YT YouTube2Text

YouTube2Text → Transcripts → AI Is a New Colleague, Not PowerPoint 2.0 | Annika in der Beek, Statista | Leaders in Talent #34

AI Is a New Colleague, Not PowerPoint 2.0 | Annika in der Beek, Statista | Leaders in Talent #34 — Transcript

by Matchr · 7,138 words · 1,023 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:01[music]
0:05When I started, we were around 700
0:07employees big and one and the half years
0:09later, we were at around 1,600
0:13and did an IPO. So, it was crazy timing.
0:16If you're coming in as the first ever HR
0:19leader from outside in a transformation
0:21phase where it really gets messy and
0:23tough because you have to have to make
0:26decisions like closing down offices for
0:28example because you you change the
0:30approach that you're selling and you do
0:32not need 13 offices anymore but rather
0:35seven. It's very important to always
0:38listen and to take your learnings but
0:41always see that you only apply the
0:43things which really match the culture,
0:45the interest and the maturity stage of
0:47the company. There is no blueprint that
0:49we can apply. We all now are there in
0:53the very first time and there is nothing
0:55that we can see and they did that
0:57already three times and that was the
0:59right way. We all have to somehow try it
1:02out which is interesting. We are all at
1:05a stage where we have to admit that
1:08we're new to something. It's an
1:10artificial intelligence. It's a new form
1:12of colleague like a digital coworker and
1:15it's not like a more advanced reffatured
1:18version of PowerPoint that you're
1:19working with. But this is a new
1:21colleague with different strengths with
1:24different skills. So we should do we
1:27should look how we can best employ that
1:29colleague. Right? So how can we best
1:31work together? All right, ladies and
1:33gentlemen, and welcome to another
1:35episode of the Leaders in Talent
1:37podcast. And today I am welcomed and
1:39joined by Anuka in their bake. Anuka,
1:42welcome.
1:43>> Thank you so much for having me. You're
1:45really looking forward.
1:47>> So Anukica, before we started the actual
1:49recording, I realized you're my first
1:51guest that has an actual podcast
1:54microphone. And little did I know that
1:56I'm speaking to a podcast expert. Tell
2:00me, tell me, tell me a little bit more
2:01before we dive into your background.
2:03>> Okay. Wow. Pressure is on. Not a podcast
2:05expert and I have to admit whenever I'm
2:07a guest on a podcast, I'm still nervous.
2:10So also practice doesn't make it vanish
2:13completely. But yeah, since I think two
2:16and a half years together with my
2:17colleague and friend Angelie, I'm
2:19hosting a podcast also an HR podcast.
2:21It's in German unfortunately. Only
2:23interesting for all German listeners
2:25today. It's called HR Unming the Al
2:28podcast. Basically HR unfiltered, the
2:31honest HR podcast.
2:33>> So good. So that's why you have great
2:35sounds, great lights, which I'm which
2:37I'm always very happy to see when when
2:38guests come on. Um,
2:40>> we're too lazy to record with video. We
2:42always only do sound.
2:44>> We only do sound. [laughter] Yeah.
2:46>> Maybe this this podcast is going to
2:48change that. Let's let's see. Hey Anuka,
2:50can you give our listeners a little bit
2:51of context about your background and
2:53your current role as the CPO at
2:55Statista?
2:56>> Yeah, sure. I'm now in a role as a CO at
2:59Statista for around a bit more than two
3:02years already. Time is flying. And what
3:05is Statista? What are we doing? We are
3:07one of the leading global data platforms
3:10who provide access to different data
3:13reports that you might need for use
3:15cases. for example in consulting market
3:17research whatsoever. I think the
3:20majority of my friend circle whenever I
3:23tell them that I work at Satista or when
3:25I told them that I'm joining everyone
3:27was like oh wow I know them from my
3:29master thesis from my bachelor thesis
3:31from my time in consulting because
3:33everyone was pulling data from us to
3:36back their research. So this is one part
3:38what we're doing that we're known for.
3:41Then we also so that means whenever you
3:43see for example top employers rating
3:46with companies like Forbes like Newsweek
3:49most innovative companies so they also
3:52most of the time done together with
3:54Statista which is called Statista R and
3:57then we also of course provide market
3:59research. So customized market research
4:02that we do throughout consulting
4:04projects for our clients. So everything
4:06basically data and also with our mission
4:08we empower people with data. This is
4:11what we do and we will dive into that a
4:13little bit deeper later because of
4:16course with GPT etc coming up we are in
4:18the middle of a transformation and this
4:20is not everything that we do anymore.
4:23That's so far about Statista. As I said,
4:26CHO since more than two years and time
4:28is flying middle in the transformation.
4:30And before I joined Statista, I am a
4:34psychologist somehow gone business. I
4:36started my career in mergers and
4:37acquisitions where I spent two years in
4:40Hamburg and London. Then switched to
4:42Accenture where I started in um HR
4:45consulting and then in 2020 exited to
4:48about you an e-commerce tech company
4:50based in Hamburg as well. joined them
4:53middle of the pandemic 2020 in August
4:57and when I started we were around 700
4:59employees big and one and the half years
5:01later we were at around 1,600
5:05and did an IPO so it was crazy timing
5:08yeah and now happy and busy at Statista
5:12>> how big just of context how many
5:14employees does Statista have
5:15>> yes we are now at around 1,000 employees
5:18and we split about seven countries
5:21>> yeah and predominantly in Europe or also
5:23across different continents.
5:25>> Across different continents. So
5:26basically if if we look at a follow the
5:28sun time zone approach, we we sell our
5:31products 247. So our headquarter is in
5:34Hamburg where also the majority of our
5:37people sit, but again we have regional
5:39headquarters as well. For the Americas,
5:42our regional headquarter is in New York.
5:44It's also our second largest office for
5:46EMIA region. This is this is London. And
5:50for APEC, our headquarters Singapore and
5:52we have different different offices in
5:54Tokyo and in Japan.
5:57>> Great. Awesome.
5:58>> In Tokyo and in Japan. No, in Tokyo and
6:00in India. So that's other than that.
6:04You cut yourself. You cut yourself
6:05there. You mentioned prior. So recently
6:08Satisa was a media company, but now the
6:11company's transforming into a data and
6:12software business. Tell me a little bit
6:14more about the fundamental change that's
6:17happening within the company and how you
6:19think about people, talent and and orc
6:21design.
6:23>> That's a very good question. Let me
6:25start with the fundamental change. So if
6:27I look at statista from the outside
6:29perspective coming in back then statista
6:32was a lot split in silos. So giving you
6:36the example one team that worked only on
6:41research data. So everything for example
6:43that you have from external parties that
6:46you acquire data that you then put
6:48together in a report for example then
6:50there was a team for model data. So
6:52everything where you're like hey we
6:54don't get the data from somewhere else.
6:57So we set up our own market model for
6:59example and create our own market
7:01estimations and these were two separate
7:03teams bringing out two separate
7:06products. So pretty siloed from each
7:10other, right? And this is everything
7:12that you could also see throughout the
7:15whole organizational setup. Why did it
7:17make sense? So if we look into the
7:20history of Statista, it's seen
7:24tremendous growth. So it always has been
7:27a big success story with a with a
7:30strongly growing top line. If you look
7:32at that, it it just looks amazing. And
7:35whenever there was a crisis in the
7:36market so where other companies faced a
7:38dip for example COVID the demand for
7:42data just grew right. So everything like
7:45high growth hypers scale success story
7:49and I think also keeping keeping things
7:52in silos with the business model that we
7:55had back then that was super beneficial.
7:58Sounds a little bit counterintuitive but
8:00what does a silo do? It gives you the
8:03opportunity to innovate and to really
8:06put up speed on some topics that you
8:09really need to move fast because there
8:11is less alignment. You should not miss
8:14the moment to turn this around because
8:16then there is the phase where you need
8:18more synergies to to get things done.
8:21And I think this is exactly the case. So
8:24really thinking about not how do we do
8:27things and how we do things we also push
8:29them out to the market because clients
8:32exactly want the products as we provide
8:36them to really switch towards a
8:38different market environment with
8:40different data access layers where
8:44for example speaking about access layers
8:47if you now put in or if you're looking
8:49for data I think some years ago You
8:53might have put that into Google and also
8:55with the marketing concept of Statista,
8:58we were perfectly tagged. So the
9:00probability that we were ranking
9:04in the first non-paid Google
9:07advertisements was pretty high.
9:10>> Yeah.
9:10>> Now if you put that into CHPT or CL for
9:13example, this is a complete different
9:15different game, a different deal, right?
9:17That means we have to rethink how we can
9:19best group our data together. No matter
9:22from which source because in the end
9:25when you're a client what's your use
9:27case? You have a question and you need
9:29verified data to answer it. If it's like
9:32own modeled or own researched or if it's
9:35like from a third party report, I would
9:39assume you generally don't care as long
9:41as it's valid, as long as you can take
9:44it and if it's a credible source that
9:46you're using. And this is a complete
9:49different way and journey um how to do
9:51things. And yeah, this also of course
9:54means to really put together to to
9:56really put together our teams more to
9:59increase collaboration to break down
10:01silos. Number one. Number two, of
10:05course,
10:07moving from a more media publishing
10:09oriented company towards a software as a
10:13service company also requires different
10:16type of skill profiles and talent. For
10:19example, if you run a platform that you
10:24can download reports from, for example,
10:26as we did, yes, that's a tech product.
10:29But how many platforms are out there
10:32that you can download products from,
10:34right? So, I would say yes, a tech
10:36program with strong traffic, but there
10:39are more sophisticated tech tech
10:41products out there, right? If we now
10:44look into the development of our
10:46business model, for example, our
10:48partnerships that we have with OpenAI,
10:51that we have with Claude, where via an
10:53API key, you have our data accessible in
10:57one of the most leading LLMs in the
10:59world. Of course, this is a complete
11:02different traffic that you're facing.
11:04This is a complete different business
11:05model and this is a complete different
11:08layer of complexity because you're
11:10building a different product. And this
11:12is something which yeah which which was
11:14also new to the company and yeah where
11:18we had to do a workforce transformation
11:20basically. So, so tell me let's zoom in
11:22on on the practical aspect, right? So,
11:23so this is the broader business context,
11:25but but what does that mean from a a CPO
11:28CRO perspective on how you look at this,
11:31right? Because
11:33>> breaking through silos like what does
11:35that mean in reality? Like how do you go
11:37about of of that transformation, hiring
11:40potentially new people, looking at your
11:41org charts? Yeah,
11:44bring it bring it a little bit more in
11:46terms of the day-to-day what that looks
11:47like for you. It's actually a mixture of
11:50a lot of topics. So number one asking
11:52which skills do I need? What are the
11:54topics? What are our must-win battles
11:57and what are topics and skills and
11:59people that we need in the company to
12:02make this happen and this is on one part
12:05hard skills that we need to look into a
12:08lot but also in a transformation phase
12:11in a difficult market environment. It's
12:14a complete different deal which mindset
12:16you need to succeed in that environment
12:19than for example in a very successful
12:21all funny hypers scale environment right
12:23so you really need to rebuild and
12:26transform things from scratch and
12:28transformation can be messy and can be
12:31tough and especially if that's not
12:33linked to direct success out there
12:35because the market is not always playing
12:37in your favor and there is a lot of
12:39pressure. So that's like number one when
12:43it comes to talent. So do we have the
12:46right capable leaders from a skill
12:48perspective also from a mindset
12:50perspective and from a resilience
12:52perspective because it's really it's
12:55that the last two years have been tough
12:58for us in the transformation. Then also
13:01from from the Oakchart perspective on
13:03the other hand how do we best
13:05collaborate together? Do we still have
13:07the right sea level units in place? Do
13:10we have the right VP level units in
13:12place? Do we have for example
13:15>> what do you mean with sorry Ana? What do
13:17you mean with VP units in place? What do
13:19you mean?
13:20>> What I would mean with a VP unit is if I
13:22have a seale unit people and then a VP
13:25unit would be for example people
13:27operations or talent talent acquisition.
13:30So basically the teams below that are
13:32they still working together in the right
13:34way? So giving you one example um if I
13:38work with silos and this was the same
13:40for my role in the beginning although
13:43the company back then was also plus
13:471,000 people nearly 1,400
13:51across 13 countries there was no central
13:54HR there was central HR to to do
13:57contracts and to do payroll but what you
14:00had was different people in different
14:02divisions and and different continents
14:04sitting and and doing HR with the
14:07regional lead or with the divisional
14:09lead who did not have an HR background.
14:12So which I which is which is not bad in
14:16general. However, when you come there
14:18and you just see okay yeah there are
14:20some HR teams but these HR teams are
14:24like nine different ones and they all
14:26have different performance approaches
14:28and different hiring approaches and
14:30different cultural approaches on how
14:32they do things. This is not one company.
14:35This operates like nine companies. And
14:37this is what you find in a day-to-day
14:39business. If you speak about breaking
14:41down silos, you will see that across the
14:43whole organization. And this is really
14:46something where you have to see what are
14:48topics that are more beneficial to
14:50centralize, how much do we centralize
14:52them? and which topics you deliberately
14:54leave within the divisions because you
14:57say okay they still need to function
15:00individually because of speed because
15:03they don't have a lot of collaboration
15:06touch points into other departments. So
15:08all these design decisions this is
15:10something that you take on a day-to-day
15:13basis with one goal in mind but which
15:17basically does not stop. So it's not
15:20that you sit down one week and you're
15:21like this is how the Octad looks like
15:23and let's execute next week.
15:25>> This is something which always is a
15:27step-by-step bystep approach.
15:30>> And coming back to your first question,
15:32I think this is also one part that that
15:35transformation really showed me again
15:38transformation and change doesn't stop.
15:41So that means it's a continuous process
15:44and if we like it or not. So I think
15:47humans in general I would say by our
15:50nature we're not the biggest fans of
15:52change and having everything different
15:54every single week but we need to get
15:57through it and we also should not lie
15:58into our faces like okay we now do the
16:01transformation and then change will be
16:03over because it's not true. If we do the
16:06transformation or first part of the
16:08transformation and three months later
16:11with the current technological pace
16:13everything still can look different
16:15again. So you need to retransform. So
16:18this is this this is basically yeah the
16:20working mode being a big advocate for
16:23that transformation is fun now right.
16:26>> Yeah. So so it's interesting right? So
16:27so so you joined Satissa two years ago.
16:30>> Yeah. Did you know coming in like hey
16:33your mandate is leading this
16:34transformational organizational change
16:36because of the different different
16:38markets out there? Was that something
16:40that was already ongoing? Um and and how
16:43did you decide in terms of what that org
16:46chart was going to look like? Yeah. So
16:48sorry many questions packed in in one
16:50but take me take me back to to to the
16:52moment that you arrived and and where
16:53you are now on that journey.
16:56So I have to say I of course didn't have
16:58full transparency around what's waiting
17:01there. I think mayor so I think in
17:04general because of one one thing the two
17:07founders of statista stepped out by the
17:10end of 2023 and our new CEO took over to
17:14lead that transformation beginning of
17:162024
17:18when they hired me was end of 2023. So
17:21that means I still had my interviews
17:23with our investor, with the former
17:25founders, with our new CEO. But of
17:28course, our CEO also just started and
17:31could not tell me everything, but could
17:33share with me first hypothesis that that
17:36are going on. Secondly, I also think
17:38that at least myself, I completely
17:42underestimated the depth of how AI is
17:46changing, how business models are
17:48operating, and how we're doing work.
17:50That's that's the second the second
17:52layer that I was like, "No, I did not
17:55completely know what I was buying into."
17:58And how do you decide it on a on a day
18:00by day? Honestly, it's a lot about
18:04listening, knowing where you want to go,
18:08working together with your colleagues.
18:10Because in the end, it's not me to
18:12decide, yeah, this is the ORC chart.
18:14It's me to decide how the org chart
18:16looks like for my specific division but
18:19not for the overarching org. Right? So
18:21this is something where I have to work
18:24closely on a day-to-day basis together
18:26with our CEO and together with our
18:28colleagues to see how to design the best
18:30orc setup and how to best move forward.
18:34>> Hey, and now you're you're two years in
18:36this journey, right?
18:37>> What are some what are some of the
18:39things now reflecting you're in the
18:40midst of it, right? But what are some of
18:42the things like reflecting back that you
18:44would advise your 2-year younger younger
18:46self on on going on that journey?
18:49>> I would advise myself first of all to
18:51not work like 24/7 over a month because
18:54you want everything to be done in in no
18:57time and want to set up your own
18:59department because you will not be able
19:00to give from an empty cup. So that's
19:03that's that's one part. The other part
19:05is which was pretty tough for me is of
19:09course you know that you're not the the
19:12most like person in the whole company if
19:15you're running HR because you also have
19:17to stand up for a lot of a lot of tough
19:20decisions that if you're coming in as
19:23the first ever HR leader from outside in
19:26a transformation phase where it really
19:28gets messy and tough because you have to
19:31have to make decisions like closing down
19:33offices for example example because you
19:35you change the approach that you're
19:37selling and you do not need 13 offices
19:39anymore but rather seven then of course
19:43the organization will not always like
19:45you and it also will take longer time
19:48for them to build trust to to you as a
19:51person than you coming in in a
19:53hyperrowth phase where everything that
19:55you do is hiring and really seeing how
19:58you can staff up teams and come across
20:00with of course a lot of work but way
20:03more positive and rewarding topics than
20:06these SWATs and I think these were the
20:08the two the two topics that I would
20:11advise my younger self on and then
20:13another learning
20:15also you will not be able to come with a
20:18blueprint and just apply what you did at
20:21other companies to the new organization.
20:24So of course it's good if you have a
20:27certain been there done that mindset and
20:30you don't get shocked by by certain
20:32topics. However, it's very important to
20:35always listen and to take your
20:37learnings, but always see that you only
20:40apply the things which really match the
20:42culture, the interest, and the maturity
20:44stage of the company.
20:46>> This is so interesting. I just had a
20:48call with the SVP of talent of an AI
20:51data center hyperscaler
20:53>> and he and he he's had a tremendous
20:55career. He worked at the largest tech
20:58companies on the on the planet, always
21:01in in head of TA roles, and he's seen
21:04what these big companies look like.
21:06>> And
21:07he sat down with the CEO on his second
21:10week and he was like, "All right, this
21:11is my background and I'm going to do
21:13this and this is what I learned from
21:14there and we're going to do this and
21:15we're going to do that." And he shared
21:18the story is that the CEO stopped him in
21:20his tracks and he said, "No, I do not
21:23want you to do that." And he was a
21:24little bit taken back. is like well I've
21:26learned from the best right and he's
21:27like we are we are not that company we
21:30are doing something very different I
21:32actually want you to write your own
21:35canvas and your own playbook on how to
21:37look at this and yes of course you come
21:39with experience but let that not cloud
21:42your judgment of what is needed here and
21:44I thought it was such a a profound story
21:46of coming in somewhere with with your
21:48your bag of knowledge and and also still
21:51having an open mind of what works and
21:53what doesn't work and I think that's
21:54very similar to what you just shared in
21:55terms of that blueprint.
21:57>> Yeah. And it's super hard because you
21:59really have to of course this is a
22:01default mode because especially if
22:03you've been somewhere and a certain
22:05thing worked like three times already
22:08and you are under stress, you have a
22:10full agenda, you need to get done.
22:12Then of course your first default is
22:15okay this is how L & D worked the last
22:18three times. I apply it here as well. It
22:20would it will work as well. I think this
22:22is this is completely human and it's
22:25pretty hard to to wake up yourself every
22:28time and be like nope have you really
22:30validated this this is hypothesis for
22:33now
22:34>> let's test it and see where you can get
22:36sparing where you can get data to to
22:38back it and then decide how to move
22:41forward and it's not [snorts] that I'm
22:43saying for every single concept we
22:45should always validate everything bottom
22:48up with the whole organization and do
22:50everything on on a on a white piece of
22:52paper. But I think to really be open,
22:56stay open, that's a big reminder. And I
23:00also had to set the reminder to myself.
23:02Yeah. Quite often because every every
23:05organization is different. And also my
23:07my former employer and Statista
23:10basically they're both sitting in
23:12Hamburg. They're both not founded too
23:15far away. They both have a similar size,
23:18a different setup. One was one was a
23:21listed company but with lower like
23:23regional complexity. Statista is not
23:26listed but had a higher regional
23:28complexity both initially and we still
23:31have have a corporate investor about you
23:34was backed by backed by the auto group.
23:36Statista is backed by right so a lot of
23:38similarities and 20 minutes walking
23:41apart. So honestly
23:44>> I thought coming in it will be a little
23:46bit same same but different with a
23:48slightly different business model of
23:50course but cannot be cannot be such a
23:53difference but yeah why I was wrong.
23:56>> Yeah yeah yeah so good so good. Hey and
23:58transformational change the big words
24:01and often that I speak to speak to my
24:03guests about is AI. So let's let's also
24:05like hear a little bit more about what
24:07is happening within
24:08>> your world and how do you look how does
24:11how do you look from your people lens to
24:13AI adoption where do you fit AI into the
24:17organization
24:19and and and what does that look like
24:21right now? Yeah, that's a good one and
24:23and let me start with a part how do we
24:26fit AI into the organization. I think
24:28one part which plays into my cards as a
24:32CHO in in terms of that adoption is the
24:36fact that AI is impacting our business
24:40model directly and at a big big pace. So
24:46it's not only that we think about the
24:49very basic step like oh we have AI which
24:52is a new technology let's look at how we
24:54can be more efficient but we really had
24:57to reinvent ourselves so that means that
25:00in general there is a high sense of
25:03urgency in our company and also my
25:05colleagues in the sea level all have a
25:09pretty strong affinity towards AI
25:11because they have to do it to survive in
25:14their jobs which is a complete different
25:17game than being in a company which is
25:19not that much affected by AI yet. We all
25:23don't know where it's coming and where
25:26your task as a CHO first of all is to
25:30create a sense of urgency for us. It was
25:32there quite quickly because it was
25:34one-on-one reflected in the revenue
25:36numbers that that we have to do
25:38something other than that how how how do
25:41I see it and how do I embrace that
25:43journey? I think at the moment I'm a
25:46continuous learner. So what I do a lot
25:49is really being open-minded and seeing
25:54what what are different use cases, what
25:56are topics that we see in the market,
25:58where are development leading to not to
26:01not lose sight and and and pace on
26:04what's going on there to see how I can
26:07of course transform it and carry it into
26:09the organization. And
26:12that's that that's that's not the most
26:14inspiring answer, but basically at the
26:16moment it's just for me what's my view
26:19on AI. I always get I'm always
26:22astonished by how many people I see out
26:25there in the market claiming about
26:27themselves being AI experts.
26:30>> And I'm just like the more I get into
26:33it, the more I feel that I absolutely
26:38know nothing about what's going on
26:39there. and that I'm back in the
26:42continuous learning process again. There
26:44is no blueprint that we can apply. We
26:47all now are there
26:49>> in the very first time and there is
26:51nothing that we can see and they did
26:53that already three times and that was
26:55the right way. We all have to somehow
26:57try it out which is interesting.
27:00We are all at a stage where we have to
27:04admit that we're new to something also
27:06as senior leaders as sea levels where
27:08we're just like we don't have the
27:10answer. We have to figure it out
27:11together and we have to we have to sit
27:15down and we have to see what's best for
27:16the business. I think
27:19that's of course you have that in in in
27:22in different other business situation as
27:24well but not at that tremendous stage as
27:27we have it at the moment. So therefore,
27:29I think it's all about being open, being
27:33being flexible and really not not
27:36playing entitled and not being like I'm
27:38an expert and I know it all. Um but
27:42yeah, really kneeling in what I also do
27:46for example that I can see how to better
27:49drive transformation.
27:50I'm really kneeling in hands on when it
27:53comes to AI. Like I'm really blocking
27:56time on my calendar for example. some
27:58hours per week to really also to to
28:01write agents to learn new stuff about
28:04clo for example right so everything
28:07where I'm like is this just a bus is
28:10this a noise out there in the market or
28:12is that something which might really
28:14make sense for me I don't know but to
28:16judge it from the best perspective I
28:19have to dive in knee deep and really try
28:21it out and see what it can do for me so
28:24this is something how how I view it at
28:26the moment when when it comes of
28:28transformation
28:30>> is within the company right so so the
28:31business model is is as you said you
28:33guys felt it early early on how how are
28:37you driving AI adoption within the
28:39organization within your teams and do
28:41you maybe have like one or two user
28:43cases where you've built an agent or or
28:46the company has done something where
28:47you're like oh wow this has really made
28:49our work more efficient
28:51>> so I think would would answer that on on
28:54two layers again as as we See, yeah, AI
28:58is completely disrupting our
29:00organization. And what I also feel is
29:02that a lot of people are underestimating
29:05AI. So that means that they're like, oh,
29:07it's a new tool coming in. It's like,
29:10oh, we now we now work with JPT at work.
29:13Cool. Or we now have use Gemini where
29:15I'm like, it's an artificial
29:17intelligence. It's a new form of
29:19colleague like a digital co-worker and
29:22it's not like a more advanced reffatured
29:25version of PowerPoint that you're
29:26working with but this is a new colleague
29:29with different strengths with different
29:31skills. So we should do we should look
29:34how we can best employ that colleague
29:37right. So how can we best work together
29:39and I think on concrete use cases last
29:42year at Statista we did a pretty
29:46pretty tough restructuring and
29:48reorganization. So we completely
29:50remodeled our content department. So
29:53basically how we produce data and how we
29:56upload data. And that also meant that
29:59really looking into what are human
30:01skills and what are AI skills and which
30:05processes or process steps are better
30:08sitting with AI because this exactly is
30:11the strength that AI brings to the table
30:14when working together. And these are the
30:16strengths that a human is for example
30:19bringing to the table that AI will at
30:22least in the foreseeable future from our
30:24stance not be able to do in that way.
30:27And this is how we remodeled every
30:29single subprocess on from the content
30:31department before we before we made the
30:34made a call of the restructuring. So it
30:36was not just like okay let's see how
30:39many efficiency gains we can get in
30:41there but really like if we do it we
30:43once want to do it right and then we
30:45also really want to redesign how we do
30:48work and then see what it brings us in
30:51terms of better speed better quality but
30:53also of course in terms of efficiency
30:56that's one use case I also I always
30:59wonder when when I talk to to some
31:01companies where it's like it it brings
31:04us a lot of efficiency gains
31:06Well, also agents don't run for free,
31:09right? You you also have to pay total
31:11costs for agents.
31:12>> So, this is the other part where I'm
31:14like you also have to do the math when
31:16it comes to efficiency because maybe
31:19then it's a personal cost efficiency
31:22which is then built up on the IT side
31:24again because it's like all materialized
31:26by token costs, right? So, this is
31:28something which which is which is
31:30important as well. But this like more of
31:33a use case of of the broader company
31:35level. Then of course another use case
31:38on the business model side which is our
31:41API business unit. So basically building
31:44APIs between statista data and and
31:47partners like clude like Microsoft
31:50copilot like chat GPT so that you get
31:53our data transferred into the interfaces
31:56that you're using anyways which is an
31:58interesting one which is a concrete AI
32:00use case on the business side and then
32:03as we are a people podcast going through
32:06different use cases that that we did as
32:08a team recently
32:10I actually think they're not Not a lot
32:13of rocket science, but those were like
32:15simple things, but where we really saw
32:17that they that they moved the needle for
32:19us.
32:20>> Number one, a SharePoint agent where we
32:23put together everything like your
32:25personal HR operational assistant from
32:28how do I request vacation? So, what was
32:31the office policy again? Where can I
32:33change my IT equipment? I need to buy
32:35new software. Where do I put it in? So
32:38everything which is not only HR but
32:41which you need to function in your in in
32:44your daily working life.
32:46>> Yeah.
32:46>> To put that into one agent. That agent
32:49sits in Langdoc. So so the software that
32:51we're using at Statista and you just ask
32:55it like you would have a conversation
32:56with someone. And I think on that one,
33:01of course, having an agent to automate
33:04some requests that usually go into the
33:06first line of a service center, that's
33:09not rocket science. You could do that
33:11with different softwares since years
33:13already. However, I think what AI did is
33:17to
33:19make this type of technology
33:22easier to access at a lower cost so that
33:25you could not only build it or buy it on
33:28a very fancy tech stack, but that you
33:31can now just do it with a little bit of
33:33practice by yourself and then comp like
33:37innovate and improve as you go. Another
33:40piece at the moment we are in the middle
33:42of our performance cycle for example and
33:45last year that that was my let's say
33:47first agent project I set up a
33:50performance management coach for our
33:52leaders. So what we put in there was
33:54really
33:55>> a prompt to also not just
33:58>> empower and emphasize to our leaders
34:01that everything that they're doing is
34:03wonderful and they're right, but really
34:05also challenge them according to our
34:08level criteria, our ways of working and
34:11all the different knowledge pieces that
34:13we use to build our culture and build
34:16our our processes um at a company. so
34:19that you can use that agent as your 247
34:23business partner to really prep
34:25according to how do I put right
34:28performance feedback in
34:30>> y
34:30>> and how do I how do I assess that person
34:33correctly what we didn't do
34:35>> oh yeah sorry go on yeah
34:37>> what we deliberately didn't do with that
34:39agent um where I'm getting a lot of
34:41questions from is that agent although
34:44technically possible is not filling out
34:47your performance review Because this is
34:49something where we're like this is
34:50better done by a human because also AI
34:54will not have the performance
34:55discussion. AI will be able to challenge
34:58you and to bring all the knowledge to
35:00you and be like hey think about that
35:02one. It's not in line with like the
35:04following way of working principle that
35:06that we have at Statista. But AI will
35:09not have the performance conversation.
35:11So I should not be filling in the final
35:14rating and the final topics but the
35:17human should and also the human should
35:19should press the submit button. Right?
35:21So this is something where we also
35:23deliberately thought about how to split
35:26>> the different the the different process
35:28steps deliberately between human and
35:31between the new AI colleague.
35:34>> The performance agents also join the
35:36calls to record the actual conversation.
35:39No, it's just for the preparation.
35:41>> Have you thought about that of of
35:43recording some of these conversations
35:45and and you know saving them into an AI
35:49brain in terms of how these
35:51conversations are done in terms of
35:52coaching and also in terms of examples.
35:56We thought about that yes but at the
35:58moment we don't see it yet for for
36:00performance discussions I think for the
36:03way of confidentiality right we have
36:05these coaching topics for example
36:07enabled for our sales colleagues if both
36:10sides are opting in client and also our
36:14sales representative we have recordings
36:16for example of their calls and then also
36:19the system behind that gets them gets
36:21them some insights about how they can do
36:23better how they can prep better next
36:25time and also guides them about topics
36:27that the client for example is most
36:29interested in for a follow-up
36:31conversation. But we don't do that yet
36:33for performance interviews
36:36because what we see even in an AI
36:39company or not an AI native company but
36:42a company being part of the AI ecosystem
36:45is still of course a lot of also healthy
36:48skepticism around that. Right? So I
36:51think this is something we deliberately
36:55I would say every single quarter me and
36:57my core leadership team sit down and
37:00really look into topics how big the
37:02appetite for example is to enhance AI
37:05capabilities further for performance
37:08management that would be for example
37:10discussing do we want to further
37:12automate for example filling out
37:14performance reviews or an agent joining
37:17the discussions or don't we want to do
37:19that and if we don't want to do that.
37:21Why? And this is something at the moment
37:24that we say we think that AI in the
37:28workplace and also in our daily lives. A
37:31lot of people including myself
37:34are not yet mature enough when it comes
37:37to the appetite in terms of AI. So, I
37:40actually would be
37:44I don't know if I would like it if like
37:46an agent would sit there beside me in a
37:48performance conversation or if I would
37:50by now find that a little bit more
37:52disturbing. However,
37:55>> I cannot tell you if I would say the
37:57same in 12 months or if I'm like that's
37:59the most brilliant idea ever. Let's do
38:01it. Because I think also that appetite
38:04and where you accept AI as like native
38:08and natural in your in your daily
38:11interactions that's changing quite
38:14quickly.
38:14>> I there's this famous clip when the
38:16mobile phone came out and and people are
38:18interviewed on the street and they ask,
38:20"Hey, would you like to have a mobile
38:22phone?" And every the clip is cut in a
38:24way that everyone that that gets
38:26interviewed says, "No, why why would I
38:28want to have a mobile phone? Why why why
38:30would I want to be made available?" No,
38:32I'm I'm I'm happy to to not be disturbed
38:34all day. Right. And it's fascinating
38:36that at that point in time people
38:38responded the way that they did. And now
38:40looking back, right? Of course, everyone
38:41has a has a mobile phone. And that's
38:43also the question, right? Like who who
38:44who knows how far we are going to go
38:48even to the to the extent which is true,
38:49right? Some popular podcasts are are
38:52only run by AI, right? They explain a
38:55specific topic
38:56>> and it's done by AI and people are
38:58listening to it.
38:59>> Yeah.
39:00>> Absolutely. And and I also think
39:02following up on the example that you
39:03just gave on the mobile phones, right?
39:06And back then everyone was like, "No,
39:08the hell I don't want something like
39:09this." And today there are like even
39:12retreats advertising for the fact that
39:15you can go there and that you will be
39:17without mobile phone and connection for
39:19five days to once rewind.
39:22>> Yeah.
39:22>> Which is funny. So therefore I also
39:25think about AI. Never say never. If you
39:28would have asked me 10 years ago to be
39:31like, "Hey, Aneka, this is an agent that
39:34you will write down on a train ride from
39:38Hamburg to Munich and that you will
39:42write down that you will test and that
39:44you will then send to like your core
39:46leadership team to test as well. And it
39:49will be literally bulletproof except for
39:51like two minor adjustments." I'm sorry.
39:55I would have looked at that person and
39:56would be like are you crazy? That's like
39:59a nice idea. Cool. If I would be able to
40:01do that, but I don't think that I will
40:04be like studying computer science in the
40:06near future. So I don't think so. But
40:08but nice idea and and now it's possible.
40:11So therefore never say never. Maybe in
40:14five years this is complete business by
40:16usual and this is something um that um
40:19yeah AI can help us with. So Anukica,
40:23that being said, in five or 10 years, we
40:26are going to organize the first AI free
40:29retreat with actual human beings. How's
40:32that?
40:33>> How's that?
40:34>> That's that's that's a that's a that's a
40:36very good idea. just I think how how
40:38quickly you get used to AI also for for
40:41like the most pardon but stupid
40:44questions that I have in my daily life
40:46that I think very long time ago I would
40:50have like looked it up or thought it
40:53through then I asked Google and if there
40:56is now something just giving you like
40:58strong articles or something where I
41:00don't have the time to read it I think
41:03some years ago I would have bookmarked
41:05them and sat down on the weekend for
41:07example to read them through what am I
41:10doing now okay sounds interesting I
41:13don't have time I put it into chat GPT
41:15and ask her to send me a summary period
41:18so therefore that that will be an
41:20interesting retreat then to see how
41:22humans then go ahead without without any
41:25AI
41:26>> yeah and then we actually have to talk
41:28to each other be like oh wow what's
41:30going on anam this is such a such a
41:32great conversation thank you so much for
41:34for taking the time I know you have a
41:36lot going on. You're in the midst of
41:37this transformation. So, I really
41:39appreciate you at this point in time
41:42sharing what what your world looks like.
41:44What's the best way for listeners to
41:45connect with you over LinkedIn?
41:47>> Yes, that's true.
41:49>> Great. And then for German listeners, we
41:51will also share your podcast so that
41:53they can have a have a listen to your to
41:55your podcast as well.
41:57>> Thank you so much.
41:58>> Thank you, Anukica.
42:02>> [music]

About this transcript

This page contains the full transcript of AI Is a New Colleague, Not PowerPoint 2.0 | Annika in der Beek, Statista | Leaders in Talent #34 by Matchr, generated from the public captions YouTube serves with the video. The transcript has 7,138 words across 1,023 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.