Podcast #29 Driving data success with people, process & platform at flaconi

Welcome to the Data Masterclass podcast where data leaders share their unique stories. In this episode today we talk with Agat Polar, who is VP Data at Flaconii. We deep dive on driving data success with people, process, and platform at Flaconii. Let's welcome your host Dr. Alexander Borich and Alexey's Plotnikovs. Hello leaders, this is Alex and Alexey's from Data Masterclass and we are here to build a better world through Data and AI and we think our organization, explore how organizations can be both effective and fulfilled, create fulfillment at the same time. We have a wonderful story with you to share today and somebody who really can potentially inspire you, we see her as a role model for doing exactly what we're here about, you know, to explore. It's Agat Polar, she's the VP Data at Flaconii. And for those who are not familiar with Flaconii, it's a beautiful, out-to-medium-sized company, Ecommerce, who really became a big player in the beauty business and brought it to Ecommerce. And it's quite a success story actually and it's also a wonderful team. Alexey's and I had a pleasure to work with you for a while now and it's a wonderful company with great culture. And at the same time Agata, you came from the business and moved towards that role and you were a commercial leader. And I think, you know, I think this is a great story because you're kind of shocking. A lot of data leaders, hey, this is how you can do it when you're a bit more courageous and simply do the right thing and do step by step. And to frame it a bit, you had to transform the team on multiple dimensions during and the organization around it. And you've done that at SPEED, you know, within one year, quite big changes that we have seen. And we will discuss a number of topics and for everyone who's listening, I want just to give you a brief point as to what we go and discuss so you can understand how this can be relevant for you because I think what Agata will show is both the mindset and the approach, but also kind of some tools and hacks and how to get there. So let's let me quickly just unpack the dimensions. So she looked at analytics and she transformed analytics. And she's in the middle of that journey. It's not done yet. It's never done, but she did big progress on that in terms of unifying the way of how to look at KPIs, building a semantic layer, exploring much better self-subs going from an at-talk reactive analytics organization to one that is proactive and has the destiny and their own hands. And I think that was a very powerful move. At the same time, she has an AI team and that actually has brought AI mindset also to the company and an intentable result within a small team, but big result of a small team, which is quite powerful. And you also migrated an entire platform at the same time and re-skilled your team. And while doing that, I think you shifted also a lot of your culture and mindset and inside your own team to show up at the best and how you work, how people show up and so on. I think that you see everyone who listening, you understand that's kind of three-dimensional at the same time or four-dimensions. It's the analytics transformation. It's the AI transformation. It's the organizational transformation. And at the same time, the architectural kind of platform migration and transformation. I think this is quite quite an achievement in one year or one and a half years. Maybe you can give us some more details to glares. Maybe in the beginning, the context I got to, at which moment came you in, what was the situation and where are you now? So we can then later unpack how you got there, but just so people understand that journey. Welcome, I got it. And so glad to have you here. Welcome. Thanks. Thanks. And like, thanks. I like, say, it's for the nice intro. Definitely. It's been a journey and a fantastic one and a fun one with the data team at Flakony. So basically context-wise, Flakony is highly successful. It's a medium-sized company in terms of number of employees, but it's definitely a German-isleading online retailer for beauty, which just had amazing revenues with over 28% revenue growth, profitable growth, just a fresh off the press because it's a Q3, you have two date numbers. And we launched markets. We have, we've been having an expansion focus since 2024 and we're launching more and more markets each year. We're now live in 12 countries. We're launching seven mores next year. And definitely that's adding to the context of why it's crucial to be efficient in your data organization and how you work with data as a business. And so my data team, when when I joined, due to many reoggs and leadership changes, they inherited from the disadvantages of having a central team at the central team at the same time. So both central and decentralized, but not yet federated. And as well as, of course, a big technical debt and some form of focus issue in maintaining healthy, healthy platform, healthy architecture. And so the symptoms of that when I took over the department was that our data delivery was delayed equivalent of one week per month, right? It's, it's, it's tremendous. Most reports were sharing conflicting information. That's also something that was definitely hurting. And we had much like siloed ways of working, as mentioned. Also, don't want to finger point why these are that. There have been many factors to this. But as a result, the business had changed, the data team had changed. No one knew each other. A lot of people in the business didn't even knew. We had data science and AI in the department. And yes, not to mention inconsistent ways of working and approach across domains scattered knowledge. So it was actually, yeah, a bit challenge for me. And I'm glad we've been through this journey also with my with my great leadership team. So you're coming to this situation. There is no trust. There is everything random, you know, no good deliverables. Where do you start? I'm like, how to start to clean the mess. Yes. So there is, there were two goals. We gave ourselves the first one was restore trust. Trusts not only between the team and the business. It was also trust in the data and trust in our platform. Like it's real reliability. And the second goal was to generate business impact. And that was also tricky because we needed the business impact to be both quick and long lasting. And as we all know in like us data leaders, these are things that are very tricky to balance. So yeah, of course, when I joined, I had to assess the situation. I hear about everyone's perspectives, also inside the data team, outside of the data team to get some form of good understanding of the status quo. But for sure, I couldn't make it some form of consensus on the way forward. Right. We needed to move fast. We needed results. And so I as leader needed to be very decisive. So we built the strategy. I built the strategy around this three dimensions. You mentioned Alex that relate also to your to your modules at data master class. But how I name them are the three piece, right? People platform process and aligned it off court with the board with the data leadership and kind of off we went right very top down doing some things that that might be simple at start to think of but very difficult to execute. And I think this is where the secret it's not it's not about something fancy it's about something you carry through through the end. How that was perceived by the team. You know, so you're coming you want sincerely to do the impacts. You name this restoration of the trust also with the team. So how did the start of the team and actually went for. Yes. So for sure. This was not like perceived very it's never popular amongst team, right? When we have a very top down approach someone coming from the outside that's never perceived positively. But something that actually was also needed I could see is to restore the human connection. So make sure like we understand we know each other. There were a few people in the team I knew from before and that was also like a great way to rebuild the trust basically something I also decided is to have some form of radical transparency, right? Because it's not that we we have like top down decisions always have a rational and people don't always ask for the rational but oftentimes it makes sense to them. So that's always why we've been decisive and top down that always explaining the why always being transparent about our actions and consistent in our actions as well. And as well as data driven I mean sounds obvious right? But when you look at health statistics when you look at the quantity of dashboard and how many of them are actually used stuff like that these are very convincing facts for a team. Why things needs to change. So on that journey I got in the beginning you know you kind of outlined what needs to be done where who you want to become and so on. I guess there was some in a resistance because it didn't match the identities of the team you know like it didn't match the identity of the people that they had. At which point in the journey you you made a kind of breakthrough where people started to move towards that new identity and what are maybe your reflection on what were kind of the key things that that that that triggered that I'll help to help to hear. - Yes, thanks for the question. It's something I've been reflecting on quite a lot, but I believe there were a couple of breakthrough moments. One was when the most experienced individual contributors in the team who really had the respect for the entire team because of their technical knowledge, because of everything they built so far started to kind of like publicly accept and praise my leadership approach. I think that definitely helped, especially the more junior ones, we were maybe like more resistant to change and more comfortable in the previous setup. So that was definitely a moment. And then of course I have to say when our leadership team was complete with Marta, Nico and Sasha, that was definitely also a very good moment because we started to be just many more people having a shared vision and also the competencies to actively implement everything in all the domains. So you started by yourself out of necessity and once your leadership team was completed, you activated that leadership team. Absolutely. Very good. So I think one of the things that our listeners would like to know is simply on these dimensions, where you know, where you told us where you started, but maybe on each diamond, could you just tell us where were you when you started and where did you got it to today, you know, and like to understand just the journey, you know, and of course it's you're still in the journey because we are on the journey and it's the wonderful thing about it. But I think you also made very good progress on this journey. So it will be great to see that kind of from A to B. Yes. So how we, like in those three dimensions, maybe if I structure it like that, we have on the people side a principle at Flakunis that we apply for everything we do, which it's called wecomers. And it's a core principle that says that true success can only be achieved through connections and relationships. And we it's actually with your team, with your partners, with your internal or external partners and your customers. And so that's basically how we started by having like also reconnections with our tech stack providers and getting people to, you know, in the office and talking to each other, having new roles being shared, like expectations. Real role expectations, right? Especially because before it was a bit of a blurry line, what's the analyst role? What's the engineer role? And we actually didn't just redistribute those hats clearly. We also help people get there. We invested in upskilling. We had some of the some of the engineers have scored the analysis, some of the analysts like share some context on their domain to the engineers. So it was really just something happened, completely behind the scenes, but it was not just putting an orchard on the slide and hopefully go, right? It was really an investing time and doing things together once and having projects where I was also very hands on and doing it together to kind of show it's possible, right? So that's something we definitely did on the on the people side with of course this focus on transparency, lots of like feedback surveys and collecting their feedback and suggestions along the way because it's not because you're top down that you cannot also consider people's perspectives and adapt your decisions based on how if you see whether it's working or not. So that's definitely like it's been a journey. We're still getting there. We got to a very, very good point from my perspective, but there are always room upstairs. And that's something we will just keep on doing. We invest more and more time on feedback. Yes, that's something related to the people and profile we have in the team. And then in parallel to that, we have the process side, right? So I think something that was also needing a big change in our approach was the way we do. We select the projects, we prioritize them and we carry them through the end. And so we revamped our road mapping process and road mapping cycles and how we also execute on projects so that it's ownership and impact driven by design. So we have a strong endorsement by our especially our CTO and CEO when it comes to which projects, where do the capacity go in terms of big projects? And as parallel to this, like, how do we build for the future and what's our road mapping, how is our road mapping helping with that? We improved our daily business support and how we manage incidents and how quickly we can react to things are still not to the target state and we have to still deal with that in the moment. So that's something we also did on the process side as well as when it comes to especially a domain. So we serve all the Flakuni domains and many of those domains work with similar data and they have to look at it the same way. Sometimes not for some reasons, but there are some North Star metrics that everyone should be should definitely be looking at the same way. And for that, we launched several KPI governance initiatives and we implemented it. So it started, it always starts as a project and then it becomes the process, right? And it's important that you always launch something, you launch initiatives, but you measure it and you're iterated. And because we have those road mapping cycles that each road mapping cycle, we have a look at where did what are we doing with that? How, how well are we performing and what needs to change? And that's actually how you get from this project to process phase and it's just stabilized and you can keep that with you for quite some time actually. And eventually it's you're not the middle man in the process anymore. And it people see the value of it and they take ownership and it falls on its own. And last, I would say platform platform is basically we decided to accelerate, accelerate our migration to, to, to Databricks. And what was actually needed to properly and pragmatically migrate is that we reflect on our entire data model and we tackle the technical depth. And that means because the business like this data platform we have right now was built, started to be built over seven years ago and the business has changed so much since, right? New business logics, new systems, even new like leading source for data creation. And this is something that required then not just the migration but the migration and the big refactoring. But for that again, like transparency to the business, looping them in on the decisions, making sure we also have this like restored communication with our tech stack providers, making sure the team gets the proper onboarding and upskilling to implement things fast on the platform and we get best practice. So that's basically if I like I was reflecting on it that these are the, these are the core things we've done that I believe are now putting us in a very healthy, a very healthy state. And yeah, we delivered quite a lot of new features along the way. So that's, that's what I considerate the success. It's really this balancing of immediate success and long term, long term wins. I'm super, super curious now because you spoke about this 3p. It's great story, you know, a lot of achievements. And it feels like for, I think for every data leader, it would feel like, okay, there's a clear inner focus, you know, on the team and the processes on the platform. But meanwhile, there is this business impact, right? So which kind of not waiting, you know, you have to take care of the team. So you still need to deliver something on the business. And I mean, I guess for you specifically, as you come from, you know, from the business background, from the commercial background. So what was happening meanwhile with the business impacts, right? Like was there something you can share as a common theme? Like, where are you trying to make this impact immediately deliverable? Why are you taking care of a team and other processes in other platforms? Yes, for sure. So one thing I can think of is like we, we had a very hard like focus on restoring our data delivery, SLA and our reporting consistency as quickly as possible. So this is something tangible, right? This is something where people start to say, oh, there are no conflicting numbers anymore. Is that random or is that, is that a bug or is that a feature kind of a question? It's a miracle. It's not a miracle. And so exactly that's one thing that we started kind of like really results driven and it implied some really manual work, right? Of going through core reports, identifying them, cleaning them, like this really this auditing and cleaning phase that I really time boxed because this can go like four years. But I think time boxing your, your fixed basic space is always a good, a good practice. That's helping like, yeah, let's just stay keeping the lights on. In a better way. Parallel to that, of course, thanks to this roadmap being cycle, there was always a discussion of how much of the of the capacity goes into what I call data, excellence topics, right? Operation, excellence topics that are really for the future and what do we do for the for the business right now? And we actually have quite a lot of new features, new systems being integrated, new logics, new new products. And each, each form and reporting on them and also seeing the adoption rising. And so that's actually a matter of deciding a bit top down, where does the capacity go and make conscious trade offs of like, okay, this needs to be done really, really now for our platform. But then we can of course support our internationalization. We can of course support the the big item on our products and engineering roadmap because it's, yeah, it's a matter of alignment. So of course the business is always a bit frustrated because we are a shared resource. We have to prioritize. We cannot get everything through, but the flip side of that is that everything we deliver is used as impact. And the team also benefits from it, right? Because of course, they see their work is not being deprecated or archived eventually. So it's a positive circle. Yes, yeah. And maybe we can, you know, I think it will be good to unpack a bit how you have approached. Let's take the metric management and getting from a talk to a more proactive, you know, self-service provisioning of analytics. I think that's a highly relevant topic. But I think I would like also, I would like to use it for two reasons. One is to simply show what you can do to improve metric management and, and, and, and, you know, self-suffer, move to better self-service. So, what did you do? Yes. So, the, the, the problem we're trying to solve is that we have actually, we have a great, in our legacy platform, we do have great capabilities because basically everyone accesses all data, all layers of the architecture from very raw to very process, I should say. And, we start that mean that everyone who had access to a bit of SQL knowledge and a tableau license could basically create reports and, and metrics. And basically, if you, if you want to create something as simple as revenue, starting from the raw data, you can look at it like, zillion of different ways, right? With, without VAT, like stuff that are very, very needy greedy in the details. And so, that's how, for coming from people in the data team, but also outside of the data team, they were always different perspectives on the numbers. And, what we've done is to say, okay, we need a list of, like, I mean, it's also, I'm not sure to well extend this is known for everyone, but doing it is actually the, the, the difficult part. We need a list of like, we actually had, I think it was 120 metrics on that list, the top 120 metrics that have to be, that are crucial for steering the business that cannot be wrong, that have to be stick what voice. And actually, we, someone from my team, Maria, prepared, prepared that, like, suggestion, and we reviewed it together with our board directly. So we had the, I think it was a two hour workshop with C, C, C, O, C, T, O, and we just challenged all of the metrics went over them. So very top down, but then very efficient, right? Because it was a, was though in one meaning that we knew, okay, this is how we want to measure this now, this, there is no blurry space. And because we did it with people like, like me who, who knew this KPIs through and through and also Maria or lead expert who also had quite a, quite some experience in, in their buildings, those KPIs in the different systems. We just, exactly, we could, we could create some very black, all white definitions of our KPIs. And that's how we rolled. And based on that list, we just gave us some, we, we, of we went and we looked at all the data products with highest downstream application or highest usage, depending whether it was like, reporting views, for example, but also based on the really like, yeah, how, how frequently was this data being pulled? And we, that's how we started from the top. And, and we started to fix everything. Top, like, from the, from the, yeah, the most used products, very copy-based teamwork, of course, it's, it needs to be refactored always everywhere. And that's the, that's the quick fix. That's how you get quickly to this, the state of every, the reporting speaks one truth, but it's not ideal because you still hasn't built, you still haven't built the proper pipeline in your platform to get to that KPI in one, in one and only way, basically for the external consumers who take this data. And we had to, like, that's why we did this exercise a bit in parallel, right? We did the customer-facing part, which was, like, not helping to reduce the technical debt, but fixing it operationally and parallel designing a data model and a concept where these things will just by design not happen anymore in the future. And I think it kind of shows how very nicely, how you balance always short term and long term, because I mean, you, we're talking a lot about team transformation, so on, which, which I mean, was really well done. But at the same time, you created all these quick ones and you always look, does it have direct impact? How can I do impact now, like, real tree now, like in the next six weeks, next 12 weeks, that's how you look at things. And so one thing that I noticed working with you is, and I haven't seen that much to be honest, like, and that, that kind of, I would say, determination is, let's focus on the next 90 days, let's focus on the next few weeks. What's the one thing we're going to deliver, she is, you know, and then, like, everything, and it needs to be super clear, this is in, this is out, this is like, what's the benefit and so on. And I think that creates a clarity and also the focus to get things going, to break down maybe a bigger vision into something that is substantial directly, you know, and where you create momentum. And trust. I think this is, this is, I had the same notion, exactly, is that it's probably also helped to rebuild that real trust, you know, because you're not talking about the vision, you're not talking about, like, yeah, I understand the problem, give us six months and we will deliver. But yeah, there will be more and better in six months, but we start from, you know, this week and in a couple of weeks, we already have something tangible, which people appreciate, you know, business or our team and so on. Yes. So we've been exactly, and this is something we've been quite good at for a majority of projects. There were still, of course, those like, black sheep's where we had, we didn't take that route and it actually didn't help restore the trust. I mean, always happens. But I think overall with like what, with that's giving me a good feeling because I, for example, when I joined the team, they were just out of having this dig cleaner, pexaricize to clean up the table reporting landscape because it was overloaded. And I was like, how come we get into a state where we have to do such a massive cleanup, right? How come the products are not continuously maintained for usability and for, for speed and for impact. And this is typically the type of thing. If you want to implement the way that it's actually a, like, an operating process, you still need to start with a cleanup. And basically the only thing you have to, like, usually people do it two ways, right? Either they send themselves a reminder, now it's time to clean. That's the best way or they just wait until problems happen. Now it's time to clean. But then what, what, like, the way I like to work is actually you maintain things. It's a production line, right? You maintain it all the time and then you don't have that type of problem anymore. And that's actually where the direction we're going. And I'm very happy about that because the distribution between the, this, like, build the production line and just, like, clean the old one. It's just shifting now. We started from a very reactive audit cleaning, like restoring phase. And now we're in the building phase more and more. And I think this is that we're also getting faster at building. And this is for me the, like, where I can be the most proud of my team, basically, because it's been a successful transformation with that regard. We also, like, restoring trust, right? Something I wanted to share because it's important. I have been such a pain in the ass to my team to tell them you have to be close to the business. And that also means being very responsive, like, just like on, you know, Slack meeting, make time for them. And everyone was like, yeah, but we're a service department. And I was like, no, no, no, no, it's completely different. It's not about pleasing them or doing, you know, stakeholder preferences. It's about what's doing right for the business, but still having a relationship so that when you communicate things are faster and efficient and put a lot of emphasis into, of course, we are a service department in the sense that we enable others to generate revenue to, like, whatever it is, be more efficient operationally. But we are, but I see the value of it as being we are proactive and anticipatory service department. And this is where we are now. And I'm very happy about that. So now also in the way projects are being like pitched in the beginning, it was, it's sitting this into the business, listen to what they need, prioritize what's most important and deliver that. And now more and more, the key to the analyst are in a continuous conversation with their business partners and they make suggestions. And we actually had this trimester like this one mapping cycle around some projects that were actually ideas, ideated from my team. And this is this, we are a service department, but look at this. This is how we, how we generate long term, this long term part of the, of the impact. And so I'm very, very happy and proud about that. Absolutely. I think you're becoming, you know, that real steer in all to the business in a good way, because you know, the data from the inside and you know, also what business needs. I have a question actually. So for further minutes, so far, it's a beautiful story of transformation with people, with processes, with business impacts. So everything seems to be perfect. I'm pretty sure there was a moment which weren't perfect. Right? And where? So I mean, from a learning perspective, like what are the things which kind of, you know, Finally, did not happen as you've. or we're unexpected those kind of down moments where you have to overcame something. You had to provide the leadership or a different thing because we see it was very complex. It was very multifaceted, transformational work. So surely there was this moment. And it's more from the learning for the hours as well. When the data leaders see this story, like, "Okay, that's just beautiful, I'm inspired." But where is my attention points, also like, what to watch where it may break or slow down unexpectedly? Yes. So for me, there was definitely a journey hasn't been a perfect point in the park. But the for sure, something I would definitely recommend the data leaders out there to do is to keep an eye for their senior leadership and making sure that they are on board with what they do at all times because that's always like helping because it's always a matter of like, conflicting priorities that are by nature, conflicting between what you have to do for now and for later. And for me, what was especially challenging, I mean, on top of, of course, this part of transforming the team who was very like used to very different ways of working, there was also that the business was super out of patience already, right? So in the beginning, we had to deliver to get to a state where it's not about people recognizing the efforts. It's about saying like, "I finally," you know, like, "Wow!" And that was really difficult because you need to keep people to stay motivated kind of as the recognition still needs and the trust still needs to take some more achievements. And I got some feedback. It's never enough. It's never enough or we'll never, like, you know, it's very one sided. And I, yeah, I could not say no to these things because I completely understand why they happen to be. And yeah, I just had to acknowledge and keep, like, keep on, like, sit tight somehow and keep on going. So that was like definitely difficult. And another thing for me, personally, I briefly mentioned before that some projects were not following that, like, iterative process of like quick deliveries and iterative deliveries. And definitely, I didn't pick the right ones on there, right? So when I joined, there were two big problems. One was definitely on the platform side. So this migration need, this refactoring need. And the other one was that we had a lot of scattered competencies in the company, governance, no governance processes around everything related to all tracking data, marketing data. And there the trust was like the quality of the data was not sufficient. The trust was minimal. The knowledge was very scattered. And some, some of the crucial people who knew about that in the business teams and in data were gone. And the few that stayed, had a lot of pressure and were kind of the process, right? The people that had the experience everywhere were clinging on to them. And this was, this wasn't what was difficult for sure. And what I've done, basically, I kick off a project to improve the marketing data quality. And I was staying a bit too far away from it for too long. And the project was not in that delivery moment. And that was like, I still regretted it to this day. And we changed the approach. But yes, it's definitely been something where the learning for me would be, if the reason why I didn't do it is that I'm not, like, I come from commercial operations. I'm very familiar with everything that's like assortment sales. Yeah, it thinks that attached to the physical side of e-commerce, let's just say. But all the digital marketing space was absolutely not my area. So I kind of also completely delegated, but too much from my perspective. And I think it's also for data leaders, they cannot be experts everywhere. But something I want to share to them is that it's okay because you will never, if you're the best in your team, if you're the expert everywhere, then you have a problem with the people you have in your team, basically. So, but it's okay to also, yeah, be hands-on in the projects where your expertise is not there. Because that's actually how you can ask the stupid questions, maybe, or ask, and make sure you see progress. And you understand what's going on. And this is something I, yeah, this was a call learning for me. So, stay, be very hands-on in a project that is critical when you don't have the expertise be hands-on because it's your role to coach even though you're not the expert. And I want to unpack this because I think this is one of your leadership superpowers, which made you this journey, which helped you to make so much progress. You're managing hands-on, and it's something I often hear during our workshops when you do think, you know, what does this mean? Because as a leader, you know, when you're two hands-on, you know, you're kind of over-stabbing your micromanaging, you're kind of taking over work, so you're taking away the room to grow. And that's definitely not what you're doing here. So, tell us, what's the right balance? What does, it seems like hands-on is for you, like what I sense is the middle ground between letting go completely, and kind of micromanaging. You're trying to find the right balance. And how are the approaches yet? Yes. It's a fine line between micromanaging and being hands-on for sure. The difference, I would still see, and why I do consider my approach as hands-on everywhere, because I'm also very curious, right? I want to understand. So, like, people who work with me, they know that I'm hands-on by nature. But the difference with micromanaging, micromanaging is when I won't leave any room for any other decision to be taken, that what I think should be taken. And I think this is the difference between when understanding what's going on, asking people what's their next step, understanding why they're doing this, and by when it should be ready, and why they made that decision, is something different from telling them what to do and kind of following up that they actually do it, you know what I mean? So, this is kind of the difference, the core difference for me. So, for like, in that case of that project, for example, concretely speaking, there were a couple of projects in that area, and I started to get hands-on with one, related to how we set up our new tracking for analytics on the platforms. And here it's about, yes, frequently having frequent discussions with the people who are the experts, bringing some structure, repeating what they say, you know, like it's a bit, I mean, feels like saying it's just coaching in the sense, but coaching very applied to the project, right? So, what are you doing? What's your next step? Why did you, like, where did you document it? Is it not documented? Why is it not documented? How about you create, how would you structure it if you were to document this? What are the steps? Explain it to me like I'm five kind of, and sometimes this is exactly what it takes to make people realize that actually, yes, maybe they were needing some help, you know, what it is. And they are perfectly capable, they are perfectly capable. It's just a trigger, I could say. So, it's a need to create a figure. And you use that, you know, to actually create results, and I, hello, actually, I'm learning from that. And I think one of the results, big results was that you move from this reactive mode to a more proactive role, in particular in analytics, you know, where it's usual that people throw, throw, requires over-defense and say, I need a red button, I need this thing here and this thing there. And I think you've done it in a very general way. So, it's not. And one of the things you, I think your chief to do is through you build this, a bit of a guiding coalition, you know, you got a big business team on board, I think it's the international team, and your company, you know, and you kind of joined forces, yeah, to do some transformative work that actually, you know, business transformative work, I would say here. Yes. So, exactly. What's something that I set on as a flakony data leader from the beginning is that I believe the true way for a business to be data-driven, if people don't need a square to access the information they need, and they don't need an analyst, and they don't need. So, this, why we, we like, I wouldn't say, especially for like, steering the business and understanding why did this go up or down or making decisions on the on the daily basis very operationally, a dashboard is not always the answer, but you need quick and efficient and reliable access to data, even if you are not a technical profile. Otherwise, you will always create a bottleneck on that technical profile, whether it sits in the central team, or whether it sits in your business team, right? That's also this like, when I was in the flakony of 2020, let's just say, they were a decentralized model where every team had their analysts, but they were actually the only reliable or knowledgeable, let's just say, person when it comes, when it came to technical things and TPCI and everything, and when that person was not there, the rest was a bit helpless. So, that's not a matter of where the people sits in the organization, and, and for sure, we have an international expansion focus, as you know, and these people, like the rule of optimizing something for the majority doesn't work with them, right? You're looking at the most choose reports that are global, doesn't work. They just have small volumes, but they need much, much faster growth than the other teams or the other functions, because they need to be data driven to come and convince, I don't know, maybe a pricing, this is a bit off for their country, and despite the small volumes, they should look into it. And so that's how, with that preliminary work, we've been doing on the KPIs and the definition of the prioritization, we decided to explore and launch MVP for a drug and drop self-service, which eventually maybe is AIBI or AI feature, could already be, but we decided to take it step by step. And actually, we had this week some form of training and exploring the tool, and it was super interesting to see how quickly, like the team needed this, right? And that's the thing, when you design it together with them, because of course, it's not like we designed something in our bubble, and then eventually said, yeah, you should use this window better than you. This is absolutely not that. It was rather hearing their pain points, being with them on a day-to-day basis, when they need those quick and fast support for very small, what actually should be answered with a very quick query or very, yeah, something that should not be complicated and taking long. And by observing that and asking them and working with them, and having them feedback the MVP, that's how we're gonna build, I think, something that's like that's beneficial, not just for team-international, but for all the departments. So, and here again, we cheated, right? Because this data is using the very much upper layer of our data model, right? Because it should be the most processed and the most golden, like the most golden form of layer that we should have that should be used for non-technical, by non-technical users, it's actually not the case, we kind of fake it until we'll make it and we will make it. And it's not fake in the sense that it's not true. It's just say it's not coming from the upper layers of our new tech stack, but it is in the new tech stack, and it's working and it's correct. And then that gives them some time to play with the solution, give us feedback and ask to build, yeah, the thing we don't see about it, basically. - So basically to move faster, and I hope everyone sees the pattern here. You know where you wanna go, but to move faster, you also make compromises, but very aware compromises to be able to move faster. And so what you've done here is basically to use the server layer directly for this, because the golden layer is country because of the migration still being built. And this way you solve a huge business problem and you enable a large, important commercial team, basically, and you also know what to build later, which is pretty cool. - And you need to, that's the part, right? You need to be consistent and deliver on it because otherwise you just add in depth. But I feel like I've been reducing depth since I'm here, but it's just so enormous. So sometimes I create a little depth on my own decision, but I know we'll get back to it eventually. I do hope though that the commercial team, the international team is having fun with the, with the tool now, and if not, then of course we'll be there and we'll iterate until it's eventually useful. But I could definitely see the quick adoption. I think we have some limitations in the setup we have now, but it's for sure getting the, getting the ball rolling. So that's what I like about it. - Good. - I have one question I got. And it's, you know, it's built on what you say so far, you know, in really driving this organization, driving, you know, data being in front of the, of the business and, you know, business leaders realizing much faster the value of the data. And my question comes, you know, from the discussions which we had as a part of our data masterclass engagement with Flakona. And you may remember I asked you about, what's your three-year vision? And specifically like where you see, you know, the data team and even yourself as a leader of a data team, you know, like how do you see this, you know, this going kind of three years? You know, I'll take a bit more strategic outlook. And you know, the reason I'm bringing this up here because when I heard your answer back then, I was surprised and encouraged and really, you know, very thinking through them, you know, like how do leaders drive this? You know, so maybe you can share a bit of this direction as all. Yes. So I mean, I have to admit, I don't know what I answered back then. But I hope I'm going to be, I'm going to do what I always do, which is being consistent and hope for the best. So my like division and basically where I want us as a data organization and as a company to be in three years, but I hope it's coming into as data-driven, data-powered decisions and processes across the entire organization, right? So data product helping us through and through be more efficient and helping the people in the business be more efficient. As well as of course, like, yeah, all the possible applications, but it's concretely going into being more efficient or generating new revenue. And how we've been, or this is, this hasn't changed. And where these hasn't changed since, I think, and I really, I'm confident this is what I told you, but I have to call for me as well. Where I see myself in that is really like eventually enjoying, in an enjoying seat because we don't have depth anymore. I have trust this will come and will keep on like, delivering data products with impact speed reliability scalability, which are also our principles or data standards, but where the team will be. So I think we now have a central team. The business is kind of like small enough on the human size to kind of allow it, but actually this is something that why I will stay true to this vision because this is actually all like, yeah, where we should be as an organization, we'll keep on iterating and adapting. And maybe in one day we have needs the rise that we need to decentralize again, but then we'll have a super healthy setup for a federated approach for like data, like data-driven data management. And this is something that, yeah, as long as we have this, we can just measure, adapt and iterate. So is that now you have to tell me? Is that what I told you a year ago? - You told that even more, you know, bolder. You basically that day you say that, you know, I see, you know, it's so beautifully federated and running that, you know, your work maybe four hours a week, you know? - Well, yeah, that's, yeah, no, that sounds like something. I did it more, that's my first thought. I was like, where I see myself enjoying myself working less because I'm not needed anymore. - Exactly. - Yes, I'm having a lot of fun right now. So for sure, eventually if I'm not as needed and if I can, like I would still be hands on, but if I can be hands on just for my own curiosity and knowledge, then yes, goal achieved. For sure. - You know, I think leaders like you are got will always be needed and you will just get even bigger task. - Mm, afterwards. - Yeah, it's kind of what happened so far, at least in my Flakony journey. So, I see how that he works. - Wonderful. - Yeah, I was thinking, you know, you've been, I think you've, you know, we've been working with it together. You've been also to one or two masterclass. I think you've been to DataMaskers. You're all, remember, you've been on the panel, you know, I was just wondering, you know, because we usually cover so many different things and topics. What did Sticco, was there anything that's ticked from these things where you said, hey, you know, this is something which I used in my work which I, where you started to see things differently. - Through participating in the masterclass for sure. So, I think something I have to think, to be thankful for, I think it's Kinder, her name, the governance coach. She gave such an insightful presentation on KPI governance and how to kind of have this like, yeah, structure the approach to it. But it was of course a strategic presentation. So I, I wanted to know like very detailed, the how a bit like you asked me, right, with the list of KPIs. I mean, that's not for me. It's really coming, coming from her. So I came at the break and I came and I approached her. And I was like, like, just tell me what, what tool did you use, right? And she said, we used conference. I'm like, what? Yes, of course, we used conference to start. Conference to start is always okay. Spreadsheet to start is always okay. It gets you to the point where you have some concept, you can show others and there was you on this. And then of course it becomes the meta data. It becomes a data catalog. It becomes and and this is exactly that. This balancing, yeah, of course, we will have a data marketplace. We will have a catalog, but you need to create that. And if you just create, you just create it with AI with no business understanding, you'll lose time. You just need the first very focused business understanding. And then you iterate and then you bring the technology. And that's when it becomes very interesting, I think. Well, I'm so glad and can I will be proud when she hears that. For sure. We will share it with her. And of course, I mean, on the, like, each time I can only advocate for this format, I think the participants and the people in the community of coaches, they, they, yeah, exactly they're very inspiring. And they've been, they represent very successful companies that are very different at the same time. So it's sort of stick to exchange with them always. So cool to hear actually. And what we say, you know, personally, I believe we are in a quite important point in time. You know, like, if you look at the acceleration rate, you know, like, you know, I was discussing this. I was just at a bar event in, in Vitzburg. and I was discussing that with some leaders. We said, "Look, for a long time, not much has happened. "Personal computers came and for a long time, not much happened." Then suddenly, the internet came and accelerated. Then mobile came and social and so on. And all of that was the past 15 years was pre-fast. Things go on. Comies like Flaconis got up and running because of that. And now we have AI. And the last few years of AI has been like mind-blowing. And we all know, right now, kind of, the future is shaped. What do you think leaders in our community, our ecosystem, could do either better or could simply invest and focus on communities like ours, like in the Dana-Mask class? What could we focus on to help leaders in shaping that future in a positive way, in a way that that is desirable, like I said, fulfilled and effective at the same time. That's why we want to have organizations. Yes. So AI speedboars are very rare. But for sure, something, I mean, it's funny because indeed, we were discussing this with Marta yesterday. AI is adopted faster than any other technology by humans. That kind of shows it's useful. I think we would just kind of show it's somewhat useful. So what you need to do as a data leader is to manage to funnel the energy in one direction that is aligned with your business school, I guess, if you do this for business. And exactly. So it's a matter of also encouraging, like, I honestly think, if you are not using AI tools kind of on a regular basis for your individual productivity, you don't realize the potential it can have on machines kind of, right? And especially for leaders, well, maybe a bit further away from deep technical work of their domain. That's where the change needs to happen. That's where data leaders should help them understand and get that angle into your business experience. You know your domain better than anyone. But maybe because you're not the one, I don't know, uploading this or doing that or checking that on a daily basis in your domain. And you're not using AI to do your emails, then maybe you don't see how quickly it's getting better. Because for me, for example, I'm using AI a lot for my individual productivity. And something I said maybe last year about how it cannot help me with that still today it can. And you can only, it's a empirical field and you can only see that if you try. And so as leaders, the best we can do in data is to support the transformation also from that other side. So of course you have to adopt the approach of thinking in terms of what are your big bets, thinking in terms of how you can generate results with AI so that you can showcase that and use it. I mean, we've been doing that very, very well in the data science team with Marta. And at the same time, on the week on our side, on the people's side of things, convinced them with those mini examples that are more relatable. Because I think otherwise it's just not relatable. And yeah, that's one thing I noticed, basically, because people are adopting AI, but they are not all leaders of big companies, right? They're just everywhere, all types of true files. And yeah, if they do it, it's just that it's helping. And that's the, I don't know if that's answering your question, but that's for me that's the one tip I would give. So I will try, and I'm trying to, you know, I'm trying to kind of sound back, but you know, what I think, what I heard in terms of the essence of what you said, which is super powerful. I think the key thing to create better organizations with AI is to empower leaders to become AI leaders, you know, who are today maybe business leaders, but not today I did this. Yes, it's a good summary. And I think exactly that. Just like for every other domain, just like in data, you cannot just, you know, use the eye. You have to use it for a purpose. You have to know why you're doing it. And for what goal? And you have to know what you're doing. I think I think you can draw a paradigm with vibe coding, right? I think the good vibe coders are actually people who understand what they're doing. And that's something, yeah, you can have the same if you just apply AI on the random business domain and you add a process that was wobbly. No, AI is not going to make that better. But if you show them that if they actually know how their process should be, yeah, I can help that in a much more efficient way. Or even think, honestly, I'm not the, I'm the very down to earth person, so I'm very, I find it challenging to find a like new possibilities with AI, but that's something that, for example, more creative profiles can do very, very well. It's just need, they just need to be confronted and experiment with you. And that's actually something we do. So I think Marta's currently planning some form of studying again. Marta is planning a workshop with our senior leaders when it comes to exactly a bit in a, yeah, through a different angle, looking at how AI can help their departments. And this is something that then will follow up on in the next year, it's part of our strategy. So doesn't mean we will stop doing e-commerce in the way we've done it. And it's just an empowerment somehow. Super cool. I think it's connects now also with the curiosity of mentioned several times in the beginning. It's a be these curious, you know, to promote that curiosity in the leaders so that they explore, you know, how to be an AI leader while being a business leader. Yes. And it's going a long way, right? I remember that time when we were having like an offline workshop with like Sticky Notes. And I actually think these are good to foster like team building and creativity and like doing something with analog. It's very good for building new solutions as well. And even if the solutions are technical, like digital. But the one thing I realized at the end is that people were like, as usual taking pictures and saying like, I'm going to digitalize this. And I was like, you know, you know, you can do that very, very quickly with AI, right? And they were like, oh, how? And then I showed it. And then they like, I received a sac message 12 hours later. I'll be like, well, you just changed my life because this person just was actually, you know, using doing workshop very organizing and moderating workshop in a very, very regular way. So it's something that just needs one person to show you. Yeah. Let's look into your crystal ball. You know, we almost at Christmas, the new year will come. You had an amazing 2025 of so many big changes. What's going to happen in 2026? So I hope my team is not listening. I want to be finished with that migration. What? Oh, we're going to time box it. We're going to, I mean, it's maybe not the thing of 2026, we actually said we already said targets for where we want to land by the end of 2026. But we want to, we want to be in the new world for sure. And that's something on top of, on top of, yeah, more AI for sure. That's something I also anticipate. We also grew the team. So that's something I'm very happy about. So yeah, that's these are the, the, the main things. But then honestly, I'm also looking for stability. Keep on doing what we do. Being constantly getting better at it. This is also something that I enjoy very much. And I also kind of predict for next year. Well, not knowing that you are hands on leader. Your prediction probably will come true. In some form. Wonderful. Alexis, any, any more thoughts on this. I think it's amazing. Great. It's a great point. So, so, so finished for today, you know, with exactly that the prediction will come true. So I love it. Wonderful. I got to you, you're such an amazing leader. And we are so grateful that you're sharing that with other leaders. Because I think a lot of all, you know, the people who work with and the people in our community and the wider ecosystem, they need stories like this, you know, because it's not only a story of inspiration, which it is. And it's also a story like it's like almost like a training session. It's a true masterclass. There's this podcast, you know, for people to understand, you know, how they can maybe learn maybe some of the leadership hacks and mindset you shared with us that can be so powerful and getting things done, getting things, you know, getting the vision into reality fast and and then iterate and then continue and so thank you so much for sharing these and for being so authentic and open. Thanks for the game words and thanks also for the, yeah, indeed, being confronted to the leadership style for now over a year and being adaptive and reactive to it. Yes, I still have something I also want to add, maybe as closing notes because we discussed it with Alex, and I found it was important to also share. Stay true to your vision. So look in the future, build for the future, and yeah, don't forget to also look back at where you started, and I'm very grateful for having done this exercise, because when I looked at the starting point, like inconsistent reporting, data delayed, every second or third day, this was really the starting point, right? And then we far away from that now. Still one day, I want to get to that, like self-service, self-service, or non-technical, everything AI-driven, but yeah, I mean, it's a journey, and it would not have been possible without the entire team, without the sponsorship of my leadership, without my peers in the business teams, who are also very patient and also very open, and yes, this has just been a great journey. - Wonderful, thank you so much. Thank you everyone for listening, and hope you take some inspiration home, and thank you, I got an Alexis, for being here with us and driving this conversation. - Thank you. This was amazing. - Thank you very much. - Amazing session. - Thank you so much. - Bye. - We have great luck, everyone, on your journey, and this amazing thing's going forward. - Data Masterclass, where great data journeys begin.