Leadership Espresso with Stefan Götz
In a world of constant change and complexity, true leadership isn’t about control— it’s about curiosity, clarity, and courage. Join Executive Coach Stefan Götz - www.stefan-goetz.com, as he helps accomplished leaders unlock their authentic leadership compass, inspire peak team performance, and achieve sustainable success—without sacrificing life beyond work. Whether you’re a seasoned executive or an emerging leader, this podcast will equip you with the mindset, strategies, and tools to lead with confidence, purpose, and impact.
Leadership Espresso with Stefan Götz
The AI Revolution in Business Leadership: feat Patrick C. Freyer AI Strategy @BCG | GenAI Builder & Developer
Artificial intelligence isn't just changing what we do as business leaders – it's transforming who we can become when we leverage these tools strategically. In this provocative conversation with Patrick, a leading AI expert from BCG, we explore how professionals across industries can build their personal AI organization to dramatically increase their impact.
Most discussions about AI focus on efficiency, but Patrick reveals a deeper opportunity: the chance to reshape your entire role around what you do best. By thoughtfully distributing tasks between yourself and AI assistants, you can focus on the aspects of your work that truly energize you while automating the rest. This isn't about replacing yourself – it's about becoming an entrepreneur within your organization who builds and manages a team of digital collaborators.
The path to this transformation begins with structure. Since AI systems originate from the highly organized environments of software development, creating similar organization in your work makes these tools dramatically more effective. Simple practices like documenting your thought processes, maintaining structured folders, and recording your lessons learned create the foundation for AI to understand your needs and preferences over time.
Perhaps most revolutionary is how AI changes the distribution of responsibility within organizations. Patrick advocates moving from assigning KPIs to distributing genuine ownership of outcomes. When team members have stakes in results rather than just tasks, they're motivated to build AI systems that create true value. This shift requires leaders to become coaches who help their teams navigate expanded responsibilities and leverage new capabilities effectively.
For industries with high quality standards like German automotive manufacturing or management consulting, the challenge becomes balancing the speed of "80% solutions" with the need for excellence. Creating processes that support rapid iteration after initial AI output represents a new paradigm that many organizations are still learning to navigate.
Whether you're in marketing, R&D, or finance, the future belongs to those who can leverage abundant intelligence to pursue their goals faster. Start building your AI team today, and discover what's possible when you focus on what truly matters in your work.
Listen to the Leadership Espresso Podcast:
https://open.spotify.com/show/4OT3BYzDHMafETOMgFEor3
View the Leadership Espresso Podcast:
https://www.youtube.com/@Stefangoetz_Global_Leadership/videos
Connect with Stefan Götz on LinkedIn:
https://www.linkedin.com/in/stefangoetz/
Check out Stefan's Executive and Team Coaching
https://www.stefan-goetz.com/
Hi Patrick, it's great having you on the show, and today it's all about AI. You are one of the masterminds in AI at BCG, based in Denver, colorado. Happy to have you on the show.
Speaker 2:Thank you so much, Stefan.
Speaker 1:My pleasure to be here we are allowed today to talk about the future of AI actually, and what kind of impact that makes on business, on organizations and on leadership. Now, what do you believe in Is AI going to make?
Speaker 2:us faster or better? You see, that's a good question, and I mean to take a brief step back right. With AI, there's almost only future to talk about. Ai history is so far fairly limited, so we have an exciting time ahead of us and definitely few historic precedents, which makes it even more interesting to talk about it. And in terms of speed versus quality, right, I think we can, at the end of the day, probably get both.
Speaker 2:But the question is what you personally need and what you're trying to get out of it.
Speaker 2:And with AI, there is currently a focus on how do we broadly apply it, how do we bring it to organizations? And to me, the much more interesting discussion is how do we make it work for you as an individual, right, and what can you do to make it work for yourself? And then it comes down to what is your optimization function right and what can you do to make it work for yourself? And then it comes down to what is your optimization function right? Are you planning to be a lot faster at what you're doing, in which case it might just be a matter of building the right outsourcing functions with AI around you, or are you trying to become much better at what you're doing when you are perhaps in a highly quality focused role, right? I've spoken to doctors who now use AI tools to go much deeper into their assessments and patient history and so on, so both is possible. I think there's a lot of potential either way, and it comes down to what makes sense for you.
Speaker 1:Yeah, but now the question of what makes sense for me. You know you spoke about outsourcing. It feels like I'm outsourcing myself, myself. So what do you think?
Speaker 2:about this I? I think it's an opportunity to outsource all the things about yourself that you're not necessarily thrilled to do every day or of which you would like to do even more right? Um? So, naturally, coming from bG, something I spend a lot of time on is preparing presentations and preparing documents, and I wish I could outsource more of that so I can focus on the actual content, on interacting with the people I care about, on leading and working with my teams. Right, and AI gives me that unique opportunity to build an ecosystem for myself, almost like building my own little organization within my job description to allow me to shape what my everyday looks like.
Speaker 2:Right, in a way, you can take your job description, ensure you get everything done within that, perhaps even more beyond the job description, but you can now decide what you want to get done by yourself and what you want to push to one of your new AI colleagues, and it all comes down to creating an environment that makes it work for you. Right? That means having structure in place, having files in place. Yesterday, I had an interesting discussion with female leaders across leading German companies and they asked so what can I do day-to-day now to ensure that once the genius AI breakthrough comes in two years with universal general artificial intelligence or something like that. What can I do so that it'll be ready to push me to 100X? Right, and I genuinely believe it starts with the basics right Ensuring I have the structure in my life so that if a summer intern came in tomorrow and had to take over parts of what I'm doing, they would be ready to go, because, at the end of the day, that's what AI is going to face.
Speaker 1:Absolutely. Now let's be more explicit. Now let's take an example. You are a R&D leader in the company, you are a marketing leader in the company, You're a controller. So how does this apply your structuring to these maybe three functions?
Speaker 2:Yeah, great examples, Look. The first thing I believe people need to understand is that all AI is being produced by software developers and these frontier AI labs. And foremost, they develop systems that work for them right, Because that's the environment they know. They are used to working with text files in this very specific environment that software developers are in.
Speaker 1:We call them nerds. You know, we don't think they're somewhere on the Philippines or somewhere and they produce software. But how does that relate to me?
Speaker 2:Well, I think they're increasingly in SF and not in the Philippines.
Speaker 2:But, yeah, well, the question is how can you ensure that you create parallels between your structures and your work and the way these teams are set up?
Speaker 2:Because if something works for them, the question is how can I set up my workspace and what I do to make sure it also works for me?
Speaker 2:So one basic example I gave to those ladies yesterday is look, in software engineering, it's best practice that in every folder I have a little text document that describes what this folder is about and what documents you can find in this folder. If you have massive folders with many documents in your organization, it would be a great idea to add a little text document to each folder that says hey, this is what you can find in this folder, and it might be helpful to you if you're trying to do X, y, z right. So, really starting to add these little snippets of infrastructure. That again would make it easier for a summer intern to find their way around the project as well, because the same thing is going to work for an AI agent, and AI nowadays might be able to navigate your messy emails and folder structures even without these little helpers, but it's just a question of whether you wanted to spend the effort on navigating all your file structure or whether you want to put it in a place where it can help you get your best work done.
Speaker 1:Now this sounds a little bit incremental. I mean I would expect more for it. But let's be it Again. Take marketing leader, r&d leader, controller. Let's apply your wisdom to these functions.
Speaker 2:Yeah, so one thing I do myself and which I think, for example, in marketing, would be very helpful, is start to think about how can you put more of your thoughts and your ideas and vision and so on on paper.
Speaker 2:Because one problem you're going to have when working with any new coworker and this applies to AI just as much as it would to a human is they come in and you have all this history of things you've tried that didn't work, and experiments and ideas and graphic layouts in mind and so on, and you've never really sketched them out. Or maybe you sketched them out on a piece of paper, but it's lying somewhere at home and your coworker will never see it right. And so how can you put that on paper so that, on day one, your coworker which in the future, very well might be your own AI assistant right is able to find all these things right? So put your thoughts and learnings and so on on paper, take a note. So I know a lot of people who've been very successful at making AI work for themselves have had a great digital diary the last few years and they were able to read in that diary and AI knows them very well and can help them pursue their own personal goals.
Speaker 1:It sounds like creating, you know, putting my genius to paper and make it available to AI.
Speaker 2:To some extent right, at least your learnings. So maybe it's not a personal diary necessarily. Maybe it's something like a job diary. I already request all my team members to every day write down two things they've done and three things they've learned or would do differently next time. If you could do the same thing every day in your job, then that would be an amazing tool set for someone else, like an AI, to learn from right.
Speaker 1:So this sounds more I invest into the future. Actually, it's taken me probably more time. I'm already working 60 hours. Actually, I talked to a COO in the Chinese automotive market. He said, Stefan, if you can find me a way, a strategy, a leadership lesson that I can reduce from 100 to 60, that would be great. So you're asking me to do even more.
Speaker 2:Well, I will admit there might be a bit of upfront investment here and there. Well, I will admit there might be a bit of upfront investment here and there, but at the same time, I think if you start using AI today, there are very immediate time gains that should just make up for these sort of things, right, Whether it's scanning documents more quickly getting a quick overview of what happened in a meeting doing much more rapid online research drafting the for you, summarizing the emails for you, right, you know we are familiar with all these uh already.
Speaker 1:I think many work with co-pilot or chat, gpt or whatever. To summarize emails, right, or you know faster.
Speaker 2:But what is it?
Speaker 1:you claimed it's like uh, you become more an entrepreneur within the organization, so how could ai shape that and create a revolution?
Speaker 2:well. Look at the end of the day, it's about building your own team, and in the past, a well performing employee might have started staffing interns or building up people who can support them, who learn from them, and in the future, that will still be part of success, but it will also include building up a little AI org that works for you right. So, if you are that marketing leader, I think it would be a great investment to sit down on a Friday afternoon or Saturday morning or whatever and write down. These are all the things I currently spent my time on and I think, very quickly you will find things where it would be great if you could, you know, start handing them to someone else. And the great opportunity with AI is that it's incredible at learning from your notes and so on. So my encouragement to everyone would be just get started right.
Speaker 2:If you spend a lot of time reading and drafting emails, put Copilot or a custom gpt or something in place today, and it won't do the job perfect job perfectly yet, but you will keep giving it feedback and you will keep iterating on instructions and so on. It'll come better and better over time, and so, the same way as you would build up a human org where you don't expect everyone to come and be perfect on the first day. You have to go about building up an AI org that works for you, and what's really fascinating to me is that people with humans are entirely ready for onboarding someone new and giving them six months to develop, or something With AI, if it doesn't do the job perfectly on day two. You're like well, what am I paying for? Patrick?
Speaker 1:that's the normal case, everybody's patient you talked about kind of you know you check your case. Is it about, you know, being faster or higher quality? Now I think we're pretty clear about faster. This is what you elaborated. But it's much more interesting. What you are provocatively saying is like AI is going to give you 10 times scale or even more. So let's apply it again to marketing, to R&D. Maybe R&D would be a good, nice case now. So how could that tenfold or x-fold my quality or breakthroughs?
Speaker 2:yeah, and look the reason I've been uh trying to stay away from from your r&d example a little bit is that r&d functions are so different in my experience that they're incredibly hard to advise.
Speaker 1:I want to challenge you.
Speaker 2:That's great. I I appreciate it. So let me indulge in your example of R&D right, and I'm totally with you. The exciting thing about AI is that you can 10x or even 100x if you manage to build out the right freedom for yourself right, and if you can work with leadership to give you that freedom. And I do genuinely believe that one of the biggest opportunities now in companies is going to come from leaders giving their teams the infrastructure and the liberties to experiment and push the envelope much more than they would have done in the past absolutely yeah.
Speaker 2:So how can, how can I ensure that my team feels comfortable using these new AI tools and experimenting with them and trying something out and sending it to me, even if it might only be an 80% version? Because the exciting thing with AI is that we're now extremely good at creating 80% versions very quickly and most orgs just aren't accommodating of 80% output. And I think this is especially something I felt when working with German clients and German teams.
Speaker 1:Right, you mean Chinese? Go with 75%.
Speaker 2:I think Chinese might be more similar to Germans than many other countries. I think Chinese might be more similar to Germans than many other countries. But so you know, how can we create an environment where an 80% starting point is acceptable and even encouraged, so that we can move much faster and the exciting thing, right? Let me just briefly finish on this 100x point. Right, the 100x speed will often come at some trade-off, and what's fascinating to me is that there's this dual function of a. Can we create an environment where we start accepting these trade-offs? Right, because we are no longer bogged down on one very specific metric. And b where do we reach the tipping point where a hundred x speed gain and or quality gain in one area makes up for these small disadvantages in other areas? Right, and I'm just going to use one more consulting example, because that's that's where I spend a lot of my time at the end of the day.
Speaker 2:Right, powerpoint, slides and presentations. I'm a big advocate that there is going to be a big leap forward in presentations and I personally believe it will not be PowerPoint for various technical reasons, but naturally there will be trade-offs. Right, we now have AI tools that can start generating slides 100x faster than the best VCG or code, but you might not be able to change the color of the footnote as easily. And is that the tipping point right? Is it enough to be 100x faster overall but to have this limitation that you can't change the font color or something, or do we need to push further and enable some other feature? And I think you're going to find the same thing everywhere, whether it's R&D marketing.
Speaker 1:I mean, your example is a little bit hingy, as you know. Changing the color of the footnote may be important for some people, but overall doesn't give an extra value.
Speaker 2:That's a hot take, Stefan. Yeah, I know, I know extra value.
Speaker 1:But if that's a hot take, stefan, yeah, yeah, I know, I know I rest my case in in your case, but again, as I'm, you know, like kind of many of my clients on the automotive market and now we're facing in germany at vol, audi, porsche, mercedes, bmw, everywhere, we need to reduce our development cycles, we need to take more risks, because we can't squeeze in what we've done before faster, just faster, or even leave our test cycles and say, okay, a Chinese car doesn't need to run 150,000 because it would never run 150. It just goes. You know, it's a bill for infotainment, not for running driving. So there the you know the trade-off having an 80 solution with all the risks combined, how would ai work in that case?
Speaker 2:I mean, look, I'm not an automotive expert, um, but I've, uh, I've had the pleasure of being a passenger in some chinese taxis abroad, where the drivers claim that they're now heading towards 200 000 miles and they're very happy with their car.
Speaker 2:So I don't want to talk down their, their abilities and I you know.
Speaker 2:I think that it's around creating processes and structures that support this experimentation and 80% work right. So, for example, I think a reason that a lot of car startups have been very successful in recent years, despite often launching with an 80% product, is that they've been able to implement structures where they can iterate on their product past shipping date. So if you have a test slide, it's going to come with all sorts of software faults, if in doubt, in the very beginning when a new model is released, but they can keep releasing those updates that make the car better and they've created a service model where it's very convenient for users to have changes made to their car or improvements made to their car, and these sort of structures are probably what you'll need to accommodate the 80% and move fast and break things approach. That is becoming more prevalent with AI and maybe it's stacking different AI solutions on top In engineering. You now have all these engineers building AI solutions very fast and then having bots that go over it and keep iterating past the initial development date.
Speaker 1:I see that point and I want to go further. I want to go now. What impact would that have? Organ of an organization and on leadership? As if we apply AI system, we train them taking notes every Friday sitting, drafting and creating a digital library. This, of course, will make our work probably, over time, faster. As a, I will kind of bring out the best of all things. So you have a overtime of gain times of coming up with 80% ideas, results whatever. So the time gained looks like you can now apply to make things even better or come up with another solution, because part of your time is not wasted on repetitive or other coming up with something any longer. But how will we deal in organizations with that trade-off, with that risk, with that? You say liberty in leadership. You know now, if you allow that kind of 80%, you know, to a German automotive company, that is hilarious, that's insanity. So what do you see? How can we handle that kind? What does it take? What is your idea?
Speaker 2:So, look, almost everyone I've worked with in a role that has been faced with more AI support or interaction has shifted their focus and their activities from task execution to A task planning and structuring and B task reviewing, right.
Speaker 2:And so, as a leader, the question is how do you support a team and build the incentive structure where people are focused on structuring and reviewing, and that probably takes a different structure, a different reward system.
Speaker 2:I think you need to pass down ownership in many ways, because reviews are always going to be half-hearted if you're not the one responsible for the outcome. So you need to see your team members less as contributors and more as their own little business owners, right, maybe even P&L owners, if we go back to the marketing example. Really give everyone a piece of very specific but tangible ownership and responsibility that they can hold on to so that they can build their own little AI work and business around fulfilling that target and owning that little stake of the business. And so, as a leader, my biggest question would be how can I start chunking up my business, not into KPIs and abstract metrics, but into pieces of ownership that I can assign to individuals and tell them do whatever you think is the best thing to do to get this not just done, but to own this and what would be pieces of ownership?
Speaker 2:like comparing to kpis yeah, and so I think, for example, if we go back to the marketing example, how can you branch out from where someone has an abstract KPI to optimize search engine performance to optimizing sales through search, things like that? And then, if I give them that piece of ownership sales through search I start to move them into a P&L position where they have profit and they can invest into AI tools or whatever else they need. But they now start to own their own little P&L.
Speaker 1:So you would chunk them up, you would come with the overall target, rather than give them, you know, pieces, elements yeah, and and I think it's it's about giving them pieces of a bigger target, giving them stakes in that target.
Speaker 2:Because, as we established, right, making this ai org work for you takes investment, and people can only make good investments if they have some sort of return function as well. I mean, um, so I I hate it when I see organizations where people are given specific budgets for ai investments or something that doesn't make sense. Right, the goal is here's what you need to do and this is how much money it's going to make for the organization or whatever your metric is, and then, based on that, you can invest, and obviously you shouldn't invest more than whatever you think you can bring in with your improvements now, patrick, I'm gonna put you down on this one.
Speaker 1:Um, how does ai work for bcg? If you're allowed to talk about it, but maybe you can talk about how does it change your life within BCG and your teamwork?
Speaker 2:so we get a sense and a feel from real daily practice well, look, I'm very lucky and proud to be at the forefront of much of the AI work we do, so full disclaimer. My experience isn't yet representative for the broader company or let alone industry, but for myself it has totally upended my work. Yesterday I held a keynote which was entirely ai generated. Instead of creating the slide deck page by page, I sat down for 20 minutes. I wrote down everything I wanted to say and the points I wanted to make in a presentation, and then I handed it over to an AI tool and said now build the slides with these talking points, and I did zero iterations. I took the slides and went to present. This is not yet your everyday BCG and it's going to be a while. Maybe it will never happen for many of our projects, right?
Speaker 1:MDPs don't listen.
Speaker 2:Right, but this is the case study for myself as to what it could be like, at least for very specific types of work and for your consulting work as a team leader.
Speaker 1:how does the daily setup look different now with the pieces of ownership and so on?
Speaker 2:Indeed this shift that I mentioned from execution to structuring and review. This has already happened in many of our teams, I dare say I think for a long time, the biggest challenge for project leaders was all of a sudden, my team is generating things and having AI do research. How do I get them to review the outcomes resulted in the follow-on responsibilities of, for example, presenting that work or working with the team to implement that work that you come up with being passed down to more junior team members. So that's something I'm a big fan of and I think it's an exact example of this ownership shift. Right, because as a project leader, it's becoming harder for me to review some of the work the team is doing. So I tell them look, guys, I you know I'm happy to give advice and provide the overall structure and so on, but if you're doing a lot of ai work here, then you need to be comfortable and present this and and I think it's your role entirely.
Speaker 1:Uh, you know, first you give ownership, you give to liberty, you need to give space and you need to become a, a real great um, a person who is capable of finding the next question, of finding it. Could this be true of reviewing, of asking good questions to your juniors? So how does that work?
Speaker 2:Absolutely. And look, I think it would be fair to say everyone has to become a bit more of a coach.
Speaker 1:Yeah, I love this one. I train coaches.
Speaker 2:There you go. And so yeah, absolutely. I personally believe that there is an exciting opportunity to shift responsibility downwards and to have people own things earlier and more thoroughly. And then your responsibility as a leader starts to become A how do I help them get used to this responsibility that people face early on? And b you know how do I act if things go wrong? Because things will go wrong. Right, we talked about the 80 approach. Ecg is not a fan of the 80 approach, as you can, as you can imagine, right we're definitely a 100% company.
Speaker 2:That is our value proposition, and so if we are a 100% company and we have tools that are being used by people, whether we want or not an associate is going to use ShedGPT, there's no point trying to stop them, and we realized that early on, luckily and if we have people in our org that are going to use tools that produce 80% outputs, but our target is to have 100% service, then how do we distribute the responsibility and the tasks needed to take the 80% that goes in somewhere and convert it into the 100% that we need to bring to our clients?
Speaker 1:Absolutely, so you gain time. Obviously you know people are faster in producing first time results 80% results. So you have some time to review, to challenge, to whatever. So how does work? How is it in your daily work Like how?
Speaker 2:You know it's a lot about deep reviews, discussing, digging deeper. I think it's an opportunity to dive deeper into topics than you ever would have before. That's entirely true in consulting and outside consulting and ensuring that you become an upscaling a quality upscaling function right and that you again build structures that work for you.
Speaker 1:Okay, let's check out what I learned about it.
Speaker 1:We have new opportunities.
Speaker 1:We have the choice between or actually both maybe to become faster and to have a leverage for more quality.
Speaker 1:But our role as leaders is going to change and it's going to help the company actually become not just faster with 80% solutions, but using the time gained to build up a digital library that makes you better over time, makes your team better, and your role is more of giving space and of allowing your team members to dwell deeper, to ask more provocative questions, even to turn things upside down, and this is going to create value, as you mentioned, because in the end not just in consulting, but I come back to my R&D If I can come up with something that drives mobility in a smarter way, this is value creation. So I see what I learned is it's a coexistence between AI and myself as a leader, but also my juniors. What I learned is not about first an 80% version and be creative with it and not judge it, not blame it, but allow the team the time to take ownership, give them greater pieces of ownership that will create value already and not just pieces of KPIs. Did I miss anything?
Speaker 2:I think those are all great points. The last one I would love to add is my favorite talking point about leverage, and I'm sure you're waiting for it already, because I can never end an AI conversation without mentioning leverage. To me, it's about taking who you are, who you want to be, and leveraging all the resources you have to get to your own goals and ideals faster. In the past, you would have perhaps raised money to get the job done, hire more people, buy a more expensive caterpillar or whatever it is. Today, you have a wealth of abundant intelligence and it is up to you to leverage that intelligence and build not on capital, but build on smarts, ideas and the very fabric of innovation, if you will padre.
Speaker 1:it's uh very inspiring for today and I take we have to talk maybe in one year's time and check out.
Speaker 2:At the rate of AI, we can talk every other month.
Speaker 1:We make a series. You know the AI exploding and I want to hear and see more examples, real daily examples, so people can figure out on how to apply that stuff to their own business. But for now, thank you for your time, thank you for your experience. Thank, you so much, stefan, and let's keep on rocking.
Speaker 2:My pleasure. Thanks for having me have a great day.
Speaker 1:Bye, bye.