A New Colleague, Not PowerPoint 2.0: Annika in der Beek on Leading Statista's AI Transformation

Show notes

In this episode of the Leaders in Talent podcast, host Adriaan Kolff interviews Annika in der Beek, Chief People Officer at Statista. A psychologist who, in her words, somehow went business, Annika started her career in M&A in Hamburg and London, moved into HR consulting at Accenture, and joined About You in August 2020 at around 700 people, leaving after the IPO with the company at 1,600. Since early 2024 she has been Statista's first Chief People Officer, hired at the end of 2023 as the two founders stepped out and a new CEO took over to lead the transformation.

Statista is moving from a media company to a data and software business, with its data now available inside ChatGPT, Claude and Microsoft Copilot. Annika walks through what that does to the organization underneath: why silos made sense during hypergrowth and why you cannot miss the moment to turn them around, what she found on arrival (nine separate HR teams across 13 countries and no central HR beyond contracts and payroll), and why transformation does not end after the first restructuring. The second half is about AI as a new colleague rather than a more advanced version of PowerPoint: remodelling the content department sub-process by sub-process, a SharePoint HR assistant built in Langdock, and the performance coach agent she wrote herself for Statista's leaders, which prepares and challenges them but is deliberately not allowed to fill in the review or press submit. She also does the math on token costs, explains why Statista does not yet let agents into performance conversations, and admits she blocks hours every week to write agents and learn tools like Claude Cowork.

Timecodes

00:00 Cold open

01:32 Welcome, and Annika's own podcast, HR Ungeschminkt

02:56 What Statista does, and a psychologist who went business

05:12 1,000 people across seven countries

06:02 From media company to data and software business: why silos made sense

08:49 When data discovery moved from Google to ChatGPT and Claude

11:21 Breaking silos day to day: nine HR teams, one company

15:30 Transformation does not stop

16:25 Joining as the founders stepped out, and what she underestimated

18:33 Advice to her two-year-younger self

20:46 "We are not that company": the blueprint trap

23:56 AI at Statista: urgency you can see in the revenue numbers

25:44 Continuous learner: blocking calendar time to write agents

28:29 AI as a new colleague, not PowerPoint 2.0

29:32 Remodelling the content department, sub-process by sub-process

30:56 Agents are not free: doing the math on token costs

31:27 Statista data inside Claude, Copilot and ChatGPT

32:02 The SharePoint HR assistant in Langdock

33:39 The performance coach agent, and why AI never presses submit

35:33 Should an agent sit in on performance conversations?

38:14 The mobile phone clip, and never say never

40:22 The first AI-free retreat

41:44 Where to find Annika

___________________________

Connect with us on LinkedIn: https://www.linkedin.com/company/matchr/

Get in touch with us: https://matchr.io/contact

___________________________

Connect with Annika in der Beek: https://www.linkedin.com/in/annika-in-der-beek/

Connect with Adriaan Kolff: https://www.linkedin.com/in/adriaankolff/

___________________________

Resources mentioned in this episode:

HR ungeschminkt, der ehrliche Personal-Podcast (in German), hosted by Annika in der Beek and Anjuli Mauer: https://hr-ungeschminkt.podigee.io/

HR ungeschminkt on Spotify: https://open.spotify.com/show/6ZGgg4z9YmNeXJiiqxHIUw

___________________________

RSS feed: https://media.rss.com/leaders-in-talent/feed.xml

“

Searching for Global Talent? Expand your reach with Matchr's expertise, connecting you with top-tier talent from around the globe.

Transcription

[00:01:32] Adriaan: All right, ladies and gentlemen, and welcome to another episode of the Leaders in Talent podcast. Today I am joined by Annika in der Beek. Annika, welcome.

[00:01:43] Annika: Thank you so much for having me. Really looking forward to it.

[00:01:47] Adriaan: So Annika, before we started the actual recording, I realized you are my first guest that has an actual podcast microphone. And little did I know that I'm speaking to a podcast expert. Tell me a little bit more about that before we dive into your background.

[00:02:03] Annika: Okay. Well, pressure's on. I'm not a podcast expert, and I have to admit, whenever I'm a guest on a podcast, I am always still nervous. So practice doesn't make it vanish completely. But yes, for about two and a half years now, together with my colleague and friend Anjuli, I'm hosting a podcast, also an HR podcast.

[00:02:21] Annika: It's in German, unfortunately, so only interesting for the German listeners today. It's called "HR ungeschminkt, der ehrliche Personal-Podcast". Basically, HR Unfiltered, the Honest HR Podcast.

[00:02:33] Adriaan: So good. So that's why you have great sound and great lights, which I'm always very happy to see when guests come on.

[00:02:40] Annika: But we're too lazy to record with video. We always only do sound.

[00:02:46] Adriaan: Maybe this podcast is going to change that. Let's see. Hey, Annika, can you give our listeners a little bit of context about your background and your current role as the CHRO at Statista?

[00:02:56] Annika: Yeah, sure. I'm now in the role of CHRO at Statista for a bit more than two years already. Time is flying. And what is Statista? What are we doing? We are one of the leading global data platforms, providing access to different data reports that you might need for use cases in consulting, market research, and so on.

[00:03:19] Annika: I think the majority of my friend circle, whenever I tell them that I work at Statista, everyone was like: "Oh, wow, I know them from my master thesis, from my bachelor thesis, from my time in consulting," because everyone was pulling data from us to back their research.

[00:03:37] Annika: So this is one part of what we're doing that we're known for. Then we also have a ranking business. That means whenever you see, for example, a top employers ranking with companies like Forbes or Newsweek, or a most innovative companies ranking, they are most of the time done together with Statista, which is called Statista R.

[00:03:56] Annika: And then we also, of course, provide market research, so customized market research that we do through consulting projects for our clients. So everything is basically data, and also our mission: we empower people with data. This is what we do, and we will dive into that a little bit deeper later, because of course with ChatGPT and so on coming up, we are in the middle of a transformation, and this is not everything that we do anymore.

[00:04:23] Annika: That's so far about Statista. As I said, CHRO for more than two years, and time is flying, right in the middle of the transformation. Before I joined Statista: I am a psychologist who somehow went business. I started my career in mergers and acquisitions, where I spent two years in Hamburg and London, then switched to Accenture, where I started in HR consulting, and then in 2020 I exited to About You, an e-commerce tech company based in Hamburg as well.

[00:04:52] Annika: I joined them in the middle of the pandemic, in August 2020. When I started we were around 700 employees, and one and a half years later we were at around 1,600 and did an IPO. So it was crazy timing. And now I'm happy and busy at Statista.

[00:05:12] Adriaan: How big, just for context, how many employees does Statista have?

[00:05:15] Annika: We are now at around 1,000 employees and we're split across seven countries.

[00:05:21] Adriaan: And predominantly in Europe, or also across different continents?

[00:05:25] Annika: Across different continents. So basically, if we look at a follow-the-sun time zone approach, we sell our products 24/7. Our headquarters is in Hamburg, where the majority of our people sit. But then we have regional headquarters as well. For the Americas, our regional headquarters is in New York. Our second largest office, for the EMEA region, is London.

[00:05:49] Annika: And for APAC, our headquarters is Singapore, and we have different offices in Tokyo and in Japan.

[00:05:57] Adriaan: Great.

[00:05:57] Annika: In Tokyo and in Japan. No, in Tokyo and in India.

[00:06:02] Adriaan: Tokyo and India. You caught yourself there. You mentioned before that originally Statista was a media company, but now the company is transforming into a data and software business. Tell me a little bit more about the fundamental change that's happening within the company, and how you think about people, talent, and org design.

[00:06:23] Annika: That's a very good question. Let me start with the fundamental change. If I look at Statista from the outside perspective coming in, back then Statista was very much split into silos. To give you an example, there was one team that worked only on research data, so everything that you get from external parties, data that you acquire and then put together in a report, for example.

[00:06:50] Annika: Then there was a team for model data, so everything where you say, "Hey, we don't get the data from somewhere else, so we set up our own market model and create our own market estimations." And these were two separate teams bringing out two separate products, so pretty siloed from each other.

[00:07:10] Annika: And this is something you could also see throughout the whole organizational setup. Why did it make sense? If we look into the history of Statista, it has seen tremendous growth. It has always been a big success story with a strongly growing top line. If you look at that, it just looks amazing, and whenever there was a crisis in the market, where other companies faced a dip, for example during COVID, the demand for data just grew.

[00:07:44] Annika: So everything was high growth, hyperscale, success story, and I think keeping things in silos with the business model that we had back then was super beneficial. It sounds a little bit counterintuitive, but what does a silo do? It gives you the opportunity to innovate and to really put speed on some topics where you need to move fast, because there is less alignment.

[00:08:12] Annika: But you should not miss the moment to turn this around, because then there is the phase where you need more synergies to get things done. And I think this is exactly the case. So we are really thinking not only about how we do things, but also about how we push them out to the market, because clients want the products as we provide them to switch towards a different market environment with different data access layers.

[00:08:49] Annika: Speaking about access layers: if you're looking for data, some years ago you might have put that into Google, and with the marketing concept of Statista we were perfectly tagged, so the probability that we were ranking at the top of the non-paid Google results was pretty high.

[00:09:10] Adriaan: Ja.

[00:09:10] Annika: Now, if you put that into ChatGPT or Claude, for example, this is a completely different game and a different deal.

[00:09:17] Annika: That means we have to rethink how we can best group our data together, no matter from which source. Because in the end, when you're a client, what's your use case? You have a question and you need verified data to answer it. Whether it's our own modeled data, our own researched data, or from a third-party report, I would assume you generally don't care, as long as it's valid, as long as you can take it, and as long as it's a credible source that you're using.

[00:09:47] Annika: And this is a completely different way and journey of doing things. This also of course means really putting our teams together more to increase collaboration and to break down silos, number one. Number two, moving from a more media and publishing oriented company towards a software as a service company also requires different types of skill profiles and talent.

[00:10:19] Annika: For example, if you run a platform that you can download reports from, as we did, yes, that's a tech product. But how many platforms are out there that you can download products from? So I would say, yes, a tech product with strong traffic, but there are more sophisticated tech products out there.

[00:10:43] Annika: If we now look into the development of our business model, for example our partnerships with OpenAI and with Claude, where via an API key you have our data accessible in one of the leading LLMs in the world, of course this is completely different traffic that you're facing.

[00:11:04] Annika: This is a completely different business model and a completely different layer of complexity, because you're building a different product. And this was also new to the company, and we had to do a workforce transformation, basically.

[00:11:21] Adriaan: So tell me, let's zoom in on the practical aspect. This is the broader business context, but what does that mean from a CHRO perspective, how you look at this? Because breaking through silos, what does that mean in reality? How do you go about that transformation, hiring potentially new people, looking at your org charts?

[00:11:42] Adriaan: Maybe bring it a little bit more into the day-to-day, what that looks like for you.

[00:11:48] Annika: It's actually a mixture of a lot of topics. Number one is asking which skills do I need. What are the topics? What are our must-win battles? And what are the topics, skills and people that we need in the company to make this happen?

[00:12:03] Annika: On the one hand, that's hard skills that we need to look into a lot. But in a transformation phase, in a difficult market environment, it's a completely different deal which mindset you need to succeed in that environment than, for example, in a very successful, fun hyperscale environment.

[00:12:23] Annika: You really need to rebuild and transform things from scratch, and transformation can be messy and can be tough, especially if it's not linked to direct success out there, because the market is not always playing in your favor and there is a lot of pressure. So that's number one when it comes to talent.

[00:12:44] Annika: Do we have the right capable leaders from a skill perspective, from a mindset perspective, and from a resilience perspective? Because the last two years have been tough for us in the transformation. Then also, from the org chart perspective, how do we best collaborate together?

[00:13:06] Annika: Do we still have the right C-level units in place? Do we have the right VP-level units in place?

[00:13:14] Adriaan: Sorry, Annika, what do you mean with VP units in place?

[00:13:19] Annika: What I mean with a VP unit is: if I have a C-level unit, People, then a VP unit would be, for example, people operations or talent acquisition. So basically the teams below that. Are they still working together in the right way? To give you one example, if I work with silos, and this was the same for my role in the beginning, although the company back then was also over 1,000 people, nearly 1,400 across 13 countries, there was no central HR.

[00:13:54] Annika: There was central HR to do contracts and to do payroll, but what you had was different people in different divisions and different continents doing HR with the regional lead or with the divisional lead, who did not have an HR background. Which is not bad in general.

[00:14:16] Annika: However, when you come in and you just see, okay, there are some HR teams, but these HR teams are nine different ones, and they all have different performance approaches and different hiring approaches and different cultural approaches to how they do things. This is not one company. This operates like nine companies.

[00:14:37] Annika: And this is what you find in the day-to-day business if you speak about breaking down silos. You will see that across the whole organization, and this is really something where you have to see which topics are more beneficial to centralize, how much you centralize them, and which topics you deliberately leave within the divisions, because you say, "Hey, they still need to function individually because of speed, because they don't have a lot of collaboration touchpoints with other departments."

[00:15:08] Annika: So all these design decisions are something you take on a day-to-day basis with one goal in mind, but it basically does not stop. It's not that you sit down one week and say, "This is how the org chart looks, let's execute next week." This is always a step-by-step-by-step approach.

[00:15:30] Annika: And coming back to your first question, I think this is one thing that transformation really showed me. Transformation and change don't stop. It's a continuous process, whether we like it or not. I think humans in general, by our nature, are not the biggest fans of change and having everything different every single week. But we need to get through it, and we also should not lie to ourselves like, "Okay, we now do the transformation and then change will be over," because it's not true.

[00:16:05] Annika: If we do the transformation, or the first part of the transformation, three months later, with the current technological pace, everything can look different again, so you need to retransform. So this is basically the working mode. Being a big advocate for transformation is fun now, right?

[00:16:25] Adriaan: It's interesting. You joined Statista two years ago. Did you know coming in that your mandate was leading this transformational organizational change because of the different markets out there? Was that something that was already ongoing?

[00:16:41] Adriaan: And how did you decide what that org chart was going to look like? Sorry, many questions packed into one, but take me back to the moment that you arrived and where you are now on that journey.

[00:16:56] Annika: I have to say, of course I didn't have full transparency around what was waiting there. In general, because of one thing: the two founders of Statista stepped out by the end of 2023, and our new CEO took over to lead that transformation at the beginning of 2024. When they hired me was the end of 2023, so that means I still had my interviews with our investor, with the former founders, and with our new CEO.

[00:17:27] Annika: But of course, our CEO had also just started and could not tell me everything, but could share with me the first hypotheses that were going on. Secondly, I also think that, at least for myself, I completely underestimated the depth of how AI is changing how business models are operating and how we're doing work.

[00:17:50] Annika: That's the second layer where I was like, "No, I did not completely know what I was buying into." And how do you decide it on a day-by-day basis? Honestly, it's a lot about listening, knowing where you want to go, and working together with your colleagues, because in the end it's not me who decides, "Yeah, this is the org chart."

[00:18:14] Annika: It's me who decides how the org chart looks for my specific division, but not for the overarching org. So this is something where I have to work closely on a day-to-day basis together with our CEO and with our colleagues to see how to design the best org setup and how to best move forward.

[00:18:33] Adriaan: And now you're two years into this journey. What are some of the things, reflecting back, that you would advise your two-year-younger self on going on that journey?

[00:18:49] Annika: I would advise myself, first of all, to not work 24/7 for a month because you want everything to be done in no time and you want to set up your own department, because you will not be able to give from an empty cup. So that's one part. The other part, which was pretty tough for me, is of course you know that you're not the most liked person in the whole company if you're running HR, because you also have to stand up for a lot of tough decisions.

[00:19:21] Annika: If you're coming in as the first ever HR leader from outside in a transformation phase where it really gets messy and tough, because you have to make decisions like closing down offices, for example, because you change the approach to how you're selling and you do not need 13 offices anymore, but rather seven.

[00:19:41] Annika: Then of course the organization will not always like you, and it also will take longer for them to build trust in you as a person than if you came in during a hyper-growth phase, where everything that you do is hiring and seeing how you can staff up teams, with of course a lot of work, but way more positive and rewarding topics than these ones.

[00:20:07] Annika: I think these were the two topics that I would advise my younger self on. And then another learning: you will not be able to come with a blueprint and just apply what you did at other companies to the new organization. Of course it's good if you have a certain been-there-done-that mindset and you don't get shocked by certain topics.

[00:20:33] Annika: However, it's very important to always listen and to take your learnings, but always see that you only apply the things which really match the culture, the interest, and the maturity stage of the company.

[00:20:46] Adriaan: This is so interesting. I just had a call with the SVP of Talent of an AI data center hyperscaler, and he's had a tremendous career. He worked at the largest tech companies on the planet, always in head of TA roles. He's seen what these big companies look like.

[00:21:06] Adriaan: And he sat down with the CEO in his second week, and he was like, "All right, this is my background, and I'm going to do this, and this is what I learned from there, and we're going to do this, and we're going to do that." And he shared the story that the CEO stopped him in his tracks and said, "No, I do not want you to do that."

[00:21:23] Adriaan: And he was a little bit taken aback. He's like, "Well, I've learned from the best, right?" And the CEO said, "We are not that company. We are doing something very different. I actually want you to write your own canvas and your own playbook on how to look at this. And yes, of course you come with experience, but let that not cloud your judgment of what is needed here."

[00:21:44] Adriaan: And I thought it was such a profound story of coming in somewhere with your bag of knowledge and still having an open mind about what works and what doesn't work. I think that's very similar to what you just shared about the blueprint.

[00:21:56] Annika: Yeah, and it's super hard, because this is a default mode. Especially if you've been somewhere and a certain thing worked three times already, and you are under stress, you have a full agenda, you need to get shit done, then of course your first default is, "Okay, this is how L&D worked the last three times."

[00:22:19] Annika: "I apply it here as well. It will work as well." I think this is completely human, and it's pretty hard to wake yourself up every time and be like, "Nope, have you really validated this? This is a hypothesis for now. Let's test it and see where you can get sparring, where you can get data to back it, and then decide how to move forward."

[00:22:42] Annika: And it's not that I'm saying that for every single concept we should always validate everything bottom-up with the whole organization and do everything on a white piece of paper. But to really be open, to stay open, that's a big reminder, and I also had to set a reminder for myself quite often, because every organization is different.

[00:23:06] Annika: My former employer and Statista are both sitting in Hamburg. They were both founded not too far apart. They both have a similar size but a different setup. One was a listed company but with lower regional complexity. Statista is not listed, but had a higher regional complexity initially, and we still have a corporate investor.

[00:23:33] Annika: About You was backed by the Otto Group. Statista is backed by Ströer. So a lot of similarities, and 20 minutes walking apart. So honestly, I thought coming in that it would be a little bit same-same but different, with a slightly different business model of course, but that it couldn't be such a difference.

[00:23:54] Annika: But yeah, that's where I was wrong.

[00:23:56] Adriaan: So good. Hey, transformational change and the big words that I often speak to my guests about is AI. So let's also hear a little bit more about what is happening within your world, and how you look from your people lens at AI adoption.

[00:24:15] Adriaan: Where do you fit AI into the organization? What does that look like right now?

[00:24:22] Annika: Yeah, that's a good one. Let me start with the part about how we fit AI into the organization. One part which plays into my cards as a CHRO in terms of that adoption is the fact that AI is impacting our business model directly and at a big, big pace. So it's not only that we think about the very basic step like, oh, we have AI, which is a new technology, let's look at how we can be more efficient. We really had to reinvent ourselves.

[00:24:59] Annika: That means that in general there is a high sense of urgency in our company, and also my colleagues in the C-level all have a pretty strong affinity towards AI, because they have to in order to survive in their jobs. That is a completely different game from being in a company which is not that much affected by AI yet.

[00:25:22] Annika: We all don't know where it's going, and your task as a CHRO, first of all, is to create a sense of urgency. For us, it was there quite quickly, because it was one-on-one reflected in the revenue numbers that we have to do something. Other than that, how do I see it and how do I embrace that journey?

[00:25:44] Annika: I think at the moment I'm a continuous learner. What I do a lot is really being open-minded and seeing what the different use cases are, what topics we see in the market, and where developments are leading, to not lose sight and pace on what's going on there, so that I can transform it and carry it into the organization.

[00:26:10] Annika: That's not the most inspiring answer, but basically, at the moment, my view on AI is that I'm always astonished by how many people I see out there in the market claiming to be AI experts. The more I get into it, the more I feel that I absolutely know nothing about what's going on there, and that I'm back in the continuous learning process again.

[00:26:44] Annika: There is no blueprint that we can apply. We are all there for the very first time, and there is nothing where we can say, "Ah, they did that already three times and that was the right way." We all have to somehow try it out, which is interesting. We are all at a stage where we have to admit that we're new to something.

[00:27:05] Annika: Also as senior leaders, as C-levels, where we're just like, we don't have the answer. We have to figure it out together, we have to sit down, and we have to see what's best for the business. Of course you have that in different business situations as well, but not at the tremendous scale that we have it at the moment.

[00:27:28] Annika: So I think it's all about being open, being flexible, and really not playing entitled and being like, "I'm an expert and I know it all." But really kneeling in. What I also do, for example, so that I can see how to better drive transformation, is being really hands-on when it comes to AI.

[00:27:54] Annika: I'm really blocking time on my calendar, some hours per week, to write agents, to learn new stuff about Claude Cowork, for example. Everything where I ask myself: is this just buzz, is this noise out there in the market, or is that something which might really make sense for me?

[00:28:15] Annika: I don't know. But to judge it from the best perspective, I have to dive in knee-deep and really try it out and see what it can do for me. So this is how I view it at the moment when it comes to transformation.

[00:28:29] Adriaan: Within the company, the business model, as you said, you felt it early on. How are you driving AI adoption within the organization and within your teams? And do you maybe have one or two use cases where you've built an agent, or the company has done something, where you thought, "Oh, wow, this has really made our work more efficient"?

[00:28:50] Annika: I would answer that on two layers again. As we see, AI is completely disrupting our organization, and what I also feel is that a lot of people are underestimating AI. They're like, "Oh, it's a new tool coming in." It's like, "Oh, we now work with ChatGPT at work."

[00:29:13] Annika: "Cool." Or, "We now have used Gemini." Where I'm like, it's an artificial intelligence. It's a new form of colleague, like a digital coworker, and it's not a more advanced, re-featured version of PowerPoint that you're working with. This is a new colleague with different strengths, with different skills.

[00:29:32] Annika: So we should look at how we can best employ that colleague. How can we best work together? On concrete use cases: last year at Statista, we did a pretty tough restructuring and reorganization. We completely remodeled our content department, so basically how we produce data and how we upload data.

[00:29:57] Annika: That also meant really looking into what are human skills and what are AI skills, and which processes or process steps are better sitting with AI, because that is exactly the strength that AI brings to the table when working together, and which are the strengths that a human is bringing to the table that AI will, at least in the foreseeable future from our perspective, not be able to do in that way.

[00:30:27] Annika: And this is how we remodeled every single sub-process of the content department before we made the call on the restructuring. So it was not just, okay, let's see how many efficiency gains we can get in there. If we do it, we want to do it right once, and then we also really want to redesign how we do work, and then see what it brings us in terms of better speed, better quality, and of course also efficiency.

[00:30:56] Annika: That's one use case. I always wonder when I talk to some companies where it's like, it brings us a lot of efficiency gains. Well, agents don't run for free. You also have to pay token costs for agents. So this is the other part where I say you also have to do the math when it comes to efficiency, because maybe it's a personnel cost efficiency which is then built up on the IT side again, because it all materializes as token costs.

[00:31:27] Annika: So this is something which is important as well. But that is more of a use case at the broader company level. Then of course another use case on the business model side, which is our API business unit. So basically building APIs between Statista data and partners like Claude, Microsoft Copilot, and ChatGPT, so that you get our data transferred into the interfaces that you're using anyway. That's a concrete AI use case on the business side.

[00:32:02] Annika: And then, as we are a people podcast, going through different use cases that we did as a team recently. I actually think they're not a lot of rocket science, they were simple things, but we really saw that they moved the needle for us. Number one, a SharePoint agent where we put together everything for your personal HR operational assistant, from how do I request vacation?

[00:32:30] Annika: What was the office policy again? Where can I change my IT equipment? I need to buy new software, where do I put it in? So everything which is not only HR, but which you need to function in your daily working life, put into one agent. That agent sits in Langdock, which is a piece of software that we're using at Statista, and you just ask it like you would have a conversation with someone.

[00:32:58] Annika: On that one, of course, having an agent to automate some requests that usually go into the first line of a service center, that's not rocket science. You could do that with different software for years already. However, I think what AI did is to make this type of technology easier to access at a lower cost, so that you could not only build it or buy it on a very fancy tech stack, but that you can now just do it with a little bit of practice by yourself, and then innovate and improve as you go.

[00:33:39] Annika: Another piece: at the moment we're in the middle of our performance cycle, and last year that was my, let's say, first agent project. I set up a performance management coach for our leaders. What we put in there was really a prompt to not just empower our leaders and emphasize that everything they're doing is wonderful and they're right, but to really challenge them according to our level criteria, our ways of working, and all the different knowledge pieces that we use to build our culture and our processes at the company. So you can use that agent as your 24/7 business partner to prep: how do I put the right performance feedback in, and how do I assess that person correctly?

[00:34:34] Annika: What we didn't do...

[00:34:35] Adriaan: Oh yeah, sorry. Go on.

[00:34:36] Annika: What we deliberately didn't do with that agent, and where I'm getting a lot of questions, is that the agent, although technically possible, is not filling out your performance review. Because this is something where we say this is better done by a human, because AI will also not have the performance discussion.

[00:34:56] Annika: AI will be able to challenge you and to bring all the knowledge to you and be like, "Hey, think about that one. It's not in line with the following way-of-working principle that we have at Statista." But AI will not have the performance conversation. So AI should not be filling in the final rating and the final topics. The human should, and the human should also press the submit button.

[00:35:21] Annika: So this is something where we also deliberately thought about how to split the different process steps between the human and the new AI colleague.

[00:35:33] Adriaan: Do the performance agents also join the calls to record the actual conversation?

[00:35:39] Annika: It's just for the preparation.

[00:35:41] Adriaan: Have you thought about recording some of these conversations and saving them into an AI brain, in terms of how these conversations are done, in terms of coaching and also in terms of examples?

[00:35:55] Annika: We thought about that, yes. But at the moment we don't see it yet for performance discussions, for reasons of confidentiality. We have these coaching topics enabled for our sales colleagues, for example. If both sides are opting in, the client and also our sales representative, we have recordings of their calls, and then the system behind it gives them some insights about how they can do better, how they can prep better next time, and also guides them on topics that a client is most interested in for a follow-up conversation.

[00:36:32] Annika: But we don't do that yet for performance interviews, because what we see, even in a company being part of the AI ecosystem, if not an AI-native company, is still a lot of healthy skepticism around that. So this is something we deliberately look into. I would say every single quarter, me and my core leadership team sit down and really look into how big the appetite is, for example, to enhance AI capabilities further for performance management.

[00:37:08] Annika: That would be, for example, discussing whether we want to further automate filling out performance reviews, or an agent joining the discussions, or whether we don't want to do that. And if we don't want to do that, why? At the moment, what we say is that when it comes to AI in the workplace and also in our daily lives, a lot of people, including myself, are not yet mature enough when it comes to the appetite for AI.

[00:37:39] Annika: I don't know if I would like it if an agent would sit there beside me in a performance conversation, or if I would find that a little bit disturbing right now. However, I cannot tell you if I would say the same in 12 months, or if I would be like, "That's the most brilliant idea ever."

[00:38:01] Annika: "Let's do it." Because I think that appetite, and where you accept AI as native and natural in your daily interactions, is changing quite quickly.

[00:38:14] Adriaan: There's this famous clip from when the mobile phone came out, and people are interviewed on the street. They ask, "Hey, would you like to have a mobile phone?" And the clip is cut in a way that everyone that gets interviewed says, "No. Why would I want to have a mobile phone? Why would I want to be made available?"

[00:38:31] Adriaan: "No, I'm happy not to be disturbed all day." And it's fascinating that at that point in time people responded the way they did, and now, looking back, of course everyone has a mobile phone. And that's also the question: who knows how far we are going to go? Even to the extent, which is true, that some popular podcasts are only run by AI. They explain a specific topic, it's done by AI, and people are listening to it.

[00:38:59] Annika: Absolutely. And following up on the example that you just gave on the mobile phones: back then everyone was like, "No, the hell, I don't want something like this." And today there are even retreats advertising the fact that you can go there and be without a mobile phone and connection for five days, to rewind for once.

[00:39:22] Adriaan: Yeah.

[00:39:23] Annika: It's funny. So I also think about AI: never say never. If you had asked me 10 years ago, "Hey, Annika, this is an agent that you will write on a train ride from Hamburg to Munich, that you will test, and that you will then send to your core leadership team to test as well, and it will be literally bulletproof except for two minor adjustments."

[00:39:54] Annika: I'm sorry, I would have looked at that person and been like, "Are you crazy? That's a nice idea. Cool if I would be able to do that, but I don't think that I will be studying computer science in the near future, so I don't think so. But nice idea." And now it's possible. So never say never.

[00:40:13] Annika: Maybe in five years this is completely business as usual, and this is something that AI can help us with.

[00:40:22] Adriaan: So Annika, that being said, in five or ten years we are going to organize the first AI-free retreat with actual human beings. How's that?

[00:40:34] Annika: That's a very good idea. I think about how quickly you get used to AI, also for, pardon, the most stupid questions that I have in my daily life. Things that a long time ago I would have looked up or thought through, then I asked Google, and now, if there is something with strong articles where I don't have the time to read them...

[00:41:02] Annika: Some years ago I would have bookmarked them and sat down on the weekend to read them through. What am I doing now? Okay, sounds interesting, I don't have time, I put it into ChatGPT and ask it to send me a summary. Period.

[00:41:17] Adriaan: Great.

[00:41:18] Annika: So that will be an interesting retreat then, to see how humans go ahead without any AI.

[00:41:26] Adriaan: And then we actually have to talk to each other and be like, "Oh, wow, what's going on?" Annika, this is such a great conversation. Thank you so much for taking the time. I know you have a lot going on, you're in the midst of this transformation, so I really appreciate you sharing at this point in time what your world looks like.

[00:41:44] Adriaan: What's the best way for listeners to connect with you? Over LinkedIn?

[00:41:47] Annika: Yes, that's true.

[00:41:49] Adriaan: Great. And then for German listeners, we will also share your podcast so that they can have a listen to your podcast as well.

[00:41:56] Annika: Thank you so much.

[00:41:58] Adriaan: Thank you, Annika.

“

Need external recruitment support? Embed Matchr's expert recruiters directly into your team. Seamless integration for immediate impact on your hiring process.