Episode 75: Using Data, Analytics and AI in Talent Management (Interview with Toon van der Veer)
In research we have conducted at Insight222, since starting the business back in 2017, we found that the vast majority of HR professionals, 82%, strongly believe that people analytics drives business value. My guest on this week's episode of the podcast, Toon Van der Veer, is the Global Vice President of People Continuity at AB InBev and a strong proponent of using data, analytics and artificial intelligence in people management. As Toon explains to me in our conversation, data provides HR professionals with countless opportunities. And this is certainly proving to be the case with the innovative work he is leading at AB InBev.
You can listen to this week’s episode below, or by using your podcast app of choice, just click the corresponding image to get access via the podcast website here.
In our conversation, Toon and I discuss:
How AB InBev brings data to talent conversations
The importance of focusing on adoption and change management
Why a "You spoke. We listened. We acted." approach is a critical element of employee listening
How companies should approach the measurement and utilisation of skills data.
This episode is a must listen for anyone interested or involved in people analytics, employee experience, and HR tech, so that is CHROs and anyone in a people analytics, workforce planning, or HR business partner role.
Support for this podcast is brought to you by AG5. To learn more, visit www.ag5.com.
Interview Transcript
David Green: Today, I am delighted to welcome Toon van der Veer, Global Vice President for People Continuity at AB InBev, to the Digital HR Leaders podcast. Welcome to the show Toon, it is fantastic to have you on here, we have known each other for a while. Can you provide listeners with a brief introduction to you and your role at AB InBev?
Toon van der Veer: Yeah, of course. And super nice to see you again, David. I have worked 14 years with AB InBev. I have worked in sales and general management. I have worked in technology in our shared service centres and I have worked in people. Prior to taking this role, I was the VicE President of People in Europe, and now I take care of Global Vice President of People Continuity. The job title by itself is maybe somewhat confusing. So what are the four things that we do? It is everything that related to talent management, learning and culture, diversity and inclusion, employee engagement and attraction. So these are the 5 products that we as a team take care of and that we work on.
David Green: Well, I am really looking forward to our discussion today, especially the work that you are doing around data-driven decision making at AB InBev. Let's start by hearing your views around decision making and HR currently, and the role that data and AI has to play or is increasingly playing.
Toon van der Veer: Yeah, absolutely. I think that the first time we met, we started to explore that a little bit. I think that data in HR gives so many opportunities for us as, HR professionals. The opportunities are in terms of making un-bias decisions, empowering our line managers to understand themselves, their teams, their organisations, even better.
It really enables HR leaders to become even more strategic. The famous "you need to be a strategic business partner” I think that data by itself enables that because you get out of the day-to-day machine work, that is super important. But you are really starting to create insights and a different lens of looking at your organisation. And I think that that AI now is just one of the vehicles that makes understandable blocks of products out of all the different data points that we select. I think in HR, we collect so many different data points, whether that is from your HR master data or the cycles that you run into. There is so much data we collect and I think now it is about making sure that you start to combine these blocks into meaningful and insightful conclusions and points of reference.
David Green: Yeah, It is interesting because we have always collected data in HR, but we haven't necessarily done much with it. And I think you gave a perfect intro there, help empower managers to make decisions about their teams, but also help them understand things about themselves and how their behaviours maybe have a negative or a positive impact on their team's performance and their team's engagement. It is also so important as a manager, particularly, now for a lot of managers, suddenly they are having to manage a team that is wholly virtual, which is challenging to be honest. If data can help guide them, to be a better manager to help their teams also come through this. And then obviously that helps them and their teams, but it also helps the organisation.
Toon van der Veer: A hundred percent.
David Green: Can you tell us about a particular project that you have been working on at AB InBev, that is optimising AI for decision-making?
Toon van der Veer: Yeah, I think first of all, David, what we have tried to do in our data and analytics journey is to kind of set a north star, so we have a direction and we are all clear that we are going to build this. I think that that has allowed us over the last kind of three/four years when we really started to understand and put dedicated resources against this, to see where this was going to meander.
But that was by design, it was not by coincidence we empowered the teams and really we said, okay, let's see where we can go. But there were three things that we set from the outset. So first of all, how can we bring more data into our talent conversations? So I think we have all been there. And talking to your listeners, that, moment when somebody says this person is great. Okay, that makes sense. And we all learned to ask, for an example and all that, but you now actually have data that can help you recommend or challenge that. So, that is what we started with from the outset we call that OPR10X. OPR is our cycle that we do every year to assess the potential of our teams.
We really start to look at combining different data sources from line manager effectiveness in terms of our engagement survey to their leadership capabilities. To their business results, to their turnover, all these types of stats and how they stack against each other. We looked at peer groups and where they sit in mediums, you will hear I am very passionate about that. But that really enabled us to start to have deeper and more meaningful conversations about making sure that we put the right people in the right role. And what I mean by that is not just promotions, but also understanding in terms of what are people's sweet spots, what are their strengths. What are they amazing at? Or are they more of a generalist? How do they stack against what are their development needs, etc? So again, I think that is the part that I am the most proud of and that product by itself created other products that came after it. But we can talk about that a little bit later.
David Green: You said that over the three to four years, there were three things. One was to bring more data to talent conversations. What was some of the feedback you got from managers? And also what was some of the feedback you got from employees.
Toon van der Veer: That is a great question. I think of it as a change journey. So the first time we tried it, I think people were like, so why is this relevant? Why are you wasting your time on this?
That is a fair provocation because I think in the first iteration, your AI may get smarter, you learn and you get things wrong, right? And I think we have been again from the beginning, quite open about that and have said, this is a long-term journey where we want to start to embed more data key decisions we take within our people department and within HR in general. So the first question was, why is this? Then you had your promoters and your detractors. The detractors are like look, I know my team, you don't need to tell me the data points. The promoters were like, Hey, this gave me a different lens of looking at my team. And Hey, I didn't actually know how person X, which I don't see so much, was performing on a higher scale than everybody else in my team that I have more proximity to it.
So that was how it started but then as it evolved, it became more credible, if that makes sense because we used it, we improved it, and the output just became so much better. So we now have a product which is being used in our cycles that people really look at.
They want to understand how they can use that to take better decisions and I am proud of that. We still do our measures we do our MPS. We use our adoptions to make sure that we keep improving it, but I think it is now really something that has an adoption and is something that became a part of our normal routine.
David Green: I think you have taken a very sensible approach at AB InBev to pilot, learn, get feedback, and then improve and iterate because that is what you do elsewhere in your organisation, so why wouldn't we try and do that in HR as well?
Toon van der Veer: That is something that we, and I think that helped from working sometimes cross-functionally, if you get a bit more experiences in the tech world. I think that that whole method of trial test feedback improve, not being afraid to launch a better product, not being afraid that it actually breaks or, that people say this is useless and you sunset it. That is something that typically when I talk to HR professionals, they are reluctant to go there because people say they don't know what they are doing or whatever. But I think it is something that you need to take the first steps. Find the right scope where you start your pilots. If you are just going to roll out a product and you are going to do this with a bunch of detractors, you are not going to learn a lot because people will by default say to them you need to find the right balance to make sure that people will invested to make your product better.
But it starts by solving a pain point. And the pain point that we solved for with the product I just mentioned was people felt that our conversation at times, lacked data. So they wanted those data points. And that is where we started from how can we add value there from all the data points that we collect, okay, we put all of those down on a page how can this help to have an even more meaningful discussion?
David Green: And of course in many cases, the data will confirm what people think confirm that hypotheses which is great because at least you have got that validation. But as you said, on many other occasions, of course, the data will help teach something new, or perhaps give another perspective or add something that hasn't been thought of. If those conversations were happening previously and there wasn't any data there, then they are essentially based on opinions.
Toon van der Veer: And to build on your point there, David, when we started our accuracy was 50/50, you might as well have flipped a coin. We are now at a 92% accuracy. Our AIML has now got the ability on past behaviour, on predicting what the outcome is going to be. So it helps managers, it helps the conversations and it also starts to help the conversation if you don't follow. Why do you disagree with the proposed score? What is the reason why you think that this person ABC and D? Then you get into great, meaningful conversations.
Like first thing had a bad year or Covid really threw them off or whatever but you can get to the core of the conversations. And you spend less time on things that are rather obvious. So, you are a hundred percent right. And that journey is going from 50/50 to 92% accuracy and that has been super, super interesting.
David Green: I guess as the machine learns you get more data and ingest more data, obviously, you are iterating on an ongoing basis. You get that improvement from starting at 50/50 to 92%. What are some of the sources of data that the platform uses? And how have those data sources not just improved, but also increased perhaps the range of data sources has increased since you started to run this program?
Toon van der Veer: Yes, and without going too technical, because I am not that technical, but I think big game changer for us was actually when we started to create a proper data link. In the past, we had different stacks of data and just simply there was no link. Of course, you can connect different Excel data sources etc but the effort is so high in terms of the outputs.
And then once you did have the connections you didn't have the ability to put the AI element in that. So I think there were three parts if I'm very honest. Step number one was creating a data lake. So really starting to break one pool of information where we start to streamline with all the GDPR compliance. Of course, we are not in some world where we create our own rules, but where you have your core HR data, your engagement data and your 360 data. We are now looking at workplace analytics data so you really start to create a bit of a brain if that makes sense. With all the different data points, meta data, etc, that is the first step. Then the second step comes down to I don't believe that you can, and I mean this respectfully, learn about AIML, but we really made sure to hire some amazing people. So, we have a global analytics capability centre in Bangalore. If I talk specifically about Europe, we had a person in my team who was a data scientist, studied mathematics, and really helped there.
So the second point was to get that input in, in terms of how do you do this, rather than trying to tinker at something. The third point was really to not get too specific on what it is that you wanted as an output. The reason is okay, we want to solve this problem. How are we going to do that? So you then empower the team who then came with solutions, better problems, and great questions in terms of requirements. And the product evolved so it rolled down, sometimes it went left, sometimes it went right. Sometimes the performance was bad and we had to move data to the cloud. All sorts of things that happened with that.
But I think that these are the three points that we really used. So the data lake to your question, where you combine as many data sources as you can. Second part, bring in people who know the spectrum really, really well. And third, not be too prescriptive because honestly speaking, we are still exploring.
I don't know how far AIML, how the data, how all the insights are going to help us make better decisions in the future. So I think if you are too prescriptive, you might miss pieces of information that we otherwise would not have had.
David Green: But I think the really important part that you said there it is about solving problems, it is not just having a tool for the sake of having a tool. What are the challenges that we are facing as an organisation? What are the business challenges we are facing and how can this unlock some of the people elements of that. As you said, having that data together is clearly one way of doing that because you need to blend those different data sources together and be able to do that at scale. And not every time you have got a question, have to be like, okay, I am going to get data sources 1, 2, 3, and 5, and bring them together. Do all the analysis etc, you have got that access to draw upon, and obviously, you have the skills to actually to do that analysis in the first place. I have been to enough conferences over the years and people still say, don't start with the data, start with the business problem.
That is probably the first rule of people analytics, I would say, or anything that we are doing using technology in HR. If it's not a business problem, don't do it. I would be interested to find out how you source those business questions.
Is it something you work directly with the business on? What I noticed from your introduction is you are not just a career HR professional Toon, you have worked in sales and marketing and in technology areas as well, I guess that probably helps?
Toon van der Veer: It does. I think there is, again, I like to talk in threes so I hope that it is not boring. But I think there is three parts there that I think helped a lot. First of all, is having a natural curiosity outside of people. So, read and ask questions, ask your strategy team or your head of finance or whoever leads the organisation. Ask them about, Hey, what are your three big questions you have to ask? And this is not a bad thing to ask questions. You are going to understand better what it is they are looking for. And I think it helps us in HR to translate that into meaningful actions.
That is the first thing. I think the second one is to subscribe to podcasts, like this one, or enrol in the work that you guys do with your organisation, David, in terms of understanding what is there, appreciating what is going on, and taking those first steps to learn.
I think the third part is to, and again, I mean, this respectfully, not be afraid to have a vision or a strategy for HR. I think that the function is, and will become even more strategic in the years ahead. And that is not just because I worked in HR and I worked in sales and I did the different things but I think that more and more technology will help us to evolve with regard to the bread and butter or the core HR part will start to change and transform.
Technology will help us do that more efficiently, so the role can really elevate itself in terms of insights, recruitment, looking ahead. Where are pools of talent? How do you facilitate certain conversation? And I think just challenge yourself, close your eyes.
There is plenty of content out there and think five years from now, or ten years from now, what does the world look like? And what is the role that HR plays in an organisation? And if you want to get really philosophical in society to drive things forward. One thing that I try to challenge my team on is to look at consumers and employees as two different animals. How you interact with your bank and then you walk into an office, all of a sudden, the joys of doing expense notes and, badge reports, all this stuff, you don't do that in your day to day life. And that is going to start to blend because our future colleagues and our current colleagues, they grow up in an environment or an ecosystem where it is so frictionless. I think that that is going to be a massive opportunity and challenge for HR in general.
David Green: Yes, I love that, that natural curiosity. Speak to people in strategy, speak to business leaders, speak to finance about what that challenge is they are facing, because, we don't want to operate in a little bubble in a silo in HR. I love that kind of outside in, approach, learn from what others are doing, you don't necessarily want to replicate it because it is your own challenges in your own organisation, but you can definitely learn from the outside. And as you said, be brave, have a vision and strategy of what you are trying to achieve.
We had Leena Nair on the podcast about 18 months ago, and she said HR needs to have more swagger which is a lovely way of describing it. I think it is very important that HR is the business as well. I think we need to recognise that those of us that work in this field and listen to this podcast which hopefully isn't just my mum. We will find out, one day when we look at the listener figures.
Toon, what I am quite interested in is with the platform that you are talking about and using the technology, what is the balance that you have struck between using the algorithm and human decision-making?
Toon van der Veer: I think taking the balance between the human interaction and the algorithm for me, the algorithms is a support mechanism. It should bring insights that you, as a human, connecting so many dots that you would not be able to logically compute. But my firm belief is that AIML will not replace human interactions. So all of the products that we build help make better decisions, but those decisions will be taken by leaders because AI is helping catch a lot of things, whereas somebody's interpersonal behaviour doesn't catch that. So I think it helps make better decisions, it helps a lot to make faster decisions un-biased decisions. But I think in the end, it comes down to having honest conversations with leaders amongst each other, amongst employees about empowering with data.
I think that insight on getting those nudges on, actually for me, I love working on a Sunday afternoon. I put on some music and I prepare for the week. But as I send emails, I don't expect my team or the rest of the organisation to open their laptops on Sunday afternoon, they can do whatever they want. If they do great, if they don't that is also great. But I think that these type of nudges can help in terms of understanding how you lead. When was the last time you spoke to this person? Do you pivot more to introverts or extroverts? So to have these types of insights, I think AI can help. And I think we can really create a more meaningful workplace where both colleagues, employees, and leaders are able to work fair together, if that makes sense.
So, I think that is it. I don't believe in letting the AIML spit out things and we are just going to run with that. I think that that is too far out.
David Green: Yeah, I agree. It is there to help basically guide decisions. And as you said by putting a lot of information together, it can help make quicker decisions, or maybe hopefully more informed ones as well. I love that analogy around using it to understand with nudges effectively saying, the data suggests that if you do this, this will likely happen. If you do this, this will probably happen. And then at least it helps you to make the decision around that.
So you told us about how the accuracy has improved, what are the practical steps that the factory has been doing to actually help to do that? The question that we get all the time, how do you improve data quality?
Toon van der Veer: Again, with three points. First, you have to make data the priority and making sure it is clean.
We went on this journey in our organisation, that when we started to look at data, how do we extract value out of this? It was like I was there swimming in a pool of gold. It was a mess. Every organisation. Start dates were there, but they were wrong. Line managers need to take data reporting and make it important. It isn’t AI but it is where you take data reporting and it helps to say, Hey, your accuracy is 60%. Then as an organisation, you can decide that this is important and then start the journey to clean your data. And again, that takes time. I have been blessed to work with people in Europe and globally who actually started that before I took roles.
So again, we made it important as a function, as a people function. We call HR people in ABI to just say, we need clean data to make better decision. So that is the first step. There is no way you are going to do this by asking the people team or the HR team to clean data for you, because they will refuse to do that at a certain point.
That is not why human beings are put on the earth. I think the second part that we do every month when we do releases. So we work agile we really try to empower the teams. And we have great people. We do the retros, we learn, we look at performance KPIs, we look at the adoption, we look at the net promoter score. But we also have data scientists who help us make sure that the algorithm improves.
So you'll have the data that came out. What are the things that we saw were outliers based on the, input you then get from managers in the process? And then what can we learn from there? So we challenged our AIML to be un bias. Again, one of the barbecue stories you always have when you work in HR and you talk about AI, first thing is okay, but what if your AI is biased? Yeah. Okay. So you need to make sure that mechanisms are in place that check. That is not very complicated. The third part with regards to how you do this is start small. This didn't start with the full global organisation. Take a scope of your organisation, understand, and learn because if you roll a big bang with everybody involved and you get it wrong, it is no problem. But you will have to do so many steps back to then get the change management moving. Again, I think it makes more sense to go smaller, learn, adapt, learn, adapt, learn, adapt. Then once you feel that the product is ready enough, then you can start to scale it. Because I think then you have got the child diseases out of it, you have got your insights and you feel yourself also more comfortable with what you need.
David Green: I guess what you do with that is if you are communicating that this is something that is being piloted in a part of the organisation, you create a buzz or maybe a desire to actually be the next business unit or country that gets on board.
I have certainly seen that happen quite a lot. And I think, again, that is another way that we can learn from marketing in HR and people is how we promote our products externally to our customers. How can we promote our products as HR, to our employees, our workforce, our managers, and our leaders.
Toon van der Veer: A hundred percent. Two practical tips that we try to apply here is first of all, we have this notion of you spoke, we listened or you spoke, we act which what it starts with. Again, asking questions and sometimes we don't ask the question because we know that the answer is going to be bad, right? You have a product that everybody complains to you about wherever you are. But then we don't make it factual because the moment you make it is factual, you can say, Hey look, we took these things to improve it and now this thing is moving. So I think the first part is making sure that you make it factual, if you ask the question, did you get that data point? And then you can communicate back like, you told us that it was not good, these are the three things that we have done. Look, now people tell us it is better. It is not a hundred percent yet, but we are on a journey. That is the first thing. The second thing, what I learned here from the team in the US, is really spending additional time on change management. It is not something that is the easy, it is a skill. But I think it makes a lot of sense to explain in simple terminology why you are doing things.
I am sure there are tons of people who are listening to the podcast, who will do this much better than us. But it is something not to neglect because it is exactly what you said. Hey, this is solving our problem. Hey, this is in line with our strategy. I think it is important to explain.
David Green: Yeah. And it is a different, way of working, isn't it?
Like you said it is change Management it is changing the way you work, which can be quite fundamental for people and for managers. We talked a little bit around feedback from the managers and stuff like that, but, were there any challenges with managers, and are there any success metrics that you can share? Obviously beyond the one you have mentioned around the accuracy.
Toon van der Veer: Yeah, so we look at adoption. I have tons of ideas but in the end it is whether people use it. It is the same as brand innovations. You can put them out there, but if nobody buys them, then you did something and you need to adapt. So when we really look at adoption, which is are people using your product within the function that can be outside of the function. But that is the first thing that we look at. I think the second thing that we look at is performance. We sometimes don't talk enough about that. Performance is boring, it is process optimisation. Basically your Kaizen but using data to do it for you instead of a whiteboard with post-its. I think we sometimes don't make it important enough. Whereas if you look at larger tech companies, this is their bread and butter. They look at the old times and yet again, they set the bar for you, me and whoever is building these things.
If your Internal search system is not as fast as Google, your users i.e., your employees are going to say, this thing is slow. So I think you need to make performance important as the second layer that we look at. Then the third part is we use NPS or CSAT or whatever. I don't think it matters I think KPI wise, it is agnostic.
I think the key thing is to ask for feedback. Ask people what would make this product even better? Or the famous question, would you recommend this and this AIML to your family and friends I think is a very weird question. Because I think there are two people in the world who come home from work on a Wednesday night and say, you know, today I was at work, I had this AIML and wow, I really recommend you to look at it. I think that is not how human beings behave. Some do, maybe, but I think you need to make sure that you ask the right question to get the right input. How did this help you make better decisions? How are these connected to your business strategy?
I think then you get the insights that you need rather than, would you recommend this and this tool to your family or friends? I don't know.
David Green: It is slightly more nuanced than that isn't it. What are we doing? Well, what could we do better with this product? I guess the thing about NPS, and I know there is a debate on whether you should or shouldn't use it. What I think it does is resonates with people because no doubt you use it on the customer side. So it resonates with leaders. They understand it. You haven't got to go through a whole kind of conversation around why you are using a particular metrics people understand it.
And then they can get excited about it. They want to understand, how is that tracking on a monthly or quarterly basis? As long as you are getting the insights that you need from it to improve the product. And as you said, telling people that you have listened and you have acted, then, ultimately the product is improving and the user experience is improving.
Toon van der Veer: A hundred percent and David to pick on a word you mentioned, it is about improving. So measuring NPS as a number on a slide, you might as well not do it, but it is about that continuous improvement part. I think you mentioned it a couple of times there. I think that is key. I mean, to say that you measure NPS fine, very nice. But how are you doing this to then do the interviews, do the work, replaying your backlog. That is, for me, the work that really adds the value because that creates credibility in how people believe you are managing your product.
David Green: Completely agree.
Where are you thinking of taking it next?
Toon van der Veer: I think I will practice what I preach. So we are going to continue to look at performance. We are going to measure NPS we are going to continue to make this better. But in the background of this product, new things started to arise, for example, our succession tool. We are excited about that.
It is starting to recommend internal successors for particular roles. It helps with diversity of thought lists. This helps with understanding skills. This also starts to help in terms of, Hey, I have gaps here because the way we do succession plans and I haven't come across one company or a tool that nails succession planning a hundred percent. So I think that that is one that I am really excited about. We are kind of in MVP stages there, but when I see what the team is building I am very excited about. I think the second part is we are starting to move more in terms of how are we going to make this available to employees.
So today, it is more on a line manager level and people team level, but how do you really, again, a fashionable term, but how do you really democratise the data? How do you get a single suite where people can get their content, which is going to help them to improve. I think that these are the two parts that I am most excited about.
The two legs that I really started to do. And whilst we add more data sources to the data lake, we have always an explorer leg with the team where they look at the benchmark. We had a session this morning with the team in Argentina and India where we are really starting to think of a community of practices for people to start to share ideas.
Because again, I don't believe in a top-down approach for analytics. I think it makes sense to do a data lake top down so that you have that degree of consistency. But how people use it where, I don't care, but then really to start to share and potentially converge on larger projects or larger products that are being built, I think is very exciting.
So again, I think these would be the three things.
David Green: It sounds exciting.
Toon van der Veer: Very exciting. I think the field is super exciting. I think we are really just starting to learn, we collect so much data, but I think we are still starting to understand. And at the same time, we are going to get a couple of things super, super wrong in terms of, we might be thinking we are solving a business problem and we are not. I am very excited about where this is going to bring us and I am super proud of the team that builds that because they are the ones that really come with the ideas and the insights and they are empowered. I think that that fits very well in our company culture of ownership where people have that empowerment to come with ideas, solutions and, solve business problems.
David Green: And that leads us quite nicely to the next part of the conversation. You have mentioned agile a few times. It is a bit of a buzz word at times, I am sure you agree with me on that. But saying that, I think it is fair to say that the more progressive HR organisations are, are actually taking a more agile approach.
As you said, rather than trying to roll out a whole program for the whole company in one go, instead lets try something and see if it works. If it doesn't work, okay, we move on. We learn from it and do something else. Or as you said, you learn, you iterate, you continue to improve it on that basis as well.
How important do you think an agile approach is in HR?
Toon van der Veer: Agile is a means to deliver results. And I think it is just a method. So I tend to agree that you need to work agile, we have to train a thousand people on agile, those sorts of things. I am not particularly impressed with those things. I think really, where does agile add value? Where does Kaizen add value? I think it is a tool in a toolbox to be very honest with you, David. Where I have seen it work and it has practical examples, but to stay high level enough, we use agile as well in business as usual. So, we had a payroll challenge in one of our countries, but what we saw there and where I think it can have tremendous value is when you have a matrix organisation or you have a cross functional, problem, or you have a, a challenge where you have multiple locations in the world or in locations working together. For me, this is where agile creates so much excitement. So nobody gets excited about, oh okay lets go agile, but people get excited about solving a common problem, working better together, communicating, being empowered to solve problems. I think if I go to the core of what agile is, and agile is not about training or what a scrum is or any of these things.
What it really is, is that you understand how that can help you work better together as a team. And sometimes, what I said before, is it can be just like a waterfall if you know, where you are going and it is just about speed, it is very clear, let's go waterfall every week, check-ins, stay close, that sort of thing. But if you have to connect different people, different rooms, different functions, and the organisational design is opaque, it is not super clear.
I think agile can help a lot. If people get their egos out of the room and focus on the business problem, that is there at hand. I will be very honest with you, it was tough for me. Because when we started the change.
Actually your role is not to say anything, which is almost counter-intuitive as sometimes when you get trained as a leader, you have to give directions, but actually it is the other way around. You have to be clear on where you want to go north star wise, but then it is about the team, the backlog, and all of that. But their results are impressive. Before we used a system, we did every quarter releases, it was slow. Every time there was a fight on that. And when I left the zone, the team was running it at a hundred percent. We were doing weekly releases, performance was up, NPS was up, stability was up, functionalities kept improving, but I had no idea what was going on.
And I think that that was super cool because for me, when I spoke to the guys who helped us with the implementation, it was almost for me the ultimate goal, because that is what you want as an organisation. So anyways, I think it is a tool in a toolbox. I think it can help a lot if you have a complicated business problem that has multiple stakeholders, but it requires leaders to take a step back and you are going to go through the change curve, but you have to trust the team in doing it.
David Green: You talk about taking a step back as a leader. What other advice would you give to some of your peers that are listening or to other organisations that are looking to use agile more in HR?.
Toon van der Veer: I think look at where it can work. To run agile because you want to put it on your LinkedIn profile or on your CV that you are a scrum master, I would stay away from that. You are not going to help anybody and especially not yourself. So let me start with that statement there.
But I think where it can really help is again, start small. Understand the real problem and familiarise yourself. The second thing, don't kill it too fast in your heads. I did three trainings before I started to get it because the first time you hear it, you are like, wow, this can maybe work for the tech teams, but man, this is not for us, try to embrace it and roll it. And the third one, really make sure that you can trust your team. If there is somebody in your team who is really excited about agile, but not like the agile from the sideline. But really believes that they can make it better, trust them. Give them 9 months, give them 12 months, give them 18 months to get this thing right. Talk to them, learn from them. That is really how it went for us. There was somebody in my organisation who believed firmly that this was going to make the product better. Okay, I trust you, I will learn from you, I will be there to support them. But she was actually the person who brought the concept to reality. And from that moment, I am really hooked on it, I see that it adds value in multi-layered multi location, more complicated product and process.
David Green: One of the things that really came out of what you said is that it is a technique. It is a methodology to use certain things that you are trying to achieve. But it might be other methodologies you might use to solve other problems. So I think it is a nice way of putting it.
It is the time where we have come to the question that we are asking everyone on this series now, and you have touched on it a little bit earlier when you talked about understanding skills, data and stuff like that. What is the value of measuring skills data and how should companies approach it?
Toon van der Veer: So the question on skills, David, I have been on a learning journey in this role for the last six months with my team here. We are working together with Gloat. It is a startup that I think you guys spoke about a couple of times as well but they help us. I don't know where skills will go. What I do know is that it is going to be super important for us with the challenges that we are going to have in the future. If that is technology, innovation, speed to get a better understanding. Because on a macro level, what do we see? We see employees looking not just for fast career paths they look for mastery. Second thing, is that this whole conversation we have just had four years ago, I would not have been able to have this type of conversation like this if I would not had met people who had deep skills that are really passionate about the topic and willing to share that.
So for business problems, you are going to have some deep rooted skills. So the whole conversation about generalist versus experts, is going to shift, I have no doubt about that. Now, where it sits in an organisation, we are going to learn. We are inspired by what the teams at Unilever has done inspired with what the teams in Kraft have done. I think we are going to understand how this is going to go. And this is also the approach we are taking So we are rolling out gloat now for around 5/6,000 people, which again, is big enough for us to start to learn. But we are really focusing on the talent management, career paths, how does the organisation look? We don't go immediately to the marketplace, because I feel that the marketplace is sometimes too far out. It has so many implications on how you reward, how you set targets, how you promote, how you allocate time we first need to learn a little bit before we go there.
So we are going to take it step by step. I think it is going to be a super exciting chapter, for people and how organisations grasp that it can get quite a crowded place in terms of discussions and, philosophical debate. It took me a bit of time to find what I think is for us as an organisation, but this is going to evolve as we learn together with the team, from Gloat, with ourselves. And when we, go live with the groups that we are going to be using.
David Green: So it could be podcast part 2 in 18 months time.
Toon van der Veer: I would love to take you up on that! So then it is going to be either something that we really learned or it will be something that we said we tried, but it was not for us.
David Green: I think it is quite a fresh approach from the outside, looking in at your organisation, you do take this approach. You do try things and learn from it. Everything that we have talked about today is that approach. But I think you do that a lot with, technology companies and partners. You try with a pilot, start something small using your own words, learn from it. And then if it works for the organisation, then you will look to roll it out.
Toon van der Veer: A hundred percent. And at the same time, there is one important thing that I had to learn from a line manager who had to coach me on this is not to have pilot purgatory. So you have to define almost step by step. What is the success criteria?
Because what sometimes happens this is human behaviour, I can fall in love with your idea, and then at a certain point, you don't give up anymore and you will make it a success. I got challenged with this four years ago this whole test and learn is great but at a certain point, you need to go to scale or you need to sunset it because otherwise you are going to create a mess in your organisation for your want to build on that. It is not that we are running 5,000 pilots here. What we do is once we think something helps and converges towards our business strategy, we are super open to learn and try, but we are also open to do the sunsetting, learn from the retro and understand why it didn't work for us. Because it doesn't mean that two years later, the same product might be extremely relevant and solve the problem. But that is really the bit of the mindset that we are trying to go to. You can diverge, but you have to converge quick enough and then see how quickly you go from pilot, to MVP to scale. I think that that is something that we are still learning but that is something that we apply as well in our team.
David Green: Sadly, we have come to the end of our discussion. I knew I would enjoy that discussion and I hope listeners have as well. Thanks for being a guest on the podcast, Toon. How can listeners stay in touch with you and follow you on social media?
Find out more about your work?
Toon van der Veer: Well, first of all, big thanks and it was great to reconnect, David. I think it is impressive what you and the team are building and it is awesome to see how you are shaping the landscape. So congrats on that. I think people can connect on LinkedIn. I think that that's the easiest way.
Drop me a note or if you have any questions I am happy to help and happy to share whatever, ideas or insights we have. We are not like perfect. I don't think any organisation is, but happy to share any learners, if that helps anybody on their journey.
David Green: Thanks very much. And, hopefully we will meet in New York or wherever in the not too distant future.