AI Hype around HR and HR Tech with Romy von Roeder

Thomas Kohler:
Today’s guest, Romy von Roeder. What is the fear of people within companies that they are not future proof?

Romy von Roeder:
Yeah, I mean, I think I have proof and I think I can talk about my own experience. I can 100% agree that there is a lot to do. And that was the core of my presentation previously, that it’s not about fancy tools or like implementing everything that we come across. Right. And yeah, just focusing on the new things, it’s very easy. Things that you have to fix beforehand. For example, the data accuracy. So most of the data AI, data startups in HR that I’m talking to, they say the biggest problem that they have is that the data is not correct.

So when they want to work with the data, what comes out is basically not valuable because of the data that has been put in. So, you know, the old saying should in, should out. So what we have to do first is we have to clean up all our data. And these startups also start or have a program to help companies to clean up the data to provide real value. Otherwise companies would probably not decide to use the tool because they think, oh yeah, it doesn’t work either. Right? Yeah. So that’s one of the things that needs to be fixed early on.

Thomas Kohler:
Romy and I know each other for years, and we did a new update episode on AI in the people field and what problems to solve, she found, and also how she thinks about the hype cycle of AI overall or technology overall, and also how it translates into leveraging or using AI in the right way now in the HR field. So practical insights and also some examples for everyone, now, having AI and HR on the table.

Thomas Kohler:
Today I’m excited because I have one of my favorite when not my most favorite guest here, Romy von Roeder. We already had a big legacy of doing events together, working together at some customers, always exchanging each other, and also doing a podcast already. And now I saw, I think early this year that you did a talk on AI, and you also did a lot of, I would say, investments in HR tech and so on. And also AI is a big topic. So I’m glad that we could again do an update episode on your perspective on AI. But maybe, Romy, we can just start shortly with a quick intro of what happened over the past, maybe, I think one year or one and a half years, because this is the last time where we talked publicly.

Romy von Roeder:
Yeah. Okay. First of all, thank you so much, Thomas, for the very, very nice introduction. I can only give that back. I very much enjoy working with you and hope to be able to do that very often again. Yeah. What happened over the last one and a half years in my life, things change in a way that I started working in HR consulting and looked a little bit into what companies need in terms of leadership and culture, supporting them, being a better employer to their employees, being even better employers to the employees. And, yeah, I mean, as we speak about AI and HR, I think that’s something that changed a lot.

Right. So I joined a community, for example, that is called the GPT community or peoplegpt.com. I highly recommend to visit the website and check it out, founded by Kyuhamirani. So yeah, we are pretty excited looking into this topic and seeing what can be achieved with this new technology. And I must say, in my daily work, it’s already, it’s a huge advantage compared to how we had to develop our work or products before.

Thomas Kohler:
Yes, and I definitely agree and see that I also use it daily, I think especially in recruiting for taking and structuring notes, because in interviews, I think nobody really likes it when somebody is taking notes all the time. And I think Metaview, for instance, they had a very exciting ad about a recruiter that is just typing, typing, typing, but not really listening. And I think that’s something from the past. So in that sense, or in sales, you can really just get a virtual AI note taker and then even summarizing what was said in a certain predefined structure or trained structure. So I think that’s already a value. What else are you using it for?

Romy von Roeder:
Yeah, I mean my daily, my assistant, basically my personal assistant is chat and I ask it her everything we talk about. How do I structure a workshop? How would you define certain processes? How do you define performance? Tell me more about this. Also telling me about specifics. And it’s really good in pre structuring things. And then I think in the end it’s still necessary that there’s someone sitting in front of it and being able to do the fine work and making sure that what it says is really valuable.

Thomas Kohler:
In case you like my show, please subscribe. I would really appreciate it. So on your slides, what I have in front of me, you also had a quote, like 80% of managers expect generative AI to lead to significant changes in their companies in the next three years. What do you think will change in the next three years through generative AI?

Romy von Roeder:
Yeah, I mean, that’s a good question. By the way, when I started this presentation, I said I still, like everyone else, have more questions than answers to this whole topic. And I definitely don’t know more than everyone else probably, but I have certain assumptions. And what we all believe is that we can do the same amount of work at some point in a shorter period of time. So we will be able to, for example, search in our archives for content, and then the AI can probably put presentations together for us with that content so we don’t have to build everything over and over again from scratch. Searching will be faster and easier connecting with people.

Thomas Kohler:
Yeah, I totally agree because I think then at some point a lot of unnecessary or repetitive work, that it’s always the same, maybe doesn’t have to be done by a human or with a lot of annoyance. Right. So do you also think that ultimately discussions like maybe a four days workweek or doing less with more can then be also enabled at scale, that it’s not something pro or contra in doing it or not doing it in terms of an opinion? It of course can still be an opinion, but that this then could be really implemented without having, I would say, a lot of impact on a company’s performance or the economy.

Romy von Roeder:
Well, I mean, very good question. I’m not sure if everyone is faster, then the competition is faster. Everyone can multiply work in that scale. If then you still need to do a lot of work in like five days of a week to keep up with what the others are doing. Right. So the pressure that we are putting on each other, I think that will probably remain the same. Right. It’s not like that.

We then are going to lay down for half of the week and say, oh, well, it’s great, the job is done now. We don’t need to do anything. It’s like we send letters by mail four days before and now we are writing emails, but everything just got much faster. Right. So I think the culture will adjust. I would say four day work week is still an interesting model and we should check it out because the pace that we currently have, as we keep hearing, is not, not super healthy for everyone. So from that sense, I think it makes sense to start thinking about different work models and also probably being more flexible in terms of different.

Thomas Kohler:
And individualized because I think maybe some people, they want to work maybe six days.

Romy von Roeder:
Exactly, exactly. It also depends on, yeah, no.

Thomas Kohler:
And then they also can do more, maybe, or achieve more. Right. Because sometimes, and what I also see with people working a lot that a lot of things are maybe admin related, but if that is going away or being reduced, you can get way more to way more outcomes. And I even believe that there will be a lot of smaller companies with, I would say, way more independence to skill sets. In the past, for instance, engineering, everybody needed an engineering team and raised money for it. Now maybe you don’t need that anymore and you can build a one man woman company and just run it.

Romy von Roeder:
Absolutely, yeah. Very interesting.

Thomas Kohler:
20% believe that their company is future proof. So 22%, based on your talk, that means that a lot of companies don’t think that their company is future proof. Do you also know or have some insights on what are the insecurities or what is the fear of people within companies that they are not future proof?

Romy von Roeder:
Yeah, I mean, I think I have proof and I think I can talk about my own experience. I can 100% agree that there is a lot to do, and that was the core of my presentation previously, that it’s not about fancy tools or like implementing everything that we come across. Right. Yeah. Just focusing on the new things, it’s very easy. Things that you have to fix beforehand. For example, the data accuracy. So most of the data, AI, data startups in HR that I’m talking to, they say the biggest problem that they have is that the data is not correct.

So when they want to work with the data, what comes out is basically not valuable because of the data that has been put in. So, you know, the old saying should in, should out. So what we have to do first is we have to clean up all our data. And these startups also start or have a program to help companies to clean up the data to provide real value. Otherwise companies would probably not decide to use the tool because they think, oh, yeah, it doesn’t work either. Right? Yeah. So that’s one of the things that needs to be fixed early on. And then we have lots of other risks.

Right. We don’t know what is awaiting us. Is AI really mature enough? Do we know what the AI does with our data? Ethical consideration, especially in recruiting. When I talk about your area of expertise, is it really neutral in terms of diversity or has it been trained in a certain way that it’s not? And all these considerations. Right. So yeah, you have to.

Thomas Kohler:
The same problems, what we have now already, they will stay. Right. If you don’t fundamentally change something, because I think also a recruiting team, or a people and culture team or organization team. They also cannot really do proper work, or a finance team. If the, the data foundations is, if the data foundation is not right, that you can really forecast accurately, for instance, or have a certain level of operational excellence in what you deliver, because if you cannot measure properly what the levers are, you can also not have or trigger the levers. And that’s also a problem, of course. And in terms of data cleaning, do you see certain areas that are specifically, I would say, affected, like payroll or recruiting. And what problems do you see mostly when an organization is not ready? In terms of clean data?

Romy von Roeder:
Yeah, I think it’s less payroll, the core data that is sometimes even stored by finance departments. That’s usually the correct data. But also there, you can see that there’s a tiny mistake in the name or a tiny mistake in a number, or you have different number formats when it comes to the birth date, stuff like that. So it’s very easy things. And that usually happens from my experience, when, I mean startup world, hyperscaling, people start working and they do whatever. Right. Data must go into the system. No one cares how.

And then after a while you think like, oh, it looks very different, we have to clean up, and then you clean it up. And then maybe people leave and new people join and then they start typing stuff in, and then it’s, again, in a different way because you haven’t set rules. Right. Yeah. So you basically have to be very, very precise what has to go into the system. O and that starts with Excel. Right. When you start working with Excel, in the beginning, you have to define what goes in and how does it go into the system. And then…

Thomas Kohler:
A classical problem is definitely also in the people field. How do you define the organization also? In what abstraction layer. Right. What is a department versus a team versus something else, no?

Romy von Roeder:
Exactly, exactly. I’m doing exactly this process right now with an organization because we want to compare salary data. And of course you want to compare through levels and you want to benchmark through departments or within departments. But in the current system, there’s no departments, no teams, nothing defined. So you just have the organization and maybe the legal entities and that’s it. So that’s very interesting in terms of, well, how do we work with the data now?

Thomas Kohler:
Yeah, definitely. And I think we also have seen that in various stages. Right. And how complex it is to solve that. Right. Because it’s just not a single problem, what you solve isolated. It’s usually a problem to be solved with a lot of interdependencies and from the outside view, it’s so simple to say, okay, just enter the correct data and we have everything together. But then systems need to talk with each other.

As you said, people have maybe no training or there is no systematic approach set up predefined on how it should be, right? So a lot of touch points and perspectives from where you can tackle that. And you also said that 47% feel sufficiently informed about the potential of AI. So 50 50. Some think they see the potential, some don’t. I think that a lot of it goes through education and also the right change management ability of an organization. So did you already see some examples where it was set up properly in an organization on how to deal with new technologies or AI versus where companies struggled with it?

Romy von Roeder:
I unfortunately haven’t. So that’s the other thing that I’ve seen, that companies don’t have a proper tool and tech infrastructure strategy where you define what’s the core of our organization in terms of tech, because it’s not about AI, right? It’s the digital structure of the organization, right? What is enabled by tech and then that is probably supported by AI and whatnot. What you usually see is that it’s a totally fragmented structure. Everyone is just setting up their own tools, so what they need in this moment, they just buy. And then, I mean, the worst things I’ve seen is that you have several tools in one company doing the same thing, so it’s completely not connected. And again, that’s the same like with the data quality. You have to have your tech infrastructure in place and have a plan on what you want and need to build in the future. The special thing here is that AI is in every field of action of the organization.

So it’s also in your product, but it’s also in the supporting areas, it’s everywhere. So you would probably have to assign someone, a team to go through the whole organization and see how it all fits together, like a digitalisation strategy. And the other thing that you’ve mentioned I think is a core topic, which is educating the people. I mean, yesterday I asked my mom something about how do I clean the wall of my house? And she said, you have to ask AI. And she’s turning 81 this week. So I think the 47% is maybe not even right anymore because everyone now knows something about it. And most of the people have been working with AI probably over the past couple of weeks and months, but there’s still a lot to do, right? Can I use chat GPT in my daily work, or do we have different AI’s that we are using. What are the risks that we need to consider? Where do I have to be careful? How can I use it in the most efficient way for my work? How do I set prompts? And there’s so much to learn right now that I think a structured, as you said, change process and education process for the organization is absolutely crucial.

And the fantastic thing about this is HR has always been like far behind everyone else, right? So we were always running behind the organization because we were always understaffed and like, ah, yeah, okay, you need to do this and that. And we were like, yeah, okay, we’re gonna do this and that, and. But when we don’t know. Yeah, and this time it’s new for everyone. Everyone is on the same level right now. And here HR can take the lead and drive this process. I think that’s a huge opportunity.

Thomas Kohler:
And I think also, especially with Germany, I think Germany has a very much HR admin heavy culture, whereas if you go to other countries, it’s maybe a bit more strategic already and also maybe more separated. So I think even there, Germany could have a big advantage in leveraging that legacy and making the best out of it. And when you look at the hype cycle slide, what you had shown there, where do you see can be? Where can HR tech be leveraged most with generative AI? Because I think there is a slide where you have innovation trigger that’s in the beginning, then you have a peak of inflated expectations. Then you go through disillusionment, illusionment, then slope of enlightenment and a plateau of productivity where it’s really implemented. Implemented. So do you have some examples where you think technology is now really hyped?

Romy von Roeder:
Yeah. So let me explain that quickly. Maybe the hype cycle, maybe not everyone knows what it is. So if you haven’t heard about it, it’s a methodology to understand how a technology will develop over time. And technology basically goes always through the same cycle. And it’s been developed by Gartner that I’m referring to here right now and on this slide, or in the cycle, generative AI and HR is at the top of the cycle. And what the cycle tells us basically is that you basically, usually you overestimate a technology when it’s been introduced, then very often we are very unsatisfied with what we see, and then we underestimate the technology moving forward, which means that we should probably not tactically buy everything right now that is presented to us and then be disappointed because it’s not delivering what we were expecting, but being informed, being part of a community, talking to others, seeing what they would introduce, what they are using. Right.

So I’m not using many other tools than chat GPT right now because it’s the common tool to use. And then you have a lot of tools like whatever, greenhouse and stuff. They’re introducing their own AI technology and checking that out and understanding what it’s doing and what it can do makes total sense before wildly buying stuff without understanding what I’m really buying there.

Thomas Kohler:
In case you have any feedback or anything you want to share with me, please send me an email on thomas@pplwise.com or hit me up on LinkedIn. And in case you really enjoy the show, please subscribe. I would really appreciate it. And do you have a wish for AI? If you would live in an ideal world, maybe in five to ten years on what technology can do to us.

Romy von Roeder:
Definitely it would take my calls and read my emails and then it would do my work during the day and then I have 1 hour per day where I catch up with what it has done and then I say, okay, great work. Thank you. Send out please. And then the work goes out to the customers and everyone is happy and the rest of the day I can lay on the beach. It’s probably not that, but you asked me if I could have. That would be the wish. Yeah, I think we, you said it yourself, right? We are definitely going into that direction, right. That the AI is doing quite some stuff for us, which is interesting.

Thomas Kohler:
Thanks Romy. That was it. And always a pleasure talking to you and have a great week.

Romy von Roeder:
Thank you so much. Same to.