's say, companies who selling phones. So we have the big ones. We are testing a few others. So that's number one. Second is moving up the product. So not only phones but also tablets for example. Next in line, laptops. So the team in India is currently investigating whether they can be exchanged programs through laptops, that would open a complete new space in terms of companies that we work with outside in terms of volume that we work with. And last but not least, scalabilities in the geographies. That could be any other country within the tele-commerce solutions for one of our entities. And it also very well could be any other division. So yes, scalable absolutely, technically implemented or changes needed, of course, but possible.

All right. until we get maybe another question from the external side. Let me raise one issue, which I think is particularly valid here for the field of digitalization, supporting e-commerce because at the end of the day, this is something that also touches the operational life of the employees out there. And one thing we know about digitalization, it's for each organization that's confronted with it. change process, yes? So what's your proven approach to make sure that whatever you guys are developing and thinking of is being transferred into the operation into the organization, meeting as little resistance as possible. And are there specific training programs? Or how is your experience?

Maybe I can give it a start, especially for the field of AI and data science. This is a super complex topic and very hard to grasp for somebody who is not involved in this so far. And especially here, we have invested massively in the awareness and education of our overall workforce in this topic. We have established under our certified umbrella for training and e-learning where everybody basically with access to our learning infrastructure, tens of thousands of colleagues can do a short learning to understand the basic buzzwords, what it's about, why it's important? Why does data matter to our company? This is 1 building block. But then we also have a training program, 2 days training for our management team, where we are really going certain many levels deeper than in the short [ line ], understanding also with them what their role is in our transformation around data and all those aspects that they have to bring to the table to make it successful. And also for this, we have, for example, trained 100 trainers across the whole group and have already trained close to 2,000 managers across the whole world, basically. And we see really impact from this. It's not just about training, but about the impact it creates afterwards because our projects are more successful, and we are faster in the adoption process, which is core at the end.

Well, maybe to add, I mean, in my experience, quite often data beats opinion. The beautiful thing about digitalization is that you can track and you can pilot. And if it works, then you can scale. Now we have a huge reservoir of good ideas. If you listen, if I listen to my employees, they quite often have better ideas, so no ideas that didn't even cross my mind. So if you can then put them to the test and if then they work, it's actually fairly easy to scale.

Yes. So I mean when it comes to digitalization, nearly every single employee is somehow exposed to technology. If you look at delivery drivers, everyone has a handheld today. The route that they drive is very often predefined. And the good thing is what we just recently introduced is combining a, let's say, mathematical model how to optimize the tour of the driver together with the driver's preference. So we put both into one model. And that at the end of the day, we come to a kind of an optimized routing that's the driver accepts. So that also allows us or allows the driver to accept technology as part of their daily life. So that's what we see very often. Yes, training is a big factor here, and we invest a lot everywhere into that one. But all the next generation is anyway tech-savvy. So that's not a big problem, I would say.

Okay. Well, that's an important link how you get it into the organization and bringing benefits together and that sort of leads to another question that we got in here from Alexia at Barclays, Alexia, thanks for that. It's basically about -- yes, I mean we are doing -- what we're doing along the famous 3 bottom lines as a customer benefit is an employee benefit. And what do you think? Where do we stand within the industry? And where does the industry overall stand with regard to penetration, let me call it, penetration of digital solutions into the operating reality. Is it -- are we stretching the surface? Or are we always halfway through what's your thinking?

Yes, I can say some words to e-commerce solutions, of course, I think we are not cross stretching the surface. We are in that business already for like whatever, 20 years. The -- I think we are well advanced. When we look at the market and our competitors, we are, in some areas, well advanced. In others, yes, there is some room for improvement. But I would say just -- we just recently saw that with the introduction of generative AI, there is out of a sudden, a disruptive factor and we have to basically get back to the drawing boards and see what we can do with it. and how can we introduce these models into all of our, let's say, businesses into our delivery chain into our products. So I don't think that this technology, let's say, driven into companies will stop anytime soon.

Maybe to add there I think if you take a look at how you can purchase music or movies, that has totally changed. It comes to delivering e-commerce, that physical part will not go away. So for me, the key question is how can we infuse and connect these worlds to the success of our customers. And in that regard, I do not think that we will fully digitalize and we actually also observed that some of the players who basically do a lot of attention, fully digital, then develop some investment into the not so digital elements as well because without them, it doesn't work.

Right. Yes. I agree. I mean the goods we are shipping will also in future need potential. So we often get this example of earlier something like a digit freight solutions and say if your driver drops you off at the wrong corner, you are smart enough yourself to move across the street. Boxes don't do that, right?

Well, okay. Well, keeping time in mind and sticking to our promise not to engage you for more than 1 hour with our session today. I think we can conclude. Nabil, Ralf, Torsten and Constantin thank you very much for the insights that you have provided to us. And I think the debate that we had just had at the very end leads over very well to the concluding and fourth session in our Digi Friday series, which is next week's Friday. Same time, and it's on how digitalization supports and changes the customer experience. We're going to have a -- very good. And with that, thank you for your attention, your interest, your questions, and looking forward to see you next week. But before that, make sure you have a good weekend. Bye-bye.