DHL's GBS Rewired: Bots, Agents, and 6,000 People (Frank Schüler, DHL Global Forwarding)

I'm a big advocate of AI. I'm convinced that AI will change our work fundamentally going forward. Equally, I'm convinced that we urgently need to explore the new technology by applying it in real use cases, allowing ourselves to fail and to learn. Yet, I'm calling out to be more realistic on the use cases. We have to focus our attention more on how to manage those new tools. AI won't politely integrate into your GBS model. It will either transform it or expose its limits. Welcome to GBS Rewired with me, Sally Fletcher, had a thought leadership and community at Hypertus. This podcast is for GBS leaders who know incremental change isn't enough. In each episode, I sit down with the executives redefining how global business services operates in the age of AI, from operating model redesign to fully-agentic GBS. We challenge the big questions. Is your current GBS model built for automation or autonomy? What does an agentic GBS really look like when AI agents make decisions and not just process tasks? Where is AI delivering genuine enterprise value and where is it just hype? And are today's GBS leaders equipped for what's coming next? If you're ready to move beyond pilots and into real transformation, you're in the right place. GBS Rewired turning AI ambition into GBS reality. Welcome back to GBS Rewired with me, Sally Fletcher, had a thought leadership and community at Hypertus. Today's guest is Frank Schule, the managing director of Global Business Services at DHL Global Forwarding. Frank leads one of the world's largest and most sophisticated GBS organisations, with a team of around 6,000 people supporting operations across the globe and processing more than 180,000 service tickets every day. Under his leadership, DHL has built an impressive automation capability with a virtual delivery centre, operating around 160 bots that help streamline processes at scale. More recently, Frank has been at the forefront of exploring how artificial intelligence and the agentic AI can work alongside automation to transform GBS, quoting new opportunities while redefining how organisations manage a workforce made up of people, bots and AI agents, and we're going to dig into that a little bit today. Beyond technology, Frank has passionate about building high performing teams and a people-first culture. That commitment has been reflected in DHL's recognition as a great place to work in 124 countries, and it's continued recognition in the SSWN Research and Analytics, top 20 most of my GBS. So, they're definitely a great person to have on the podcast. Welcome to the podcast, Frank. Good morning, Sally, and thank you so much. Happy to be here. Yeah, yeah, absolutely. I know we've had, we've done a podcast before, so it's great to have you back. So, interview again. So, I'm going to dig into it straight away. I mean, as I said from the top, you process over 180,000 tickets daily and have 6,000 people in your GBS. That is quite a large number. How do you manage such a large global organisation? I can't even imagine. Well, I have to say, I feel really privileged to lead an incredible team of highly motivated and engaged employees. All of our people are really focused on delivering the service we commit to our business partners. And with that, I'm actually also extremely proud that we are not only receiving outstanding employee engagement scores from our employees, but as you were rightfully said, we are also being recognized from external great-place to work as being certified as a great place to work across all our six centres globally. So, I'm really happy about that and it is in the end all about our people. Yeah, absolutely. Yeah, we talk about AI a lot, but you know, with such a big scale, it's really important to have the leadership also, the managers in place that can inspire people, that can lead people, that can go in the right direction, that can think strategically, but also keep an eye on those transactions. So, that's great. So, let's look a little bit about the digital transformation, because obviously the theme of the podcast is all about how GBS is transforming in the age of AI. I know you have currently around 160 bots in your virtual delivery centre. Talk us through what they do and how that centre actually works. Yes, happy to do so. So, we have a centralized virtual delivery centre, actually based out of Chennai, from where we are actually supporting all our automation activities across the organization. So far, we have delivered various automations first of all over the last years, based on rule based automations, meaning RPA, supporting actually all kinds of functions, reconciling numbers, performing logistics tasks and so on. The impact of those, RPA based automations is of about 500 FTEs and efficiency releases actually today. Now, this has now been actually complemented more and more through AI and A-Birds process automation tools, like IDP tools or voice boards. These are now being delivering already 150 to 200 additional efficiency gains into our organization. That's really impressive. And I think it's also what struck me as how you're able to quantify the benefit, because quite often we speak to people and it's like, oh yeah, I put in bots, but they find it really hard to quantify the downstream benefit of the automation beyond just saying, oh, I've reduced cost per invoice from like $9 to $2 or whatever. Can I ask you how you like how can I calculate that? How you quantify the impact of your automation? Yes, sure. For me, this is absolutely crucial, because in the end, what is shared services about? shared services about actual work consolidation and driving in efficiency releases in actually how you perform the work. And therefore, you need to, when you design already a process automation, you need to think about what will you actually deliver in the end. So we have designed for that, several years back, a process which we call idea to robot, where we actually start in the initials in the idea generation phase to identify already what would be the realistic impact of the automation. Then we set that ourselves as a case, as a business case in the end. Then we configure the automation, while the monitor actually will be achieved that. And eventually, when we are live in production, we measure again whether actually this business case, this benefit case has been realized. Therefore, for all our automations, we have the exact measurements on what is actually the output of the change which we brought into the organization and how much work could we actually in particular, repetitive work could we release away from our workforce and therefore refocus the work of our workforce to more value-editing tasks. Yeah, that makes sense. Yeah, thank you for sharing. And you know, in the previous question, you touched a little bit on how you augmenting automation with IDP, for example. You know, have you been using AI and can you delve a little bit deeper into that? Well, yes, sure. So, IDP is already on AI, voice pods is also on AI and so on, right? So, I believe today you'll hear a little bit too much about AI, you know, without being actually specific on where do we apply artificial intelligence for which tools. And so, you see now for us, in the IDP case, we're on the AI supplied to actually have a template-free reading of any kind of page, so meaning OCR technology enabled by AI. We are applying that actually in large scale across various processes for us. Another good case is also voice pods. Voice pods, which are also enabled again by AI, which we're using now for cache collection and for other activities, so directly actually having AI calling our customer, you can say. Thirdly, we are also exploring and implementing now tools where we use AI then for reasoning to actually decide and make or prepare actually decision making, always then handing over then at the end to to human in the loop to make the final confirmation to proceed. Yeah, that's really interesting. Yeah, I think it's always better to have the human in the loop at the end. There's a lot of talk about governance at the moment, like who should be earning the process, who should be earning the work instructions. who should be making the final decision, whether it's you and in the loop, on the loop, off the loop. So, yes, interesting to see your perspective on that as well. Delving a little bit more into kind of the people's side, GBS organisations, as we said in Intro, are becoming increasingly blended with a mix of AI agents, bots and human employees. How do you manage a hybrid workforce? What does it mean for training and accountability? Yeah, I see. If I look back, it was actually about eight years back when we started our journey on RPA and our virtual delivery centre, that we came up with our concept of holistic workforce management. You see, if in my humble view, shared services is all about work consolidation in the end. And if you look back many years when actually shared services came up, it was all about actually bringing a lot of many employees together in low cost environments usually and put actually the work on many shoulders and then actually optimise how you can perform the work. Then eight, ten years back, RPA came up and with that rule-based process automation tools which complemented our human world forces. Therefore, you had first time actually two categories, as you could say, of resources, human resources and virtual resources, and you had the need to orchestrate actually the workload going across it. Nowadays, these resources are being complemented by non-deterministic automation tools based on AI or as they also quite often are called "agentic AI". So today, you have a few three categories of resources, human-based, deterministic or RPA-based process automation and certainly non-deterministic process automation or agentic AI. I strongly believe that in our environment of shared services in future, humans shall lead the work consolidation. Our bots, meaning our deterministic work home consolidation will perform the deterministic workload and our AI-based components will evaluate, will prepare a decision-making by actually leveraging those non-deterministic process automations. But you're absolutely right. This means also that our employees need to be upskilled. They need to be enabled to lead such a holistic workforce. They will need to understand the technology and they need to manage non-deterministic tools, tools which are not, as I would say, having a black or white outcome. They have an outcome which is more a shade of gray. But therefore, again, also in this non-deterministic world, it is very similar like a human being. So it is again going back to what we are doing already all those years, consolidating workload. I think we definitely have to be careful of, I think the term is rubber stamping. When the bot is correct, 90% of the time, not just always approving what the bots spouts out without looking at it because they definitely can get to that point where you're just like, "Yeah, it's fine, it's fine, it's fine." And then start kind of just approving on autopilot. How long do you think it's going to take to get to this kind of new GBS? Maybe five years, 10 years, longer? So it's different for every GBS. I think it will be a journey. And I see already starting happening. So for us, we are actually all our process automation being deterministic or non-deterministic, are getting into our entire holistic workforce management already as of today. So meaning our Solar Smellis Management will actually assigns the task to those three different categories of workforce. Now, over the years that will become more and more sophisticated. Over the years, we will be built also more confidence on the non-deterministic tours. And that will allow us again to actually give them more autonomous environments. Yeah, so it will change over the time. The impact I strongly believe will be big, but it will not go overnight either. So it will be a journey on which has started already. Yeah, absolutely. So you went to Shed7. That was a week. Lisbon run by SSO N. And I know that you spoke about the business case for Agentec AI. Can you tell us a little bit more about your perspective here? Yeah. So what I'm observing is a bit of a height of AI. I'm observing that the objective of process automation often seems to become secondary, while it becomes a primary focus to apply a AI, whatever it takes as I would call it. You see, I'm a big advocate of AI since I have written my PhD on AI about 30 years ago. We have to focus our attention on how we will control the output of these non-deterministic tools. How do we manage this shade of gray coming out? We have to focus on the increasing costs for applying AI. The token costs will become a major issue in concern once we scale the solutions globally across many use cases. While working on process automation, we have to refocus again our attention on the process first and utmost. In my experience, it's about 80% of any process automation can still be addressed via deterministic tools, cheaper and better to be controlled. It is only 20%, which would require a non-deterministic tool. Thus, my call out is let us focus our attention on a thorough review and design of the process which needs to be automated, rather than over-emphasizing the technology to be applied. Yeah, that makes perfect sense. I think we definitely see an increased focus and importance of the global process owner and having that end-to-end process. As you said, StreamLine improved and then eventually automated when it is ready to be. My last question before I let you go, Frank, is we talked in the intro about DHL has been certified as a great place to work in 124 countries, which is absolutely phenomenal. I know that the GBS is in about maybe six or seven of them. Can you give me maybe three brief reasons why you think this is? How do you get to be a great place to work? Yeah, my view in at least in our environment, as you're rightfully saying, we are based in actually in five countries with six centers, which we are having. My experience tells me you need to focus on three core elements. One is first of all, you need to provide trust into the organization and trust for the employees that you are able to lead in a future with a clear vision and transparency on what this would mean for each individual. Secondly, you need to engage with every individual across all levels, listen to various viewpoints and act on it, provide your employees a purpose of what they are doing. And certainly, I strongly believe we need to provide a home. We need to build a home through providing a sense of belonging and congrade ship for all employees. So those would be my three thoughts, actually. I would say. Yeah, now I agree. You're in your artwork a long time in the week. If you're full time, maybe 40 hours, probably more. And that is an incredibly large amount of time over the course of a year. If you're not enjoying it, if you're not feeling like it's got purpose, if it's not worth it in your mind, then it can be incredibly demoralising. So yeah, I completely agree with that. Well, thank you so much for joining me on the podcast, Frank. I really enjoyed our discussion. And I hope you will come back and join us again soon. Thank you so much, Celia. It was my pleasure. Thank you so much. Thanks for joining us on GBS WeWired. The future of Global Business Services is being shaped right now by leaders willing to question, experiment and evolve. If today's conversations about new ideas, share it with your team or a fellow GBS leader and keep the dialogue going inside your organisation. To stay connected be on the podcast, join the Hypertise community on LinkedIn or add me personally, Sally Fletcher and we will continue the conversation, share insights and spotlight practical examples of AI in action. Subscribe to the show so you don't miss an episode and joyous each fortnight as we explore what's possible when AI and GBS come together with clarity and purpose. GBS WeWired, turning AI ambition into GBS reality. [BLANK_AUDIO]