Episode 215: How to Navigate Challenges in Skills-Based Transformation Journeys (Interview with Christophe Cabrera)

 
 

Taking the first steps towards becoming a skills-based organisation is exciting—but what does it really take to get started? In this episode of the Digital HR Leaders Podcast, David Green welcomes Christophe Cabrera, Director and Head of IT Talent and Company Reputation at UCB, to share an honest look at the early days of their transformation. 

Unlike the polished success stories we often hear, this conversation is all about the reality of starting out—what works, what doesn’t, and how to build momentum.  

Here’s what you can expect: 

  • The biggest challenges in getting started, from outdated approaches to experimenting with AI 

  • How UCB used a proof-of-concept to build confidence in AI-powered solutions 

  • Managing the complexities of a workforce spread across production, clinical, and office roles 

  • Keeping their skills framework agile as new competencies emerge 

  • Defining what success looks like, with internal mobility and personalised learning leading the way 

This conversation, sponsored by TechWolf, is an honest and practical look at what it takes to get started with skills-based management.  

TechWolf is an AI-powered solution focused on one mission: delivering reliable skills data for every role and every employee in your organisation. 

With TechWolf, companies like HSBC, GSK, IQVIA, Workday, and United Airlines have accelerated time-to-hire by 32%, boosted internal mobility by 42%, and saved around $1,000 per employee annually on talent management. 

Visit techwolf.com for more information.  

[0:00:00] David Green: When we hear success stories of skills-based organisations, they often come from companies at more mature stages of their journey.  Those that have already successfully navigated the challenges of adopting this approach, and who have moved towards scaling skills across the organisation.  However, today we're taking a step back to explore what it actually looks like in the early stages of this transformation.  I'm David Green, and joining me on this episode of the Digital HR Leaders podcast is Christophe Cabrera, Director and Head of IT Talent and Company Reputation at UCB, a leading biopharmaceutical company that is on the journey towards becoming a skills-based organisation.   

What I love about talking to those at the early stages of their skills journey is the practicalities, the real-time challenges and solutions they encounter along the way.  And in this conversation, Christophe offers an honest look at UCB's journey so far, from tackling the complexities of traditional skills identification, to experimenting with AI-powered solutions like TechWolf.  We discuss how UCB's diverse workforce, spanning production sites, clinical roles, and office-based teams, shapes their approach to skills and how they're staying agile in a rapidly evolving skills landscape.  Christophe also shares how UCB have moved beyond a successful pilot and are now rolling out a skills-based approach to more than 50% of UCB employees.  With AI being a massive part of the skills journey, Christophe also shares how they're building stakeholder confidence in AI-driven systems while keeping the human element at the heart of their strategy.  So, with a lot to cover, let's dive in and welcome Christophe to the show. 

Christophe, welcome to the show.  For our listeners, can you start by giving us a little bit of background on UCB and also your role in the organisation? 

[0:02:05] Christophe Cabrera: Hello, David, thank you for the invite.  So, UCB is a global biopharma company, so our headquarter is in Belgium, but we operate globally in around 40 countries.  So, we are about 9,000 employees.  And one specificity of UCB is that we really have people working along the total value chain of pharma, so from research to manufacturing, and of course field force and clinical development.  So, we have a very broad and diverse population.  I mean, UCB focuses to create value and valuable solutions for people that live with chronic disease, neurological and autoimmune conditions.  So, we have a very focused and big commitment towards those population.  And my role at UCB, so I'm part of the global IT organisation.  So, I'm leading the digital technology team for HR and corporate communication.  So, it's a very long title just to explain that I'm basically supporting the HR organisation, I mean me and my team, for everything which is related to technology.  So, from hire to retire, any technology component that they need to, or they want to rely on, we are basically helping them to get to make the right choice and set it up. 

[0:03:33] David Green: Obviously, the topic today we're going to be talking about is around your skills journey at UCB, which I understand that you've recently embarked upon.  What led UCB to start this journey to becoming a skills-based organisation? 

[0:03:48] Christophe Cabrera: I think that's something that has been going on already for many years.  I mean, maybe the word 'skills' was not used back then, but this has been a need for a very long time.  And that need has grown to the next level, I would say, over the last three, four years, mainly for two reasons.  I mean, more and more, the organisation has to do very quickly decisions on people because internal decisions need to be made, so to be able to quickly decide where and how I need to reallocate people and are those people the right ones for the job.  So, that's really something that became, I mean, critical for us over the last three, four years.  I mean, the COVID time, we also had internal decisions, strategic decisions that have been made.  So, yeah, there is really a big, big evolution on that, and we need to be able to make informed decisions.  So, that's really one of the aspects.   

The second one is more the forward-looking aspects of more the long-term plan, strategic capability plan, as maybe some people name it, and to be able to really measure the impact of the effort we are making into those.  So, we had defined at UCB some strategic capabilities that we want to develop.  It's easy to define on paper.  Now, to make it happen and measure that actually the actions that have been defined have the expected impact, and that actually the capability is growing within the organisation, is becoming essential as well, because sometimes we were flying a bit blind, doing a lot of activities, putting a lot of effort and resources, without always being able to really assess and quantify the impact that those had on the UCB employees. 

[0:05:32] David Green: So, in a way, it helps you be more agile in terms of reallocating resources when you need to do that, because let's be honest, the world needs it as organisations become more agile.  And it's helping you understand the gaps that you've maybe got from your future strategic workforce plan, but not just understand the gaps, also understand how that gap potentially is being closed by the various actions that you're taking through this data.  So, it's effectively providing you with the intelligence to operate successfully as an organisation, by the sounds of it.   

[0:06:23] Christophe Cabrera: And also, how to get your internal population to evolve.  So, for example, we've been moving from traditional pharma to biopharma, so we had to move people from one department to the other, let's say, to be very pragmatic.  And of course, to understand what was the gap in terms of skillset, because they had a lot of skills in common, but there was also some skills that were different. Where is it that they are different, and do we have people that have those skills, and what gaps do we really have to close to be able to reallocate those existing resources to where we need them, rather than going out and hiring new people, was also super-important for us to understand.  So, it's indeed gap towards the future, but also internal gaps and current gaps, basically in terms of resource reallocation that we would like to perform. 

[0:06:51] David Green: And it's interesting actually, from some of your colleagues that we've had join from other organisations in the podcast over the last five years, I think, who also are on this skills-based organisation journey.  We hear that this is the thing, I think, which is so captivating about it, the benefit potentially to the individual employees within the organisation in terms of supporting their development, supporting their career mobility within the organisation, their progression within the organisation.  I think clearly this skills data is helping us to do that and personalise those recommendations better for them.  But for organisations, it's really helping us to be more agile, but it's also helping us to understand the gaps and how we should close those gaps, whether that's through developing our internal talent or sometimes obviously hiring new talent as well. 

I know you're at the early stages of the journey, Christophe.  What have been the biggest challenges that you've encountered so far in getting going? 

[0:07:53] Christophe Cabrera: I think one of the biggest challenges we had initially was actually the effort it was taking to be able to understand what are the skills of our existing workforce; and to define a plan and to say, "Okay, in the future, I want to have that skill developed in my company", that's kind of the easy part of the problem.  What is more difficult is actually how do you make sure that you actually identify the skills that currently people have, and that you keep that up to date, because of course the organisation is not static.  I mean people are evolving, I mean because of their job, because of trainings, also because people coming in and out of the organisation.  So, we might get some people today that were not there yesterday, or we could lose people that have critical skills, and we have a kind of leakage on some skills.  So, I think that was the biggest challenge we had in the past.  So, all the exercises that were made that were trying to do that were extremely time-consuming, because it was done manually, it was about talking to people or talking to people managers and trying to figure out what kind of skills we have in the perimeter.  It was not always complete in terms of, I mean, exhaustive, because of course we all think about one topic, but maybe don't think about something we've done two, three years ago, where actually we are perfectly skilled to do that.  Definitely bias existed in that way of measuring as well. 

So, our biggest challenge was basically indeed how to identify the skills, how to make it, I mean fair and equitable how we do it, and how to maintain that over time.  Because if you want to measure the evolution of your population, and it takes six months every time to do the effort, I mean first by design, your view is outdated by the time you get there; and second, most of the people were okay to do it once, but the second time they were like, "No.  I mean, we've done that last year, it was a long and painful enough, we are not going to do it again".  And then you fly by blind again, because you have one moment in time where you have the view on the skills of the people, but then what happened next, you don't know. 

[0:10:09] David Green: This episode of the Digital HR Leaders podcast is sponsored by TechWolf.  TechWolf is an AI-powered solution focused on one mission, delivering reliable skills data for every role and every employee in your organisation.  While companies often spend over 18 months identifying skills, surveying employees, and mapping them to jobs, TechWolf provides all that in a matter of weeks, without disrupting the business.  As skills constantly evolve, TechWolf also provides the tools to help you continuously manage the change, and enable you to apply that data to real-life use cases.  With TechWolf, companies like HSBC, GSK, IQVIA, Workday and United Airlines have accelerated time to hire by 32%, boosted internal mobility by 42%, and saved around $1,000 per employee annually on talent management.  Ready for tomorrow?  Talk to TechWolf, visit techwolf.com.  

Skills have always been important, I think, haven't they?  And you can tell me from obviously working in this space for a long time, but we haven't had the means to capture that information in a way that, as you said, isn't out of date as soon as we've captured it, and can recognise that people are acquiring new skills all the time.  So, how did you solve that problem at UCB? 

[0:11:53] Christophe Cabrera: So, that's where at some point we said, okay, we have to do things differently.  I mean, that way is clearly not sustainable in terms of resource allocation.  And even if we want to put those resources, the quality of the outcome is not good enough.  So, that's where we started to explore AI and skill intelligence basically to say, okay, is there another way we could identify the skills of people or prepare at least something that is already 80%, 90% good enough so that we can really kickstart those conversations and get faster to the real value-added conversation about, okay, what do we do with the skills we have in-house rather than which skills we have in-house.  That's really where we were.  I mean, we were not sure.  That was clearly quite some -- we had some doubts on would that work or not.  I mean, one of our HR leaders was using the casino chip analogy to say, "Okay, that's a risky bet and we're going to put one chip on that one.  But if it works, we can win a lot", and that was really the mindset when we started.   

We started basically with a proof of concept.  So, we said, "Okay, let's give it a try.  We're going to work on a limited population".  So, I think our initial proof of concept was around 500 people; specific domains in the organisation, so not people everywhere, but really focusing on one specific population; no data cleansing, so we really wanted to see if AI could get something out of our data without us doing any data-cleansing effort.  Because that was also something we wanted to avoid, that we were just moving the problem somewhere else, so that, yes, we are not talking to people anymore, but actually we need an army of people to clean data.  So, that also was important.  And also, to evaluate the privacy aspect, of course, because AI skills, I mean personal data processing was clearly an attention point for us to make sure that, okay, if we would go in that direction, yes, from a privacy perspective, it was also something we could explain to the employee, and it was okay according to privacy laws and GDPR.   

So, we really started with that mindset to say, okay, we're going to try to see what we can get in a short period of time.  I think our proof of concept lasted for four to five months, really from start to end.  We only did two iterations in terms of really working on the data.  And we wanted to see, okay, after that, where are we?  Are we in a good point and we see that, yes, it can work, and if we put a little bit more effort left and right, we're going to get something out of it?  Or is it just, no, I mean it's completely off topic and we don't want to go there?  One of the aspects also we really wanted to validate was, can it be specific enough?  Because of course, quite often the skills you will identify are leadership, communication.  I mean, transversal skills are super-easy to identify.  But when you go to more specialised roles, specialised jobs, that's where you want to see that, okay, the specific process or the specific methodology that is being used in a specific domain is appearing basically by itself, and not that you only stick to the surface of transversal skills, project management, programme management, which are important skills, but I mean usually people are getting frustrated because it's like, yeah, I need much more than just those skills to operate my domain, basically.   

So, that was really the starting point.  And what we wanted to get out of that proof of concept were three deliverables: a skills taxonomy; job-to-skill mapping, so mapping our existing role to the skill taxonomy; and the employee skill profile, so for those 500, mapping the employee and create a skill profile, which is unique to them as a person and not something that would be based on some kind of generic personas or something like that.  So, that was really the three deliverables we wanted to get. 

[0:15:57] David Green: So again, four to five months, that's a manageable timeline to do that in.  And let's be honest, how much data will you gather if you'd ask the employees over the same period?  You'd be lucky to get 50% probably of the information you needed.  What was the reaction from maybe your colleagues in HR, but also maybe from the business as well, when they saw the results of the pilot?   

[0:16:22] Christophe Cabrera: I think overall, people were quite impressed by the result we could get through, especially looking at the amount of effort we had to put in there.  So, I mean okay, the duration was over four or five months, but actually we also did some measurement and comparison of if we would have tried to do the same the old ways of working, what would be the difference.  And we estimated that just in terms of effort, I mean we divided the effort by four.  So, to do the same kind of exercise, it would have needed more than 100 man-days in total to make it happen across the different resources involved.  And there, the POC only required basically 25 man-days to execute.  So, we really showed, we were able to demonstrate the efficiency gain.  And of course, on top of that, also the completeness, the accuracy, the data we collected were way more rich than we would have got through the traditional process as well.  So, that was really for us, I mean, very successful and we were quite happy with the outcome of the POC.  And again, starting from the data as they were.  So, we've learned also through the POC some of the data quality issues we kind of knew already and we should work on to make it even better.   

But that was really for us a very, very important point to demonstrate the value of AI, was to say, "Okay, we are just going to give you the data as they are.  The AI engine is supposed to live with it and we'll see what we will get out of that".  And that was for us super-important also.   

[0:17:56] David Green: So next, again, you mentioned that UCB is a biopharma company, 9,000 employees, obviously the proof of concept was for around 500.  And you said at the start, Christophe, you've got a very diverse workforce, employees working in production and manufacturing research as well as office and field roles as well.  How is this shaping your approach to skills?  So, for example, maybe as part of that, how are you looking at maybe rolling out from the initial proof of concept to other parts of the organisation? 

[0:18:26] Christophe Cabrera: So actually, we are way beyond now in terms of rollout.  So, we've rolled out the platform.  So currently, it's already rolled out to 3,000 people.  I mean, we have the employee profile of 3,000 employees.  We are working to get to 5,000 and then further grow in the coming months.  And that's where it comes to which type of data you use.  So for now, we've only used HR data to do that, to apply skill intelligence on top of it.  So those data, of course, we have them for everyone, so job description, work history, learning history, those kinds of things we have for everyone, and we get to quite good results thanks to that.  Now, if we want to really go beyond that, that's where we will need to introduce more work-related data, so data that are really specific to your domain.  And that's where we are evaluating now what's the best strategy moving forward.  Most probably, we would then work more use-case-based.  It's okay for that specific group.  There is a need to go beyond in terms of accuracy, for example, or completeness of the skill profile we generate, and therefore what kind of source we could look at.   

So, if you think about the scientific population, they work on publication.  I mean, they are constantly publishing or contributing to publication.  Yeah, we could use that because if you've been working on a publication on a certain topic, most probably you know something about it.  So, I mean, in terms of skills, it's a way to identify your skills as well.  So, that's a bit where we are looking at now, but we did not move very far yet because, again, so far what we have, just based on the HR data, was already a good starting point and allows us to really start also activating skills usage at UCB. 

[0:20:11] David Green: We hope you're enjoying this episode of the Digital HR Leaders podcast.  If you are looking to continue your learning journey, head over to myHRfuture.com and take a look at the myHRfuture Academy.  It is a learning experience platform supporting HR professionals to become more data-driven, more business-focused, and more experience-led.  By taking our short assessment, you will see how you stack up against the HR skills of the future.  Then, our recommended learning journeys guide you every step of the way, helping you to close your skills gap, deepen your knowledge, and press play on your career.  

You mentioned obviously one of the challenges of doing it the old way is that the data is out of date pretty much the day after, because people are acquiring skills all the time.  So interestingly, again for our listeners, we've been able to actually surface these skills via AI.  How are you ensuring that the taxonomy stays agile and the skills data stays up to date? 

[0:21:28] Christophe Cabrera: Yeah, that's a very good point.  And indeed, that's for me one of the big risks or complexity factor in that journey, is how you manage your taxonomy.  Because I mean, that's where also we really went and started to explore what was available under the technology landscape, because we knew the taxonomy had to be dynamic and also had to be, to a certain extent, self-managed.  You can't have a group of people that will sit in an ivory tower and decide what is the taxonomy of the company; that will never work as a principle.  Taxonomies need to evolve; they are shared across the company and they must be shared across the companies.  But at the end, you will never have one authority that has the right answer, because there is no right answer.  I mean, one skill can appear in one bucket of the taxonomy, and the day after being moved to another bucket, or can appear in two different buckets, and it's perfectly valid.  So, I mean, there is no point to debate in committees forever about which bucket should own the skill.   

Also, what we wanted to get from AI is the external perspective on the skills, so that indeed, if there is skills that are emerging in the market, for example, that we see them appearing and get recommendation by the tool to say, "Hey, actually there is that skill which is, from a proximity perspective, very interesting versus what you have already in your taxonomy.  What do you think?"  And yes, okay, we can add it or we can remove it.  That was also important for us to have that outside in perspective.  There are things we know and we want to do and we will add skills because that's our ambition to develop those skills also, but what's coming together with that one more skill that we added, maybe there is additional skills that could be linked to it.  And that, yeah, the AI will help us to identify because from a proximity perspective, it will say, "Okay, you're looking for that skill.  Actually, you have also three or four other skills around that skill that could be interesting for you".   

So, that's for us something, and it's still a work in progress in terms of governance.  I mean, it's definitely not also an overnight decision that you can make.  But yeah, the idea is to make sure that there is a real conversation around the taxonomy, that the tool is also flagging the skills that exist in the organisation, for example, that are not part of the taxonomy, because that's also not the ambition to have all the skills in the taxonomy.  There is some skills that exist in the organisation that are perfectly valid, that are captured, that are managed, but that do not belong to our taxonomy because otherwise, your taxonomy becoming also a monster with 20 heads, if you try to put and rule every skill that exists in the organisation. 

[0:24:08] David Green: With AI and data, it's people data at the end of the day, there's always that extra sensitivity around collecting that data.  I don't know if you could talk to a little bit how you manage that from working with your privacy team and maybe working with any work counsels as well? 

[0:24:27] Christophe Cabrera: Yeah.  So actually, we did not collect any new data.  So, everything we used is data that used to exist and is already collected.  So, it's about your work experience, your work history at UCB, or outside of UCB if it has been recorded in one of our systems, it's about your learning journeys at UCB, which learnings did you follow, and things like that, the job description, the job catalogue, I mean the role, how they are described.  So, using only those data, we were able to do a lot.  Now of course, that's something we've done in alignment with our privacy officer.  We did review and update our employee privacy notice to inform employees about the use we are doing of those data, because it was a reuse of existing data, but the initial use was not meant for that basically.  So, it was a new purpose.   

So, we worked closely and we communicated around that also, clearly also what kind of decision or often where we will not make decision also based on those data.  So, that's something which has been done and looked at indeed from the early days of the project.  We also worked and discuss with the work counsels on that to inform them about, yes, this is what we do.  So, I think transparency is super-important because there is a managing expectation also of people that, okay, those data are not perfect.  So, I mean to use it at individual level is also very, very risky.  For now, I say that at our level of maturity. 

[0:26:07] David Green: Yeah.  Okay, well, let's talk about those two things.  So, let's go about accuracy and then talk about how you're using this data as well.  So, with any people data, let's be honest, but particularly maybe when you're able to surface this through AI, some people expect it to be perfect.  How do you manage those internal stakeholder expectations when the data isn't going to be 100% accurate, certainly at first? 

[0:26:35] Christophe Cabrera: Yeah, the data by design will never be 100% accurate.  And even that's something we are considering as well to ask the employee also, and to give access to the employee to their skill profile, so they can edit it to correct or complete it.  But if you do that, then you open also the risk of people that claim they know something that they don't know, or some kind of biases in the data that you will produce.  So, I think that's something which is very important to keep in mind is, by design, those data will never be 100% accurate for plenty of good and bad reasons.  I mean, the data quality of the system we use as source is not 100% perfect.  So, there is always things we miss or that are mis-captured or incorrect in our system, and that's impossible to get to 100%.   

The methodology itself, so skills inference, I mean it's a kind of prediction.  So, you get a profile with a certain probability level basically for each of the skills, and you have to decide where do you stop.  So, if the AI engine finds a skill but is only confident by 20% that the skill is valid for the employee, that's a parameter you have to set in your system, where do you stop.  And also, what's the weight of every source you use, and that's a discussion we still have.  So, for now, we decided to go for equal weight.  So, all the source of data we use have the same impact on the employee profile.  There is not one which is more predominant than the other.  But you could perfectly imagine that at some point, you increase the weight of the job description versus the learning history, or the work history is becoming a little bit more predominant versus the other one.  And all of that will change the outcome you get from the system. 

So, I think for me, it's important to always keep in mind that it's never going to be 100%, and what's actually for my specific population I'm looking at, what's the level of quality and accuracy of the data I have, and therefore at which level of detail I'm able to use.  So for now, our recommendation, for example, is not to use those data for any individual conversation, because we see too many outliers, positive or negative.  I mean, we have some people that have a very complete skill profile and others where the skill profile is completely empty because, yeah, data-wise there is an issue.  But more, to use it at macro level, at organisation level, to say, okay, yes, if I want to do a gap analysis of that department, I'm looking at a total department, and then I know that I have an error margin, of course, of what I see, but if I see that overall 80% of my people in my department are properly skilled to do their job, it's okay.  If I see 40%, okay, maybe I have to do something, but is it 40%, 41%, 42%?  That doesn't really matter.  And that's really not how we want to position the use of those data for now, because that percentage can change so drastically just because we fix that quality issue, for example, in one of the systems that, yeah, it will be very risky to use those data at a more granular level.   

[0:29:48] David Green: And that kind of leads on to the next question, Christophe, how are you using this newfound skills data?  Or maybe it's not newfound, it's the data that you've already had, but as you said, you've managed to surface it up to help you to better understand supply and demand and to organise your skills taxonomy.  So, how are you using that currently in UCB? 

[0:30:10] Christophe Cabrera: So for now, I mean we've focused a lot over the last two years basically to build the foundation.  So, the actual use case, we're limited.  Usually, it was ad hoc use cases or specific insight that we were asked to produce, and we leverage the skills that we have to answer a specific question.  Now, we are looking at the next step, it's really something we want to start as of 2025, how to industrialise a bit the use of those data, because it's a super-rich data set.  And we are now evaluating where we want to strike first with it basically, because you can use it in many, many, many different places, in many, many different HR processes.  But again, looking at the accuracy we have currently and where it would make the most sense, I mean, we are considering two areas in particular.   

The first one is internal mobility.  So, how can we provide and help people to find a role which is open in the organisation and where they would have a good fit?  Because sometimes, it's difficult to know what's going on in other departments, in other teams.  And we know we have quite some potential in terms of mobility for some resources that maybe are sitting in research for now, but could perfectly do a very good job in another department because the skillset required is quite similar.  So, if they have an interest for that, they could explore.  So, that's one of the domains we are looking at, is how we could bring those skills data into our internal career site basically, to really make sure that people get recommendations that are based on who they are, what they've done also in the past, because it's not because I'm in a role now that I don't know anything else, and I could perfectly be fitting for a position in another department.  So, that's one.   

The second one, which is also linked to people growth, is personalised learning.  So indeed, it is using the skills data to understand the gaps so that we can develop specific learning programmes and maybe more tailored learning programmes than what we have today.  But also, from a recommendation perspective, again, for the employee, when they get to our learning platform, that they get a more tailored recommendation, because the content is a bit more tailored.  And then if we know a bit more about the employees, we can really connect them to the right content based on their profile, their preference, their role also, what they are expected to do and things like that.   

[0:32:43] David Green: Two more questions before we get to the question of the series, Christophe.  Adoption, what does success look like in terms of adoption at UCB? 

[0:32:52] Christophe Cabrera: I think the adoption, I mean the success will look like people are using those data and they really start leveraging those.  So, part of it will be through industrialisation, so kind of a hidden use, because people will get something and they might not realise 100% that it's becoming better because we have those data.  But for me, one aspect is also to get, especially HR leaders and population, to start to use those data in their decision-making and use those insights and get into the detail and really grow there.  And that's where we have to upskill them also in terms of data literacy and AI literacy as well, because in that case, we use AI.  So, they need to understand also, I would say, the strengths as well as the limits of AI, that indeed we can give to them a certain view or certain recommendation, but it's never going to be 100% accurate and it's never going to be the full context about the employee or the organisation.  So of course, they still have a critical role to play in terms of putting those data in the context where they are sitting, in the context of the domain of the business unit they are working with to understand a bit, okay, what does it mean in practice and how and what kind of decision they can make out of that.   

But really success will be when that's going to become in the day-to-day, and you refer to finance people that are, by design, very much data-savvy and used to work with numbers and they only speak numbers sometimes.  I think that's where we need to grow that within the HR communities, that HR VPs, HR leaders are becoming much more data-savvy and comfortable to use those data also in the conversation they have with the business unit.  So, for me, that's where success will be, that it's becoming for them, I mean part of their toolbox and that they are comfortable to use the tool in the conversation they have with the business leadership. 

[0:34:50] David Green: Christophe, I think this has been really helpful, I think, for many of our listeners, like you, HR professionals.  They'll be working in organisations that are either on this skills-based organisation journey or maybe thinking of getting started.  And I think that the practical advice and story that you've told about what you're doing at UCB is going to be really, really helpful for them.  I don't know if you could encapsulate that and just maybe offer, I'll say three tips, but you could give four or you could give two, that's fine; if you had to give two or three tips to listeners that are looking to get started, what would they be? 

[0:35:28] Christophe Cabrera: I think the first one is get started, because it's definitely a journey!  It's not going to be perfect, and that's also what you need to manage in terms of expectation for the stakeholders that, yeah, it's going to take time.  And to really become a skill-based organisation and to go beyond the word basically and actually embed skills into the HR processes and really consume the information that you have been able to produce is going to take time.  It's not even a one-year journey, it's a multiyear journey before you get there.  But you can only get there if you get started.  So for me, try, learn from it, because we've learned a lot over the last two years about the data we have, about the AI, about people's maturity to use those kinds of technologies.  So, you can indeed also see based on the feedback you get.  So, yeah, for me, it's get started. 

Be open.  I mean, there is no one way of doing it, and depending on your organisation, depending on your context, depending on your domain, depending on your workforce structure, there will be many, many, many different ways of doing it.  So, it can be really, yeah, common.  I mean, there is a lot of things in common across the companies, but the usage will be always specific to who you are as an organisation.  So, yeah, try.  And I think indeed, for me the last one is also work with the expert in the room.  So, that's something we've done in the proof of concept.  So, this is not something we've done in isolation.  We had people from the business domain involved, because at some point, I mean we are not expert and even within the HR organisation, they have a good understanding, the HR VPs, or the learning partners have a good understanding of the skillset of a specific domain, but not as good as the actual people that are sitting in the domain.  So, as part of the proof of concept, we had people from the domain.  So, I mean, yeah, managers.  Actually, we had two persons which are actual managers in that organisation that were part of it, and that did spend quite some time to review the taxonomy, the employee profile and to give us feedback, "It's okay, actually it's good [or] no, no, you're completely missing it".   

I mentioned that we did it in two iterations, and actually the first iteration was not so great, because we had, and we've really started with the default settings everywhere basically.  And of course, the outcome was a bit awkward because it was not specific to the domain.  The clustering of the skills was too generic, and so on.  So, it's okay, feedback received.  We integrated that and the iteration too was way better, because we got the feedback from them.  So, I think for me, it's definitely also something that collaboration with the business unit will be critical to make sure of that, and also to maintain that over time that, yeah, you have the right understanding of what is important for them.   

[0:38:34] David Green: Yeah, and I think that last point is so important because number one, because you made a point earlier, that actually as you mature this, you want to potentially bring business data in as well.  Well, your partner in the business is much more likely to be able to do that if they can see the benefit of this already and they see this as a way to further enhance it.  And then, the second point is really around adoption.  If you want people to actually use this, you need to have the business involved in it, don't you, so I think that's really good.  So, we're going to move to the question of series.  It is definitely related to what we've been talking about, Christophe.  This is a question we're asking everyone in this series of the podcast, and actually this is the last episode.  It's the last episode before the Christmas break.  How can organisations leverage skills intelligence to make more informed decisions? 

[0:39:21] Christophe Cabrera: Like I said in my tips, there is many, many different ways.  I mean, you can embed that into your recruitment process, you can embed that into your mobility process, your performance management, skills-based reward is also a topic.  So, I think it really depends on your business strategy, what is important for you, what's the strategy of the organisation.  Because, yeah, there is many, many, many different ways to apply skill to the organisation.  So, I think that's important that indeed, the use you do of it is aligned with the company priorities and the company culture and the company structure as well.   

Again, we have a very diverse workforce because of who we are.  If you would compare that to maybe some consulting companies where 90% of the workforce is doing the same job, they don't have the same challenges as us, and they have different needs as well.  So now, there is a lot in common which is, how can you make sense of the vast amount of data you have already available about your employees, because every company has work history information, project history, learning history, job description, job architecture.  I mean, it's not always called in the same way, but those data exist and is available to you.  So, yes, you need to work on your taxonomy; yes, you need to work on all of that to really be able to connect the demand you have in terms of skills as an organisation to the supply.  And also monitor that.  I think for me, one of the big, big potential of skill and skill usage is also to detect possible leakage or shortage you have on critical skills for your organisation.   

So, that's something we've tested a little bit in our proof of concept, for example.  We looked at some critical profile and we looked at existing population that could fit that role if the person currently in that position would leave or move on to another role.  And it was quite interesting because for some roles, we could see that we had a rich pipe of people that with maybe some upskilling or reskilling, could come and take over that critical role in the organisation.  But we also saw that for some roles, we had a potential issue, let's say, that if the person would leave or move to another role, there was no obvious successor, basically, in terms of competence.  I'm only talking about skills here.  We are not talking about, I mean the whole other aspect that you would need to consider before to consider to move the person.  But, yeah, to also manage risk, leakage of critical skills, because if you invest in a training programme but you see that all the people you've trained are leaving the company afterwards, that's a bit of a missed opportunity. 

So, yeah, there is very, very, very different possible use case and that's what is great and frightening also, because then, okay, where do we start?  It's like, okay, what makes sense and where is it that we can have a real impact for the organisation? 

[0:42:27] David Green: Well, and it's all about that, isn't it?  It's where can we have the biggest impact for the organisation, and let's prioritise around that.  And I love that little bit there, thinking about how you bring skills into succession planning.  So, it's not just about, okay, the successor 
is the person that's their assistant, or whatever it is.  So, really good.  Christophe, thank you so much for sharing the skills journey at UCB with listeners of the Digital HR Leaders podcast.  As I said, I think listeners will be able to take a lot away from this conversation, so thank you very much for that.  How can people stay in touch with you and learn more about UCB? 

[0:43:08] Christophe Cabrera: I think LinkedIn is probably the best way to get in touch with me.  So, I'm always happy to have a call, a conversation to bounce back ideas also on that topic, because it's definitely a topic that is in the making.  So, I think nobody can claim that they've figured out everything.  Every day brings its own set of challenges.  So, yeah, LinkedIn, Christophe Cabrera on LinkedIn, UCB, more than happy to connect with anyone who wants to stay in touch or discuss.  I mean, feel free to drop me a message and we can have a quick chat, or online if you're not in Belgium, to discuss about that. 

[0:43:47] David Green: Perfect.  Well, Christophe, thank you very much indeed and look forward to hearing the next step of the journey at UCB maybe in the future. 

[0:43:55] Christophe Cabrera: Yeah.  Thank you very much, David.