326 - Never let an LLM do the math on comp. Do this instead.

0:00Sébastien Baehni: But what people maybe don't see is that comp's the iceberg. The decision is one part, but there's a whole heap of stuff underneath. And that stuff underneath is like the budget, the comp rules, the country rules, the calibration that's happened underneath. And that is not in the manager's control. That is part of the framework of compensation that companies use for fairness, for an equitable outcome, for metocracy, for all sorts of business-driven reasons to drive outcomes.

0:28Clare Bonham: When we say intentional, it's based on three main things. We want the AI to be, or the tools, or everything that we put AI based into the product to be explainable. Meaning that you can have and trace a decision that has been done. So audit is very important for us. I mentioned it a bit before, collaborative as well, in the sense that human is in the loop. so it's not because you have AI and all this superpower as you mentioned Stephen that you can let it go at least for now we know in a few years but at least for now I mean right now I mean we want really to have something that is comfortable the human in the loop and making sure that the human has the as the last word into that and and controllable in the sense that you can control which piece of AI you are using for what.

1:21Clare Bonham: You're listening to the Modern People Leader podcast, where we go behind the scenes with today's top HR leaders and work experts. We talk to them about how they've gotten to where they're at, what they do every day, and what they see coming next for the future of work. Hope you enjoy.

1:40Clare Bonham: Welcome, everyone. We have a group format today, so there are a lot of lovely faces on the... If you're watching this on YouTube, yeah. With group pods, there's a lot to say. So I'm going to get this rolling pretty quickly. Our first tradition on the show is what we call good news stories. It's just a way to kick things off with some positivity. It could be work. It could be personal. We all share some good news. And we prefer for our guests to go first. So Sebastian, Claire. Yeah, let's start. So I can start with no problem. So with good news, I would say it will be work-related. And we are working quite hard right now on the new version of our product.

2:24Clare Bonham: And there are some customers that are actually using it and upgrading to it. And we received very good unsolicited feedback from one of the customers that basically was saying that he has never seen a tool like that. And I think it was very good news because it's not something that we ask for, right? I mean, sometimes we ask for feedback through NPS and these kind of things. And here it was really an email completely unsolicited coming from the customer back from his vacation saying, hey, I've been playing with the tool. It's just fantastic. And congrats to the entire team for what we are doing.

3:01Clare Bonham: So I think that was really an uplifting message from this customer. Good news for me.

3:08Sébastien Baehni: Great. on my side I would say maybe more on a personal front but I have a six-year-old daughter and just with the summer passing I've seen a lot of just great changes in her in terms of her independence and readiness to go back to the next level of school and it's just been really fun to watch her get ready now for this next what will be her second year of school which we call senior infants here in Ireland so for me that's really good news she's excited to go back so I'm really happy.

3:39Clare Bonham: I feel like if I were to start a second podcast, it would be like a parenting podcast. Me and Steven are like deep in the throes of like, well, he's got, Joaquin's a seven-month-old now? Soon to be eight? Soon to be eight, believe it or not. And my daughter is 15 months old. So still babies, but yeah, it's fun seeing the development. What's my good news? I'm going to go work-related. So about like two or three weeks ago, So Steven showed me this new sales tool and I'm going to be kind of gatekeeping and not the name of it, but it's been working wonders. It's an outreach tool that helps us do things in a more, I'm being very like vague and cryptic, but it's working.

4:24Clare Bonham: Like we've booked probably three or four sales meetings here in the last couple of weeks. And I just like love to turn it out on all things sales and marketing. And so that's my good news. I'm going to keep it work. We have some technical folks on the call. So I'll keep it technical. I had a hypothesis. So in our last podcast, I don't know the sequencing of how these are going to be rolled out. So it could feel like weeks, but that was days ago for us. I announced that this week I started a full-time role. And so I am just a co-host. I've worked in the business of Modern People Leader for a year and a half, a little over that.

5:07Clare Bonham: And I recently left and have shifted back into my co-host role. And my new role is completely different, like midlife, like total career change and head of revenue for an early stage startup. And it's an AI powered music software company, non-generative. And so one of my hypotheses is that at this stage, like you don't, salespeople always want more tools. It's like, I need this, I need that tool. I need that buzzy thing. I need that. And so that's like one end of the spectrum. And on the other end of the spectrum is like spreadsheet hell, like literal spreadsheet hell. We've all got like, you know, two or three sales trackers, a growth tracker, things are coming from different places.

5:56Clare Bonham: You know, I'm never updating the things that I should be updated, but everything's shared little nightmare. And so I advocated early on, like in my first week that I could build, I could vibe code. And I've, I've, you know, I would say like, I'm pretty average at vibe coding, maybe slightly above average. And so I am like six prompts out of eight in vibe coding a CRM. And it is like, I'm just, I feel bad for the CRMs because literally in like a day, let's call it, not even a full day of like human work, I had built a very, very well-functioning CRM that takes the data from the system that Daniel's talking about, from other systems.

6:42Clare Bonham: It wanted to implement Google OAuth, so single sign on. And I was like, no, no, no, no, no, don't do that. Like we'll worry about that later. So like, I could have had that as well. And it's just like, I don't know why it keeps blowing my mind y 'all, but like that I feel like is very good news. Well, it's great news for me because it's going to help me scale tremendously as we start to like scale the difference and scale revenue, which is like, you know, the company wants me focus on closing deals, not on updating spreadsheets or making sense, trying to consolidate, roll things up to my boss who obviously wants the update.

7:20Clare Bonham: So I'm super stoked. That's my good news for today. But we're not here to talk about us. We're here to talk about Become and some of the cool things that y 'all have going on. And so the next segment is what we call your origin stories. and I also call it the Brene Brown question. And so like Claire and Seb, walk us through your backgrounds, your paths and how that ultimately came together to you working on some pretty amazing things I've become. One thing keeps coming up on the podcast and that's leave. It's one of those processes that everybody has but very few get right. And we just talked about this the other week with Kim Smith from Tilt and she framed it in a way that really clicked for me.

8:06Clare Bonham: Not every employee experiences leave the same way. Some want every detail up front, others feel overwhelmed and just want you to simply tell them what to do next. But most companies deliver the exact same experience to everyone. It's one thing to have a compliant leave program. It's another to have one that actually works for the human going through it. And that's where Tilt comes in. Tilt is a leave experience platform that helps teams move beyond one-size-fits-all processes. Using Tilt, companies can tailor the leave experience to each employee's needs, keep managers aligned on coverage and return to work, automate complex workflows like pay and compliance, and give HR the space to focus on support and not spreadsheets.

8:51Clare Bonham: In other words, the tech handles the complexity so that the people team can show up where it matters most as a human. Leave isn't just an HR process. these are life-changing moments that can make or break the employee experience if you want to learn more about tilt go to hellotilt.com slash mpl that's hellotilt.com slash mpl yeah sure I can go

9:14Sébastien Baehni: maybe first uh so yeah my background is really hrit and being a bit of a hrit nerd I would say I sat in a large multinational um I own it Stephen I own the nerdiness uh but I had great experiences there and I got great exposure to different HR tech across the HCM suites, across best of breed, across payroll. And I got to work with just really passionate people around that. So that really formed how I think about HR tech. When I then decided one day as all, as you just described, you wanted to build a CRM, I decided I wanted to build a different way to do performance management. And I partnered with some colleagues that I'd worked with, great HR directors, and we built a product called Our Tandem.

9:58Sébastien Baehni: And then Our Tandem was acquired by BCom back in 2022, which is when I first got to meet Sebastian and the team. And I've been now with them for four years and I'm the chief product officer here. So that's really how I came to meet Sebastian. So I'll hand over to him.

10:16Clare Bonham: Yeah, well, on my side, I'm a nerd as well, but more on the technological side, I would say. So I've done my studies and I'm a master in computer science. And then I went in different, I would say, companies, starting as a developer, climbing a bit later. And I mean, participating in, I would say, early startups. And back in the day, so eight years back, I've been working there quite a lot. And I joined basically Becom in order to be their CTO. It was seven years back. And over the past seven years, we've done quite a lot of different things. And I mean, we have been acquired by private equity funds, Sumeru.

11:02Clare Bonham: After that, we did two acquisitions, and one of them was our tandem. And I had the pleasure to have done the tech due diligence with Claire. So that was quite fun, I would say. And yeah, we ended up working together. So me as the CTO and her as the CPO. And here we go. And I would say in the last couple of two years, I mean, my work has been transforming due to AI, as you mentioned, Stephen, with web coding and all this kind of cool stuff. So here we are. And how would you describe what BCom does to a stranger? So say, you know, you're out and about and you meet a stranger or you meet somebody at a conference.

11:48Clare Bonham: what's like the quick 30-second pitch?

11:51Sébastien Baehni: I'll go first. Yeah, sure. I'll be honest. Daniel, it's very rare that I meet a stranger who asks me about B2B SaaS software. But if they did, I would say that we help people make better and fairer pay and performance decisions. And we like to be a tool that supports employees, managers, and HR around making those really important decisions. And I try and keep it as simple as that.

12:14Clare Bonham: Tell me about that B2B software that you've been working on, Claire. I'm just trying to imagine. I know it. Good call out, good call out. Yeah, yeah.

12:23Sébastien Baehni: But apart from the fact they never asked me, that is what I would say. If I do get to the second part of the conversation, Daniel, that's usually what I go to.

12:29Clare Bonham: I am a fellow HR tech nerd. I started my career as a compensation analyst at Ernst & Young in the global mobility space. So that's like as nerdy as you can get. I was like, not even cool comp. I was like comp embedded in a very kind of obscure, like the stepchild, the HR stepchild, which is mobility. So Claire, I want to start with you. In our prep call, you made a point that, you know, we tell managers to be a coach and to be a trusted ally. Then in the same breath, we make them responsible for whether someone gets paid more next year. The future is highly dependent on them. And so I want to start there.

13:14Clare Bonham: Like, what does a manager actually need in front of them to make a good one? Like what usually gets left out?

13:20Sébastien Baehni: Yeah, absolutely. It's a great question. I think for me, it's context, right? Which is a really broad statement to say, but it comes down concretely to things like, obviously, you need the pay history, right? Because particularly if you've taken on a team and you only know them this year, you need to understand the journey they've gone with with their pay. because as well, when you're relaying that message, it needs to be in context of how they've felt about their pay in the preceding years. It's not just, you're not just talking to them in that one year. So there's a piece about pay history. Then you need to know where do they sit currently within their range?

13:53Sébastien Baehni: How does it compare? You also need to clearly know what's the budget envelope available to you to work within. And I think you need as well, this is where I feel quite passionate. You need also that performance and talent potential information because you also, So even if it's a merit decision, you've got to make it in context of thinking about that person and how they've performed and delivered. And I think sometimes what gets left out is that softer side of the data. We can sometimes come at comp and, you know, this from a global mobility sense, it's like a mass of numbers. And I think you've got to make sure the person's there and you've got to make sure that information about them and their broader impact to the business is part of the conversation as well.

14:33Sébastien Baehni: So for me, that kind of gives the manager context to take a decision.

14:38Clare Bonham: And why do you think some of that softer side gets missed? Is it just because there's not good access to the data or, you know, managers are in a time crunch and they're just like trying to get, they're trying to make decisions? What do you think is the issue there?

14:54Sébastien Baehni: Yeah, I think it's probably two or three things. I think one is definitely data. And we see that a lot, that bringing a lot of disparate data together in a way that's contextual is difficult, right? Because you've got different systems running things. Global mobility is a great example. It is the stepchild of HR. So if you're an international assignee, you're an expat, your data is sitting somewhere else completely. You probably have two records. What's the right record to use? Does the person in the home or host country use it? So data is like an absolute key part of that problem. And then I think it's about the second part is just understanding that, you know, while pain is to be driven by really measurable factors and pay architecture and job families and things like that, that the impact of, you know, looking at talent and looking at the holistic picture of someone does need to contribute because subjectively the manager has that in their mind.

15:48Sébastien Baehni: It may not be in the form they're filling out, but they for sure have it in their head, some information. And I think it's better that every manager has that same information when they're making that decision in front of them. So I think it's the two things.

16:01Clare Bonham: I got on a high horse the last, probably two or three years ago, and it lasted for about a year, just about managers and the new, you know, the evolving role of a manager and how much is like sitting on the shoulders of said managers. And, you know, one really difficult thing that sits on the shoulders of managers, I believe is ownership of the compensation, like decisions for their team, on their team. And I feel like many times, at least in my experience, like the manager was just like communicating a number to me. It was like, yep, here's your grading. I literally had a manager at Ernst & Young that was like, yep, you get a three, you get a four.

16:47Clare Bonham: This was in the knots a long time ago. Hopefully no one's doing that today. But then it was like, all right, come into the room, you get a 5 % increase, you're getting a six. And it was like, but why? How? And it always seemed like my manager was deer in the headlights. Like they didn't know. It was clear that they weren't ready to be accountable for those decisions. It was almost like they were like the spokesperson of the decision that was being made. And like that was implied, an implied decision by it was implied that that decision was made by HR. And so I just want to like talk a little bit about like how do you view this?

17:35Clare Bonham: Like the best companies, how are they managing this? The best managers, how are they approaching it? And like, what is that subtle difference between truly taking ownership and just being, you know, the communication layer?

17:48Sébastien Baehni: Yeah, it's a great question. I think the companies that do it really well explain the roles and responsibilities and they don't allow the line to get blurred. So I think it becomes about saying the reason managers should make decisions about comp is they are closest to the work. So it really makes sense that they're part of that decision-making process. But what people maybe don't see is that comp's the iceberg. The decision is one part, but there's a whole heap of stuff underneath. And that stuff underneath is like the budget, the comp rules, the country rules, the calibration that's happened underneath.

18:22Sébastien Baehni: And that is not in the manager's control. That is part of the framework of compensation that companies use for fairness, for an equitable outcome from metocracy for all sorts of business driven reasons to drive outcomes. So what goes wrong is when the manager, I think, positions it as you say, oh, this is what I was told to tell you, or this is a three, but I can go no further. Whereas I think what should be very clear is that we should say to managers to help them have great conversations authentically, there are a lot of constraints here. And we should say it to employees as well. It's not an open checkbook.

18:59Sébastien Baehni: There are a lot of constraints. There's a way it works. It's like about a career path. It's about a job level. It's about a budget. It's about how successful the business is, which impacts the bonus part. A lot of things happen. And then there's this piece of decision and judgment that the manager owns. So we should be clear that manager owns the decision, but they don't own that envelope, right? And the minute you try to either blame the company and say, oh, it wasn't my call. It's, this is what you got. Or the manager takes like full accountability for everything. I think it costs them a little bit of credibility in the conversation.

19:31Sébastien Baehni: And I think, Stephen, probably that's what you experienced. It's like, you know, you've got this manager who you're reporting to every day, but suddenly in this area, they're either have zero influence or they're, you know, apologizing for what they've done. And I don't think it needs to be that way. I think it should just be a much clearer conversation about these are my constraints. This is what I could do within the envelope. And we're all adults here. And I think that starts the basis of people who do it well. That's the type of tone they set in the conversation.

20:01Clare Bonham: So let's do like a little bit of a role play here. So let's say Steven is your direct report and you know you're about to deliver this hard news. You've kind of laid out the framework, but like, yeah, give us like the opener to this meeting. Like, how are you setting the stage? that way it doesn't turn into like you know one of these like very yeah a tense situation like how do you avoid that sure first of all clear i'm sorry that i'm your direct report i'm a total pain in the butt so just saying i apologize but no let's roll with it this is a great role play

20:41Sébastien Baehni: because once again i'm coming to the conversation with no context of your pay your performance for anything. So let me set the tone. But I would start by saying, you know, Stephen, we're delighted, as you mentioned earlier, you're moving into a new role with new accountabilities and new revenue. What we want to see from you is delivery in that space. Your compensation is naturally tied to that delivery. So a lot about what we're going to talk about today is what we want to see with that delivery, how I can help you achieve that and get the best out of it, and what that looks like in in terms of a meaningful outcome for you.

21:13Sébastien Baehni: And then from there, having set that tone that there is conditions around what we're going to talk about, I would hand over to Stephen to reply.

21:21Clare Bonham: Everyone listening, like take note of the timestamp here and like create a script and share your managers. That was great, Claire. So I want to bring you into the conversation. And, you know, One of the things on the tech side that I think is pretty common, well, first of all, tech overload for managers, right? Managers are the key approval point in so many processes. I think with my personal opinion is that in the go-go SaaS days, I think we got a little carried away with like, you know, in our pursuit of achieving scale and straight through automation, like, do all of these things really need to fall on the manager?

22:05Clare Bonham: I don't know. We'll save that for another conversation. But a symptom of that possibly being right is that it's really hard to engage managers in tools. My wife works at a 12 ,000 employee company. And she's fairly senior, so maybe it's not that applicable. But her manager, she's a chief of staff, and she literally has to go in Workday and she does all the things in Workday. And it's like, I don't think that was the, I don't think that's the model that we're looking for here. And I'm super curious from a technical perspective, like how do you design a system that gives you what you need, but isn't overwhelming with features and functionalities or isn't like so cumbersome because comp is, you know, secure, right?

23:00Clare Bonham: It's like definitely confidential. And so there are like layers of security and access rights that need to be involved. And so I'm curious, like, how do you design something elegant yet simple that is like in the strike zone or the bullseye of what a manager is looking for? How much compensation risk do you think is hiding in your spreadsheets right now? For a lot of people teams, comp has gotten way more complex than the system supporting it. You got merit cycles, bonuses, incentive plans, approvals, benchmarking, pay equity, transparency requirements. And somehow a lot of it is still being managed through Excel, manual uploads, or HRIS modules that weren't really built for this level of complexity.

23:43Clare Bonham: And you know, that's all fine and dandy until there are three versions of the same comp file, or one person becomes the single source of truth or leadership asks why a pay decision was made and the answer is harder and harder to explain. That's where BCom comes in. BCom is a purpose-built compensation platform for organizations managing complex pay programs, especially those teams that I was just referring to that have outgrown what spreadsheets can comfortably handle. It helps people in total rewards leaders run complex comp cycles at the speed of business, connect compensation, performance, and pay equity in one place, create clear governance with clear audit trails, and use AI responsibly with humans still in control of the decisions.

24:31Clare Bonham: If your comp strategy is getting harder and harder to manage, Be Calm is worth checking out. You can learn more at becom.com slash MPL. That's B-E-Q-O-M dot com slash MPL. I mean, that's a$1 million or maybe$10 million question. So it's... No, but I mean, I think you nail it in the sense that you make this in order to have a very good experience for the end user, right? You need to make sure that the managers are willing to enter the system and they don't see it as an additional burden. And I would say it starts to understand your end user. It starts to understand your personas. So an HRBP experience won't be the same as a manager experience.

25:15Clare Bonham: So it all starts, I would say, with very skilled designers, UX experience, having direct contact and touch point with those different personas. So I don't think it starts with an engineering. I think it starts more with really, I would say, product management and designers in order to understand the pain point, understand what the, I would say, the person needs to have directly into the system. And then it's about, I would say, hiding the complexity of the entire, I would say, process. And really, I would say, making, popping up some recommendations. So for merit review, for instance, I mean, you need to have three, four things that really, really matters that needs to pop up so that the managers understand what's going on and can go through, I would say, the experience the most simple way.

26:04Clare Bonham: They don't have too much time in order to do that. So they need to have this really at a glance. They need to be able to answer questions as well. So nothing needs to be hidden, as you mentioned, right? It's complex. So nothing needs to be hidden. They need to go and potentially drill down in order to better understand. But it's really, so drill down is really something that needs to matters. And also maybe highlighting, flagging things that maybe pop up for something that are, I would say, exceptions or something like that. So I think it's quite being close to, I would say, the personas to whom we are developing for, understanding, getting back the feedback also about this.

26:42Clare Bonham: I mean, the number of clicks that they need to do in order to go into the system. They are not using that all the time. And you said as well, I mean, the problem is that comp is complex. And sometimes we have customers that just want to have 100 columns in order to understand what's going on. So making that as simple as possible, yet having this possibility is, I would say, the little tension where we are at. Seb, when we last talked, I think you said something to the effect of people don't want their comp decisions solely decided by an AI agent. And so let's just start there. How do you view how much of this should be decided by an agent and how much we should let the AI help us in this in comp?

27:31Yeah, so let me take a metaphor here.

27:36Clare Bonham: So I would ask you the question, how much you would like an autopilot to land solely the plane where you are in? I would not like that. And you need all the electronics. you need all the, I would say, I would say the software that is in the plane in order to make this happen, right? But you would like to have still a pilot in control. So I think it's where I would answer this question. It's like for me, I mean, AI should be here in order to help around the complexity of the compensation because it's so difficult in order to grasp and to understand, especially if you are in an international company with, I don't know, 40 or even more countries and I would say different multi-currencies, different, I would say, LGBT rules and all this kind of thing.

28:25Clare Bonham: So AI needs here to help and to give feedback to the, I would say to the manager or to the HRBP in order to take some decisions. But the people needs to be always in control about what they are using into AI and in control of different things and different pieces of the governance as well so that they can say, okay, no, this is not good. I will take that. I will also overwrite maybe some of the rules that have been done by AI. And so for me, it's really where I see AI as a help, as something like an autopilot that can help for basic things, for helping the managers and the teams in order to focus on what needs to be done at the right time.

29:09Clare Bonham: So this is where I would put that. I've said that I am not jealous of product people, product leaders, and CEOs of software companies, product-led companies, because with AI, it's just, it has changed so much, right? Obviously, you know, it's changed work itself. It is, it's changing all the dynamics within an organization. There's a lot of, a lot of, and there are a lot of dynamics that are being impacted in my, in my mind today in the workforce. But it's also like as a consumer, you know, of software, it's change. It's also made that complicated, right? I think that, you know, AI is a superpower that any person can access.

29:58Clare Bonham: Technical people can take that to another level. And so like, competition is just like so hard, like, you know, it just feels like there's, the channels are all very crowded with people saying they can do different things. And so, you know, when you in the team looked at how you were going to take the AI superpower and start incorporating it into your product, like, how do you look at differentiation? Like, you know, how do you look at like, okay, you know, the most obvious things, you know, that we could do are A, but like the right things to do in order to differentiate ourselves are B. Like what I think you called it like intentional AI.

30:41Clare Bonham: And I think that's really what I'm getting to here. And so Seb or Claire, like I'll leave this one open to you all, but like, tell me a little bit about like the, you know, design thinking that you all have put into how you leverage AI. Yeah. So do you want, I can, I can do it clear and you can compliment if you want. Yeah. So for us, I mean, when we say intentional, it's based on three main things. We want the AI to be, or the tools, or everything that we put AI based into the product to be explainable. So meaning that you could have and trace a decision that has been done. So audit is very important for us.

31:19Clare Bonham: I mentioned it a bit before, collaborative as well, in the sense that human is in the loop. So it's not because you have AI and all this superpower, as you mentioned, Stephen, that you can let it go, at least for now. when I was in a few years. But at least for now, I mean, right now, I mean, we want really to have something that is comfortable, the human in the loop, and making sure that the human has the last word into that. And controllable, in the sense that you can control which piece of AI you are using for what. And I can go a bit deeper after if you want about what it means, because they are different, I would say, kind of AI, and you should not use the, I mean, a specific kind of AI for a specific problem to solve.

32:05Clare Bonham: And all that for us are together. Explainable, collaborative, and comfortable for us, it has to be together. If you have only two, then you will have something that doesn't work. If you can explain, but you don't control, well, I mean, then you can potentially audit a decision, but you never really authorize it. So, okay, so what does it mean? Do you have really control on your system or not? And if you have something collaborative, but you cannot explain, then it's, I mean, the human is signed off on something that he cannot expect. So if you miss one, you don't have intentional AI. I think you have, I would say, a good intention.

32:45Clare Bonham: And I think that's what we are trying to really put into our system. And this is what we said, that it's really intentional and that you need to have those three things. I don't know if you want to add something on top of that, Clare.

32:57Sébastien Baehni: Yeah, maybe just one point, which is, I think when we thought about the philosophy of intentional AI, it really came rooted from a place of we fundamentally believe that compensation and performance decisions impact people. It impacts the way they're promoted. It impacts how they feel about their job. It impacts how they feel about coming to work sometimes as well. And for us, that is really an important factor about thinking where you use AI, because people deserve to have someone take time and think about their performance, think about their growth, think about their comp. Should we make it as cognitively simple as possible using tools like AI?

33:38Sébastien Baehni: Absolutely. And should we have great design around those things? And everything Seb said is absolutely accurate. But at a product philosophy level, for me, I still would like to think my leader thinks about my performance and my reward for that performance in a way that's reflective of the contribution I make. And I think that's what AI can't bring to the table. So for us, that's where we wanted to have intentional AI. It sits in specific places. It sits and does specific jobs. And it's controlled by those three variables that Seb spoke about. But at the heart of the philosophy was that this is something that, you know, we shouldn't take away, no matter how busy a manager is, we shouldn't take away from the fact that as an employee, you know, a manager or a person or a human should be part of those decisions.

34:29Clare Bonham: So on the explainable piece where you can go in and audit, I'm almost imagining like, you know, being in like a, this might be a bad example, but I'm imagining, you know, in a vibe coding platform, I put in a prompt and then you can like literally see it all building on the left. and you can go back and look and see what exactly they did. Is that what you mean by explainable to where like a manager can go in and see the data sources that are being pulled from and the calculations and everything? Like how granular is that? Yeah, exactly. So, I mean, the goal is that potentially about explainable is that when there is an audit or when there is a question about why did you take this decision with respect to either my performance or my compensation, then you can actually go back into the system and understand what's going on.

35:22Clare Bonham: In the sense that you don't need a project. Let me put it like that. So when someone is asking this, you can say, okay, no, wait, we have used this, I would say, in order to, I don't know, compute the model, we can go into the details later. We have used this statistical model in order to, I mean, compute the pay range where you are. We have been using those rules in order to, I mean, put you in the eligible model or where you are in the country and these kind of things. So the goal is that you can really trace back what has been happening without really praying that, well, actually, if you ask the same question, then you don't have the same answer.

36:04Clare Bonham: So it's also about not pure auditability, but it's also repeatability to understand what happened, right? And that's explainable for us. It's really to understand how it has been done and that you can explain to the people why he has this kind of pay, where is his day band, and which part of the AI has been used here. So it's not a black box. You can potentially open it and go deeper. It's like taking the, I would say, back to the concept of the pilot is that if there is a problem, then you know what happened and you can go and understand what happened. Usually you don't. You don't need that. But if you can, then it has to be audited.

36:47Clare Bonham: It has to be logged. It has to be transcript. You made the point in our prep that most people don't even realize there are various forms of AI. We just assume the bots are doing the things and they're doing everything. I'm just curious, what are the various types? What do you think are the types of AI that are most impactful for compensation and performance? Hey, everyone. Really quick. I can already tell that this is going to be an episode where you're going to want to take a lot of notes. And that's kind of a pain in the butt when you're casually listening. So we have made it super easy for you.

37:27Clare Bonham: and we've turned the entire episode into a downloadable PDF so that you can enjoy the conversation and grab the notes later. The link is in the description below, so go grab that before you forget. And now, back to the episode. Yes, so I can start and not indicate this technique here. So for us, there are three types of AI. I would say there is the fancy one, the one that we are using today, right? It's everything regarding the language layer, LLM, and all these kind of good things. Brilliant at explaining policy in plain English, summarizing, trying to help you in order to take some decisions, I mean, telling you what the definition of a compare ratio is and these kind of great things.

38:14Clare Bonham: So really advisory, I would say. Then another piece of AI for us is statistical modeling. So taking the 20 variables that you have, or even more, maybe 70, 100 variables that you have, market data, range position, performance, tenure, and all these kind of things, and also your own data, and taking all this and creating a suggestion for a potential pay raise. So that needs a lot of supervision. I mean, all that is, I mean, it's machine learning. It learns from your history and your biases. So that needs supervision in order to make sure that, I mean, those models needs to be trained. So I think it's quite important.

39:01Clare Bonham: And the last one is maybe, I mean, it's the one that started, I mean, pre-order 20th century. It's more the rule-based logic. And it has been tagged as an AI back in the days. and it's the boring one. So basically, I mean, how you create the LGBT rules, the floors, the country rules, and all these kind of things. And I mean, nobody is excited about that, right? I mean, it's over now. But it's, I would say, the part of AI that really passes the audit because you can really explain with the rules, with the, I would say, everything that has been done in the formula, basically why, I mean, you are part of this, I would say, group of people.

39:43Clare Bonham: And coming back to your last question, Stephen, it all stacked together. I don't think that there is one that you need for best for comp. For us, it all stacked together. There are three layers, one source of truth. You stack everything together. You have the rule base, then the statistical model. On top of that, you have the LLM in order to give you advisory based on the information. But you should, I would say, never miss it or having, for instance, LLM taking decision about statistical model. You need really to split these two because as soon as LLM is touching numbers or mathematics, then it's where you start to drift.

40:23Clare Bonham: And I would say LLMs are, as you know, I mean, built to be plausible, but comp needs to be correct. So that's why you need really to have these layers stacked in a nice way and not touching each other. Claire, do you want to add something?

40:38Sébastien Baehni: I think you're very comprehensive there, Seb, but you're all getting an insight to our team meetings. So I hope everyone's in there.

40:45Clare Bonham: I love this. So something we've been nerding out on the last four or five episodes, just like how the people function is going to change so drastically here in the next four or five years. And we're already seeing some of these changes, just like in the types of roles they're hiring for, the new roles that are being introduced to the team. And, you know, we were at Transform back in March and we hosted like an impromptu breakfast and the topic of comp came up and how different that sub function within HR looks today versus a year or two ago. And so I'm just curious if you guys have any thoughts on that.

41:27Clare Bonham: I know we're going, we're going off script, but any thoughts on what the future of comp looks like for the people in those roles?

41:33Sébastien Baehni: Yeah, I think that For comp, you know, historically they've spent a lot of time, and Stephen can attest to this from his global mobility days, a lot of time in spreadsheets. They've spent a lot of time looking at raw data, bringing people together, bringing single records. And a lot of the work of comp was, particularly comp operations, was really just data intensive, data cleaning, data, you know, integrity type work. and I think the opportunity now is you know with some of the tooling that's coming with the different types of AI is to take the lean off that but I still think and I think it allows them to refocus on the work of comp in terms of what are the incentive plans that matter how do we pay people fairly how do we acknowledge the new skills that are emerging and remunerate them well so I think there will be an absolute shift in the focus in the way that you see automation touch any industry.

42:26Sébastien Baehni: I think it will take out tasks that were labor intensive, combining excels and pivoting data and doing all these things. And I think there's a lot of things that can be done there. But I think the job of what's the right reward strategy for this company, what's the right way that we can drive the business forward? How do we incentivize Stephen in his new role of revenue to get where he needs to get? All of that stuff is where I think comp people naturally can spend their time and should thoughtfully spend their time with the business. So I see it as an opportunity to lift for a comp to lift because it's a massive mountain of work to organize and run comp, particularly in large enterprises as well, where the data sits across many systems, many countries, as Seb said, there's many rules.

43:13Sébastien Baehni: So I think automation is absolutely there. And I feel like that's something that, you know, anomaly detection, all this kind of stuff can really help drive great, great comp outcomes. That'd be my view. Seb, I don't know if you have any other thoughts.

43:31Clare Bonham: I think you have done a very good, very good statement here. On my side, I would say in terms of comp and what might change in the future, where I could see is that people need to really maybe focus on the compensation, maybe a bit more on the individual contributor. Before, right, I I mean, there is the management ladder and the IC role, right? And what we can see or what we could see is that the management layer, managing 10, 20, 40 people might, I mean, fade away with the power of AI because you can do quite a lot of things. At least I'm talking about maybe on the tech side, but you can do quite a lot of things and quite be powerful as an IC now in order to do quite a lot of different things without a big team as before.

44:16Clare Bonham: So I would say focusing on this track about IC, where the people will go, I think it's something that I would not be surprised that comp needs to focus more. Daniel didn't ask me the question, but I'm going to answer anyway. Because it's just like crazy how much has changed when I was just listening to you and also thinking about what my answer would be. Like the first thing, so in 2023, I sold my engagement survey company and I didn't, I went through like a kind of career and direction life crisis. Like what I had, I had over indexed my work persona in this, this company and, and the outcomes that that would, that would present me.

45:05Clare Bonham: And then you get there and it's like, well, okay, now what? And so I did growth consulting, which is ironic to now look back and see, oh, okay, the universe was setting me on a path to shift to revenue and sales. And I worked with a compensation startup. And this is 2024. And I helped bring on a customer. And we went through this crazy exercise. And I'm sure it's all different now. But we essentially, they had no job leveling, but all of that like job architecture data was required for the comp system to do all of the great things that it can do. I'm a big fan, continue to be a fan. And I literally like, you know, I was hired because I knew that the I also have a comp HR background and I knew the person that purchased the software.

46:04Clare Bonham: So they engaged me for like this side of the project to just help them with the all of those spreadsheets that had to be populated for the comp system to work. Like today, I don't think any of that needs to be like one of the things that AI does so well is like it can review data and it can like it can, you know, I haven't done it. So maybe I'm making this up, but I've got to believe that the days of doing that, like manually in spreadsheets and reviewing the level, like now it could all be done by AI and it can be highlighted and it can do it. You know, it can be done exactly how you want. So that's one thing that came up for me.

46:41Clare Bonham: The second thing that came up for me is I remember there being a point in a company's growth where like the playbook was, OK, expand the HR team and hire ahead of total rewards. And you would do that because of all, you know, partly because of that exercise that I just talked about. But then there are all these other things. And I honestly don't think I would be shocked if that is part of the playbook now, because there's just so much you can get. You were essentially bringing a consultant in-house to get you ramped up. And then if you keep growing, you IPO and then you hire some other comp, like deep comp expert.

47:22Clare Bonham: And I feel like that has got to look different now. And the last thing I'll share is like, we were our CEO and founder, we're very early stage. And he was walking us through kind of what the future may look like, and how we may grow. And you know, the reality is like, we're a company born in the AI era. So we're going to grow like an AI company. And, you know, this notion, and these are my words, not not his, like, I'll leave that part of the story behind. But like in my head, what I envision is we companies are going to be growing headcount wise slower and individuals are going to be doing more.

48:05Clare Bonham: And which what brought that what made that pop up in my mind, Seb, is was your comment that like used to have like these manager bands and managers on managers and executive. and like I just feel like we're entering an era of like you know the head of growth at Lovable calls it high c the high impact individual contributor and so those are things that like a year ago I just would not have believed any of those things having been through job architecture review and implementation having seen that playbook that I talked about and then this kind of total disruption of like bands and levels within a company just because of like what an individual can now do.

48:50Clare Bonham: And so I'm going to just leave those here, just sprinkle that to the group here and see if anything pops up for you. Claire, you're off mute.

48:58Sébastien Baehni: Yeah, I think it's fascinating what will happen with the role of managers and the role of individual contributors. And I agree, I think we're all seeing the rise of that type of high individual contributor type role. But I also and I think it will actually maybe make bring management back to being a more disciplined role because I think we got to the the point where a manager was sort of a jack of all trades it's a Swiss army knife it's like you can do a bit of people management you can be a bit of an individual contributor you can do a bit of this and that and you know I think there'll be two things I think you will have people who will who are on management track who opt out as Seb said I think it's really we already think I think we can all see that happening you know people moving off to sort of high paid, high autonomous roles if you're good at what you do.

49:45Sébastien Baehni: And I think that will grow. But I think there are also business models and workforces that will still need management and people management may rise as its own discipline again, where being a good people manager will matter. And I think it's also important that, and I really feel bad now because I can't remember what research I read about this, but I read something there the other week but to reference it but essentially it was about burnout and what does it mean if we automate away all these easy tasks and people always then have to work with the most difficult tasks so let's say you take a normal job and let's say it's a call center and you've automated away all the friction the sorry frictionless tasks and now they're only dealing with like really as humans we're only dealing with the hard problems and actually that can increase burnout and feeling of just every call it's like oh my god or every email you get it's like a big problem to solve and that apparently we as humans psychologically we need some easy problems the easy problems make us feel like we can tackle the big problems so it was an interesting article and I wish I could remember the reference if I find it I'll share it but the the idea was that we have to think about the constitution of work so we have to think about giving to make people still feel engaged and productive at work they need to have that mix of tasks the easy and the hard stuff so don't just automate everything and I think that's what will also become like the next layer of conversation is what does work look like what are the skills and that links to what we're all seeing the change in skills the change in preferences but definitely organizations are reshaping but I also think like any transformation a little bit like when in the old days I worked for a long time so I remember the days when we used to outsource offshore, near shore, you'd always go a little bit too far and then you'd need to bring some back.

51:31Sébastien Baehni: And it was sort of this ebbing flow. And I do think AI will have this ebbing flow of people will reduce workforce, try to be leaner, find out they can't be so lean and it will be a cycle. But I think it's something that's super, super interesting about the world of work right now. And I think more attention should be paid to the split between IC and managers. I think it's fascinating.

51:54Clare Bonham: When I think about our recent humans times machines conversations that we're having with CHROs, there's one theme that keeps coming up. The vision is clear. The pressure is real. But the roadmap is a little bit less clear. It's one thing to talk about the future of work. It's another to sit down with your CEO and say, here's how work, the workforce and our organization really need to change. But I get it. There are a lot of difficult questions that you need to answer. Like, what work should we redesign? How will AI reshape roles and careers? What kind of organization do we need to move faster?

52:31Clare Bonham: And these are exactly the questions explored in Deloitte's latest Human Capital Trends report. There's some interesting findings in there. The report shows that we're no longer balancing tensions. We're at a tipping point. The report brings together insights from thousands of leaders and explores how organizations can build a human edge. It's about moving beyond simply having humans and machines to creating a system where their capabilities multiply. It looks at how companies are redesigning work for this new future, how to shift from a mindset of cost efficiency to want to value creation, and how to move from static plans to dynamic orchestration of talent and technology.

53:09Clare Bonham: In other words, it helps turn big future of work conversations into practical leadership questions for a world that won't sit still. If you're thinking about how your organization evolves in the age of AI, go check it out. It's a great read. I read it a couple of times. So if you want to learn more, you can find the link in the show notes. I wish we could keep this going. I feel like we need to schedule an episode two to keep it going, but we've only got a few minutes left. And so I'm going to move us to rapid fire questions. We ask the same questions to every guest. Since we have two of you here, we're going to do a little bit like back and forth.

53:45Clare Bonham: And so Claire, I'm going to start with you. Question one, what's one piece of advice that you'd give to somebody who's starting in HR today

53:53Sébastien Baehni: great question I think the best HR professionals I've ever seen are people who can hold the context of the situation the human in the room and the data in their mind in parallel and I think sometimes HR there's a tendency to silo into one of those streams but I would say my advice to anybody is learn how to read the room remember a person is at the heart of the decision you're taking and be able to read and understand all the data in that context.

54:22Clare Bonham: Next question. Seb, if your team gave you a superpower nickname, what would it be? I would say maybe a captain metaphor. Because I like to try to explain complex things with metaphors. And if I cannot find a metaphor, it means that I didn't really understand the problem. I feel like there's like a there's like a marketing principle there yeah I love that all right next question this is for you Claire what's a strategy or practice from another company

54:54Sébastien Baehni: that you'd love to steal good one I would like so I think knowledge calibration is really powerful and I say this knowing that maybe I can't be the person to who would be great at this process but I think what Amazon do with that six-page doc and the whole starting a meeting with a reading review is so powerful because in my world as well, I go back to back on conference calls and different topics. So I'm always arriving right on the meeting and to have that minute, to have the context and calibrate with everyone in the group about what we know before we talk, I think is just a great leveller.

55:27Sébastien Baehni: So I also put my hand up, I would be the worst person for writing that six page doc and distributing it on time. So it's maybe more of an ambition than a reality, but I, particularly now and I think with the pace of work and you know individual contributors and everything the whole thing we're talking about having everyone on the same page in a meeting is hard because people are working at like lightning speeds on different topics and yeah so calibrating that knowledge that would be amazing love that last question and this is for

55:56Clare Bonham: both of you if you could think of someone who's just like really pushing the envelope in HR they deserve, you know, the spotlight, like who should we have on the show next?

56:06Sébastien Baehni: It's a great question. I think there are a lot of people really thinking about HR, but I think there is people leading in terms of thinking about AI and really pushing the envelope. And I would definitely recommend that you bring on some of the other voices in terms of people who have been reward leaders and worked with AI, working with AI now and how it's transforming their practice.

56:29Clare Bonham: Yeah, on my side, I don't have a name, but I would say someone that is going through, I would say a governance change like European, I mean, pay transparency with AI would be very interested. I know that also you have also pay transparency in California. So it might be more related to Europe, but going through this time with AI, I think it's a huge challenge. Well, if anybody is listening to this episode, and is going through either of those, hit us up on LinkedIn. We'd love to have you on. And so final tradition of the show is something we call one word or phrase close. We all say a word or a few that are coming to mind for us as we wrap up.

57:14Clare Bonham: Seb, Claire, one of you want to go first? I can start. So for me, control, in the sense that all this power doesn't really count if you cannot really steer it. So you need to know how to control it.

57:30Sébastien Baehni: Yeah, I would follow up. I'll keep it thematic with a C word as well. So I would say context, because I think everything we've talked about today is about decisions and people. And I think anytime you make a decision, the only time you end up questioning it is when you realize you didn't have all the information to hand. So I would say context.

57:48Clare Bonham: I'll go with conscious just to keep the C theme. I feel like with all the benefits that are coming and all the changes, it means we can be more conscious about like what we're saying as managers and doing and like, I don't know. Hopefully that hits. I'm going to say... No pressure. No pressure. I'm not going to do, I'm not going to stick with a C word. I'm going to go with be calm. Every time I read your company's name, it's like a reminder to myself to just chill out. I think it's supposed to be read like be calm, like comp, but I say be calm. So that's my one word phrase. Well, this has been an absolute blast.

58:31Clare Bonham: I know we need to get you all out of here. So thank you so much for joining today. It was a lot of fun.

58:36Sébastien Baehni: It was super. Thanks, Daniel. Thanks, Stephen. It was really nice. It was a great conversation.

58:41Clare Bonham: Thanks a lot. Thank you. Bye. Thanks for tuning in to another episode of The Modern People Leader. We really, really appreciate it. And if you enjoyed the show, please leave us a five-star rating. It would mean the world to us. And connect with us on LinkedIn. We want to know what you think about the show. And you can find links to both of our profiles in the show notes. Thanks again for listening and see you on the next episode.