Didier Elzinga: What survives AI

19:58Didier Elzinga:What I hope that we have done well and will continue to do is keep betting on potential because you're not just betting on what's the person going to be like on day one. And so like one of the lines I used to give my leaders is when we're considering an internal promotion versus an external hire. So you asked me at the start about going and getting a CEO from outside or having a CEO in. I would say I use a six month lens. So don't compare the internal person to the external person, because you're always going to get someone externally who's got more experience. You've got to ask yourself, what's that internal person likely to be six months from now if we give them the role?

20:35Didier Elzinga:Because it's going to take that new person six months to get up to speed and be effective. And that also helps you anchor to people that are on a fast trajectory. You want to find those people internally where if you give them another six months, you're like, actually, I reckon they might be better than the field. put those people in the role. Yeah, that's super interesting. I mean, I think there's also something here, which is you have grown CultureAmp through huge societal change around how we think about what work is. We've talked previously about this notion of like, bring your whole self to work.

21:05Like we grew companies with, you know, orientations around people through a period that we now, you know, like it's disparagingly referred to as a sort of woke era. I'm really interested in like your perspective if you're willing to go there like on what bringing your whole self to work looks like how maybe culture amp both for its customers and for itself is like added nuance to that notion and to what work is to people in a safe bounded way so much to unpack

21:36Didier Elzinga:segueing from the previous conversation one of the things that uh i realized earlier on so there was a point we're talking about growth and i was talking about wanting people to grow and I remember having a conversation. I don't know how big we were at the time, maybe two, 300 people. We want to be a place where people can grow. We want to optimize for that. We want to hire people. And somebody said, well, not everybody wants to grow in their role, like for various outside reasons. Maybe you've, you know, you've just had young children or whatever it is. There's reasons why for the next 12 months, you might not want to push yourself to grow at work at breakneck speed.

22:05Didier Elzinga:And I remember at the time thinking, yeah, they're right. Like we should be inclusive. We should be open. So not everybody's going to need to grow. About a year, year and a half later, I went, actually, I was wrong. You can't be that. You can't be a growth place and make it available. So you actually have to be honest about what you are. Like if you are a scale up or a startup and you are backing yourself that you're going to hire people that maybe don't have the rockstar CV, but they're going to grow like hell, then everybody has to be growing like hell. You can't save this one from others. So I do think it is, you know, if you're trying to meet everybody's needs, you meet none.

22:39Didier Elzinga:And so that becomes a challenge too, when you start to think about all the other stuff. So, you know, we talk about woke and, you know, I find it so funny when people, it's become a pejorative term and it literally just means to be aware. Like how can that be the pejorative thing to be aware where you weren't aware before? And, uh, you know, the benefit of having some incredible people at CultureAmp who I've learned a ton from, Stephen Huang, who was our first head of diversity, Aubrey Blanch, like just incredible people. And Orbs put me on to, she gave me Dr. Abraham Kendi's book, How to Be Anti-Racist.

23:15Didier Elzinga:And that really affected how I think because he makes the argument that rather than thinking of it as you are or you aren't, you look at it and go, are the actions that you're pursuing either perpetuating or dismantling an unfair situation? and so that's been a useful lens for me to sit down and think about when we're doing stuff are we doing things that is making the world fairer more equal you know all of those things or are we perpetuating all of the systemic issues that exist and you can't fix everything and at times I think we've tried too hard to do that and I think it's hurt us in ways but I would look back over what we've done at Coltramp and I've I would feel I do feel and I know from some of the people I've spoken to that we have been, and I think will continue to be a company that when someone needs you the most, we're there in a way that most companies won't be.

24:11Didier Elzinga:But that's different to having the world's best set of policies, the most generous employee profit policies. And I can tell you every employee engagement survey we ran, somebody would go, I can't believe we don't have this policy. This is ridiculous. We're not looking after that. And And I'm like, you can't meet all of those needs. No. And to your point earlier, I think there's sort of some people have an interpretation of this sort of area of being great to people, being fabulous to employees, treating people as they come, recognising we are all human and thinking that that's a sort of opening for workplaces as extensions of therapy, which is also unhealthy both for them and for their peers and for, frankly, building a growth stage company.

24:55and finding the balance around that. I mean, I've observed from the outside CultureAmp play in that space openly, have a point of view, but, you know, adapt over time as you've sort of learned what's worked for you and what hasn't. Was there any hard-won lesson in there?

25:10Didier Elzinga:Yeah, I mean, I've reflected a lot on this as a CEO and as a founder, and I want to come back to vulnerability in a sec. So if I don't remember it, let's go back to vulnerability. But, you know, I think coming through COVID and a lot of the stuff that was going on in the US, there was a need for organizations to talk to their people. And the backlash against that was the concept of the, I think, incorrectly titled mission-orientated organization, where people said, no politics at work. We're only going to talk about our mission. You know, it was a backlash against this need to go and talk to people about all sorts of things that was going on in life.

25:42Didier Elzinga:And I think the challenge is both ends of those spectrum are bad. It's incredibly privileged to say, we're just not going to talk about all of these topics because they're not relevant to what the organization's doing. If for half of your workforce, that's their daily reality. So I don't think you can just say, we don't talk about politics or we don't talk about any of this stuff at work. On the other side, if you say anytime anybody has something that is important to them, leadership should spend the time to talk about them, to say that they've created space for them, to hold space for them. You literally will have corporate comms on them.

26:17Yeah.

26:17Didier Elzinga:You will never get anything done because you were doing, and I do remember thinking in a moment going, I'm doing this literally every other week. I'm having to write a note to the company to hold space for a portion of the workforce. And you start asking yourself, like, am I even the right person to do that? Like, what entitles me to do that? There's a little bit where people want to know that their leaders care and you actually have to care if you're going to tell them that you care. So like, you've got to show that and make that not just performative, but you know, I'm commenting on US political issues.

26:49Didier Elzinga:I don't vote in America. I'm not a US citizen. I just have a lot of people that work there. And so what's the right way for me to engage with that? I've struggled with that a lot. Yeah, it doesn't seem to me that any organisations completely figured that one out either. And it's only gotten harder, not easier since the last couple of years. So let's talk a little bit about vulnerability then. What's always struck me about you, Didier, is that you're deeply self-reflective. You've alluded to a couple of the things, some of the practices you've already quoted from Buddhist literature and many other leadership principles.

27:22You're very thoughtful about who you are as a leader. And also you've talked a bit about, you know, this challenge of running a culture-based organization openly. I'm really interested in what you think the kind of healthy balance of that self-reflection and the openness you have around these topics, what your role is, who you are and the unhealthy part of that.

27:44Didier Elzinga:Yeah. So our first core value is have the courage to be vulnerable. That's like deep in the DNA of the organization. Right there. And it's there because, so I'm going to just step back and talk about through what lens do you measure the success of a culture? In an organizational context, like cultural transmission is to create a better world of work. And what we're trying to do is help organizations create cultures that help them achieve their mission. And their mission is not necessarily save the world. Their mission is to be a successful company, to win in a space that they're choosing. And winning doesn't just mean money.

28:23Didier Elzinga:Winning means being the best in the world at something and doing it at a really high level. And I think that's the first place where sometimes people get unstuck, where they feel like they're choosing between one or the other. And they're like, either you're committed to culture or you're performance orientated. And we've actually just done a whole bunch of research that we're sharing with a performance culture quadrant, where we have this two axis. So we've got like the engaged organization, which is like, how much do you care about the organization and your role and your craft and what you do?

28:50Didier Elzinga:And then the performance culture, to what extent is the company being successful? To what extent is the culture achieving at a really, really high level? and you want to be top of the quadrant. You want to be on both sides of those. You want those two things to come together. And so there's a subtle piece there where when we do a lot of our research, because we're coming from that mission-orientated side, because we're coming from that mission - orientated side, what people are looking for us to do is provide the research that shows why investing in culture is a good idea. And we do. And there's some fascinating findings.

29:22Didier Elzinga:Like we found that implementing a structured performance review process makes zero difference to the experience of men, but it eradicates the historical gap for women on a whole range of topics, like that my role is evaluated fairly, that we're a high-performing organisation, etc, etc, etc. So huge value in doing that from a sort of moral point of view. But actually, what we want to do and what we strive to do is subtly different. What we want to do is we want to go, if you look at the culture of the organisations that succeed at the highest level, these are the sorts of patterns that work. There is no one perfect culture.

29:59Didier Elzinga:So we're not going to come out and say, this is the culture blueprint that makes a successful company. But what we can do is we can show you the cultural underpinnings and the research that shows if you're highly engaged, but not performing, how you make that jump. If you're an organization that achieves at a high level, but is burning people out and it's actually a fear-based environment, how do you make that jump where you go back to connecting with a culture that allows you to play and do all those other things. So we're about helping people do that, but we're always navigating to that top star, which is performance.

30:28Didier Elzinga:And so that's the tension for every company. So we come back to vulnerability. Vulnerability is the heart of learning. It's the heart of innovation, of courage, of all of those things. And I think in a complex environment, you need it. But there's a line in one of Brene Brown's books, which really resonated with me, where she said, vulnerability without boundaries is not vulnerability, it's manipulation. Wow. And so for me, that's the line that you have to learn. Like vulnerability is there to support the people. If you're a leader and you're showing vulnerability, you're doing it to support your people.

31:02Didier Elzinga:You're doing it to show them that you're human. You're doing it to show them that you care. You're doing it to show them that there is more to life than work. One of the lines that I've used a lot is we borrow our time at work from people or things that are more important than the work is. But I say that as a workaholic who is dedicated to winning at the highest level. and you have to have both. The fall over is when you're using that exposing to try and manipulate the other person, or you're trying to go, you know, don't give me a hard time. Can't you see that I'm hurting here or whatever it is?

31:33Didier Elzinga:And so that line goes through my mind a lot in terms of how we manage it. And it's made that much harder when things are online. Yeah, I can imagine. Well, actually say a bit more about that. Why harder online? Well, I was talking to our head of comms about this actually, Han, she made this really great comment too, where she was saying she caught herself, you know, you and I face to face, there's so many more, so many more dimensions of bandwidth when I read that differently and I'll, I'll react to it differently. But I remember this, um, this cartoon where it's like somebody says, Oh, come to bed.

32:02Didier Elzinga:And he says, I can't, somebody said something wrong on the internet. Tap, tap, tap, tap, tap, tap. And so there's just like, you see something said, but you're not responding to a human. You're just responding to this thing. And we see it constantly. like the rage baiting, the, you know, the overly emotional response, they're just, you know, flaming each other. Online only isn't conducive necessarily to, to, to, to great relationship skills. And so that's a challenge in organizations, like how we keep our slack conversations civil and so on. But if we go back up another level, you and I were talking about this earlier, I think this comes from therapy.

32:37Didier Elzinga:I'm not sure where it comes from, but the idea really resonates with me is you want to share scars, not wounds. And so the way I internalize that as a leader is if it's still a wound and you're still hurting about it, chances are you're actually probably not going to share it in an effective way for the team or for the other people because it's still raw. Once you've kind of processed it, made sense with it, extracted the lesson, it can be really valuable to go, hey, let me tell you about how I screwed that up and how I was feeling and why that was difficult for me, but I'm doing it from a somewhat centered place.

33:08Didier Elzinga:And so that's actually been a line that I've used in my own head a lot when I'm feeling an emotional reaction. It's like, what am I sharing here? Am I sharing a scar or am I sharing a wound? It strikes me, and I've heard this in some of your research, actually. I think you've done quite a lot of research on authenticity and the importance of authenticity and leadership, which is like, you know, anyone can say or put out a perfectly worded statement, certainly now with AI, like definitely possible to do that at scale. But people pick up pretty quickly whether there's an authenticity behind that.

33:37And it seems to me that the last five to 10 years of kind of corporate reckoning has been around this notion of vulnerability, but not always in the format where like the person expressing vulnerability is doing so with authenticity. Or as you say earlier, like Renee Brown saying it can be manipulation. It can have kind of an emotional undertow that doesn't carry the honesty that people are actually seeking out of that. Can you share a bit about what you've learned through CultureAmp's experience with customers around the importance of authenticity? There's two different parts of communicating and authenticity works at two

34:14Didier Elzinga:different levels. So I think one is, you know, I used to work in Hollywood, so I'm a big believer and fan of storytelling. And I think as a leader, a big part of your job is to be a storyteller. And so part of storytelling is to make the other person feel. And you want to connect somehow emotionally with the audience. And that audience might be your employees, it might be your customers, it might be the world at large. And that is a skill. And you see it in musicians, you see it in actors and everything. The other part of authenticity as a leader that I think is really important and probably underlooked is, and there is a framework that I learned from my wife Greta with four C's.

34:52Didier Elzinga:And I can't remember what all the C's, I think it's clarity, something else, something else, but one of them is consistency. And so the reason why I think authenticity is so important for leaders is because it helps people know how you're going to show up. Because a powerful culture is what happens when you're not in the room. It's when somebody goes, oh, I'm going to make this difficult decision. Like I'm going to walk, you know, I'm going to put this customer at risk, or I'm going to spend the company's money, or I'm going to do something that somebody else would look at and go, that's risky for you to do that.

35:23Didier Elzinga:because I know that this is in line with the values of the organization. And I've seen the leader act and show out authenticity. So I know what to expect of them. And I actually remember years ago talking to a guy who told, and I won't tell the whole story, but he had this moment where he basically had to walk away from an important customer that this person had. And he went to his boss and basically said, I had to do this. And he was sort of not sure which way it was going to go. And the leader just looked at him and said, yep, you did the right thing and thought nothing on it. And he said in that moment, he had him for life because he now knew how he was going to show up in those moments.

35:58Didier Elzinga:So for me, that's what's important about authenticity is you're helping people understand how you're likely to show up because if it's authentic, it's true. It's going to persist. It's going to persist. Let's go one level deeper on this sort of role of, you mentioned earlier, online only and also this sort of authenticity as a leader. I'd love your reflections on, so Mark Zuckerberg recently kind of came out saying that he was contemplating, I think, building an AI avatar of himself as a leader that people in the organization could interact with. What do you make of it? I think we're going to see more of it.

36:32Didier Elzinga:I mean, this is actually one of the areas of research for me right now. Like I'm deeply curious about what is the role of the human in the world of work when we have more and more AI. Like one of the things I thought, it was kind of funny as I was stepping out, part of me is like, oh, you know, maybe I want to have the experience of running a bigger company than Colchamp. And, you know, at our largest, we're, you know, a thousand people. And then I had this other thought and I'm like, actually going into the future, there may actually not be any point knowing how to run a company that's that much larger than a thousand people.

37:05Didier Elzinga:I may have actually already had the experience that's going to cover 90 % of the companies we're going to build into the world. Cause I think companies are going to be smaller because that's what AI makes possible, but those ones are going to be more leveraged. So what is the role of a human in an AI world? How do AI and humans work together? What's the role of AI in amplifying humans, et cetera. And the shift that I think we're going through at the moment, and once again, something I'm passionate about and I'm really enjoying learning about in my non-CultureAmp days, is so much of AI is almost individual now.

37:39Didier Elzinga:Like what an individual can do with Claude or ChatGPT or it was incredible, but not enough of it is multiplayer. So not enough of it is how we're working together. And so I've actually, some of the stuff I've been building on my own is like, rather than thinking about it as your assistant, thinking about it as the team member you always wished you had. That's in every chat going, how do I make this chat a better chat? How do I put the spotlight on the person who hasn't had a chance to speak? How do I bring it back to the person who should be involved here? So I think we're going to see more and more of how do you amplify an individual.

38:12Didier Elzinga:And of course, CEOs are going to try and figure out how to do this. It is interesting though, isn't it? I mean, there's something there about, we are in this sort of very early phase where we're thinking about AI agents, as you say, kind of like individuals. Like our frame of reference is I'm a person with a given job, a scope of work. How do I get an AI agent to help pick up parts of my scope of work? Which in one sense also just treats like an AI, which could be in many different formats as an individual rather than like actually I'm choosing to have the AI think in a multidimensional way that's actually very hard for any given individual to like it'll be interesting to see in like 10 years time whether or not the agents we build actually don't look at all like what a human version of a computer could be.

38:58Didier Elzinga:I love that you said 10 years I mean wherever we say 10 years we probably mean 10 months. That's right. Because like the world is moving so fast but you're 100 % right. Like I'm, I keep flip-flopping on this because the prevailing model at the moment for people that are building agentic workflows is largely to say, let's think of agents or swarms of agents as, uh, jobs. And, you know, there's this great, um, uh, open source project called, uh, paperclip, which is to run a zero employee company where all these agents play the role. So there's a CEO and a CTO and an engineer and a marketer, and they talk to each other and they do stuff.

39:34Didier Elzinga:And firstly, I'm not interested in building a zero employee company. I'm interested in what does it look like to have a hundred employees and a thousand agents. But secondly, I am very interested in, do you make them jobs? Are they jobs to be done? Are they some alchemes of humans? Are they something else? And I keep flip-flopping and trying to figure out how to make that work. And I'm building some things to play that out in my mind. And I think it will change a lot over the next 10 months because at one end, people have got agents as essentially just like little snippets, do this and create a presentation.

40:08Didier Elzinga:That's not very interesting. What we're talking about is the long-running memory context aware thing that we can start to converse with and interact with. So, yeah, that's the watch the space, and that's why I'm actually excited to be an exec chair rather than a CEO right now. Totally. It feels to me like there's two dimensions here. we're constrained both by, you know, capability, like these tools, as you say, in 10 months time, like what is possible with these tools will completely shift, but also by our own imagination of how to use them. What's like one thing you've been testing out that's right at the outside of the boundaries of imagination?

40:42Didier Elzinga:I love that you use the word imagination. And it's something that I've found myself saying a lot, which is because I think we're on this exponential curve. And so particularly in coding, but also in long running agents in general, there was a step change between November and February. So people were doing things middle of last year and they kind of worked, kind of didn't. And then suddenly stuff just started working at a level that is mind blowing. And so because it's not an exponential curve, somebody that might only be a week or a month ahead of you is 10 times further ahead. And that's really hard for us to get our brains around.

41:13Didier Elzinga:And the one part is the tyranny of, or the information asymmetry. They just know stuff that you don't know. But I think the bigger problem is the imagination asymmetry. Because they're already there, they're now thinking about things that you're not even putting it to. So what I've been playing around with for myself is how far can you take agents when they largely just bootstrap themselves? You know, what happens when you give them fairly basic tools? And this is like the open claw Hermes agent type world. And rather than trying to over-specify stuff, you just keep redirecting them to update themselves to get better at it.

41:46Didier Elzinga:And it's amazing. It's scary. What's your favorite prompt in doing that? Because you mentioned this challenge around you don't want to over-specify because by definition, you're then constraining the outcome and you may not know what the best version of that outcome is that you could collectively come to. Have you learned anything about the best prompting version of the redirection without the over-constraint? So I think that the trick to do this is you've got to move from single-shot prompts to long running agents. And like the skills model has kind of unlocked this because what you're doing is you're kind of creating repeatable ways of doing stuff.

42:20Didier Elzinga:And so an agent can still be just written as a series of prompts, but you start with the simplest possible version. So you go, oh, I'll give you an example. So in the thing that I built for myself, I have this concept of a librarian. So the librarian basically looks at everything else and makes it better. So to start with, it's just simply two lines. Like you are a librarian, this is what you do, but then you tell it to bootstrap itself. So every time it runs across and does stuff, it compiles what it learnt into a bunch of lessons and then it reads those lessons and then it rewrites its own prompt.

42:52Didier Elzinga:So the trick is to get it to write itself and watch what happens. And it's actually really, really fascinating. It's both incredibly bright and incredibly stupid at the same time. But it is mind-blowing. And then the imagination starts to open up and you're like, what if you could do this? What if you could do that? And I think the most useful thing that it's helped me understand is that the usual answer to most problems is just to go meta and go back to the AI itself. So rather than trying to over-specify it, go back to it and go, oh, why did you do that? You know, extract for me the lessons of that and then let me interact with those.

43:28Yeah. Interesting. Which, you know, you mentioned earlier, like it's almost impossible to have the space for exploration and the, you know, going up the imagination curve, use your sort of way of conceiving it while also running a thousand person organization with all the pressures that CultureAmp is under.

43:43Didier Elzinga:Even running a 20 person organization. Indeed. Yeah. So actually let's blend those things together. So, you know, we can't be having a conversation about CultureAmp in this day, in this exact moment without referencing the sort of SaaSpocalypse as it's been termed. I guess first question for you is like, are you glad CultureAmp's not public at this moment? Yes, a hundred percent. But I think what's happened, the private markets are not immune either. So, you know, whatever happens in the public markets will go to the private market too. You're just... Especially for large companies that are like held to comparables and all that sort of stuff.

44:17Didier Elzinga:Like it's a really hard world for any company north of 50 million in revenue. In fact, it's any company with revenue, unless you're an AI company growing at 3000%, it's a really, really hard place to be. And what's shifted is 15 years ago, no one gave SaaS the multiples they had. because they were like, oh, I don't really understand the model. And then sort of probably 10 years ago to five years ago, people said, go, this is a beautiful business model because it's a monotonic revenue curve. People have got positive net revenue retention. So even if they don't add new sales, the business is going to continue to grow, but then you add new sales on top of that.

44:53Didier Elzinga:So I've got a pretty confident that this is going to grow at 20, 30, 40 % for decades. that's really valuable. And so it got valued accordingly. So it had a long terminal value. As AI came in and everybody started going, maybe that terminal value is zero. And so then they changed all their models. And so then all the valuations came down. And at the same time, growth got lower. So everyone's like, well, maybe they're not going to grow at 20%. Maybe they're going to grow at 10%. And we value companies that grow at 10 % in a different way. So all the multiples have come down. So there's a lot of reckoning going on.

45:26Didier Elzinga:underneath those businesses, still great businesses. They're, you know, generating large amounts of revenue. Most of them are generating cash. And so maybe say a little bit about that to the extent that you can. What is CultureAmp's AI play? Like, you know, it's easy to think about starting a company from scratch. Like, is the mental model for you at CultureAmp, imagine we weren't a thousand person organization with hundreds of thousands of customer relationships and we cared about making a better world of work. Is the mental model like, what would we build if we were starting from scratch? Or like, how do you think about both the pros and cons of having this huge business that's been around for 15 years with all these customer relationships and legacy tech and legacy ways of operating?

46:05Didier Elzinga:You definitely have to disrupt yourself. So you do at some level have to sort of create an opportunity to look back at yourself, create a small group and say, reconceive what this could be. But you also want to take advantage of what you have because you're not a five person startup. You can't just vibe code the back end. Like we have a huge amount of privacy and legal stuff that we need to do for all of our customers that we've spent a lot of time figuring out how to do well. So we've got to preserve all of that too. One of my board members has a good model, I think, for AI companies where he said there's three dimensions of value.

46:37Didier Elzinga:There's data, which is the one that most people think about. Like what proprietary data do you have that other people don't? And why will your model be better because of access to unique data? Most people aren't building their own models. or if you're building your own model, you're actually largely doing it for cost reasons. You're doing it because you don't want to be spending, you know, Opus 4.7 tokens. You want to be using your own models way cheaper. But the data that you have is a context layer that you can bring to bear that other people don't have. So Coltramp has, you know, over a decade worth of data that we can bring to bear in the organizational work context, which is super value.

47:08Didier Elzinga:So you've got data. Distribution is another one. Like just because you can build a cool AI tool, you have to still get it into the hands of everybody. Yeah. So incumbents actually have distribution. And then the last one is design. So what is the way you interact? So granola for me is the canonical example here. Does granola do better transcriptions than anyone else? No. Like their summarization is pretty cool. But the thing that they absolutely nailed was the interaction pattern. I agree. And so like as an incumbent, those three things are things that you can be better at than a startup. The thing the startup has you don't have is the imagination.

47:44Didier Elzinga:And so it's actually one of the things when I'm talking to industries and to different organizations, there's a whole thing of people going, you know, don't do AI transformation without a clear business case, you know, not what you're going to change and all that sort of stuff. I'm like, honestly, I think you've got to be in there and doing stuff because if you've never used the technology, you've got no idea what's possible. You can't write a business case for a language in a language you don't understand. And so for most organizations, they should just be doing stuff. And that's where the imagination comes from.

48:13Didier Elzinga:So for CultureAmp, we obviously have a very strong, sophisticated, complex workflow tool for managing employee listening, for performance management, for all of those things. We'll continue to develop that and we can now develop it faster and better than we ever have before. And I think it is a great time to be building product. Like product feels magical again. It didn't feel magical three or four years ago. But the opportunity on top of that and the next wave of that is what we call the Culture OS, which is you think about in all those interactions, in all those patterns, there is a context layer that can be brought to bear.

48:48Didier Elzinga:So over the last year, we built our own sophisticated coaching platform that sits in Culture AMP. And so when you close an engagement survey, it helps you understand it. It helps you figure out what to do. It helps you write emails to your team. It helps you get to action and navigate your journey on that. when you close a performance review. It helps you write the performance review. It helps you get better at giving feedback. But the next opportunity is to connect all of those things and to actually move it from being something that's kind of reactive, if you will. Hey, employee engagement survey's closed.

49:17Didier Elzinga:Can I help you with it? To being proactive. Because if you think about what's going to happen or what's already happening is every person who's an IC today is basically going to become a manager. Of agents. Exactly. And that's going to mean they're going to be context switching. They're going to have to be prioritizing. They're going to have to grapple with and build a bunch of skills that they haven't had to have before. So we see a huge opportunity to meet them where they're at and take them on that journey. And then every manager has to become a 10X manager because there's going to be less of them.

49:47Didier Elzinga:Almost all organizations are flattening. And if you think about productivity, if your span of control is roughly five to one, if you can move that span of control to 10 to one, because you can now use AI, if you look at the Jack Dorsey kind of concept of being the context layer, like a large amount of leadership and management is basically information routing. We're not that far away from AI being better at that than the managers and leaders are. That doesn't mean that managers and leaders don't have a job. It just means that they need to be focusing on other things and they can probably support people in different ways.

50:16Didier Elzinga:So what is the context layer that needs to exist to allow those managers to be superstars? That's what we're building. Very few people are in a position the way that you are, Didier, with the 15 years of seeing lots of different organisations. Like how many customers does Cultura have right now? Just under 7 ,000. 7 ,000, right. Like all over the world, every type of customer. What's your prediction on the shape of organizations in the coming months, even years in an AI enabled world? It's a great question. The question that everybody's worried about. The challenge for all of us is none of us quite know how fast this is going to happen or how it's going to play out.

50:52Didier Elzinga:And the faster it happens, the harder it's going to be. So there was two people at Davos, two of the AI people. I think it was someone from Google and Microsoft, and they got asked this question of, when do we think we're going to hit AGI? One person said two years, the other person said five. And then what they both agreed on was that it would be better for humanity if it was five than if it was two, because we're just not ready for it to happen in two. We don't know. So I'm worried about what's going to happen over the next couple of years. I think it's going to be bumpy because I do think it's going to be a huge dislocation because what AI is going to make possible is a flattening of organizations and we're seeing it now.

51:29Didier Elzinga:And the big challenge for organizations is that with all the economic pressure that they have, they're going to need to get smaller and they will get smaller. And AR will mean that they can do with six people what they used to need 10 for. But you also have to be thinking about how does AR unlock the value in the people we have? Because that's the only enduring advantage is not how do you get smaller, but how do you do more with the people you have? And how do you enable those people to achieve things they couldn't achieve before? That's the reckoning leaders are all going to have to deal with.

51:55And I'm interested actually, you know, for customers of yours, let's say, that have had organizations, let's say they're currently 500 people and actually over the medium term, they're going to get down to one of these like one to 200 person organizations achieving the same amount or more. Is it fair to say, do you think that actually the smart ones are trying to capture context and the way that people make decisions so that they can do a better job of building the agent? Like, is that, it sounds cynical, but is actually like that the best play for an organization thinking about its future?

52:23Didier Elzinga:I mean, you raised two really good questions. One is, and I think this is the hardest thing for leaders and organizations to do right now, is that you have to sit down with your people and figure out how to co-create what this future work looks like in an AI enabled world, knowing that that new world will probably need less of the people that they have. How do you ask someone to participate in that? A lot of organizations are not saying that openly, but doing exactly that. People are still coming at it in an old school way. They're like, let's sit down and process map everything that people do and then create essentially a deterministic system that will do that.

52:57Didier Elzinga:That won't last long. And so it's going to have to be a process of how do we co-create with AI and humans a system that can work two years from now, not how do we do what we were doing a year ago. And the challenge that we're seeing when we talk to all these organizations is if you go to people, most people will only say, okay, I need AI to help me get this little bit from here to here. The best stuff's going to be zooming right out and redesigning the whole value chain inside the org. And so that's what orgs are doing. It's just keep coming back to organizations having to think about, they are probably going to be taking people out of organizations, which is hard.

53:27Didier Elzinga:And it's going to be really bad from a societal point of view. I'm like quite worried about what this will do at a societal level for the next little while. If you're a 500 person company and you could run it with 200 people, your perfect world is probably to be a 300 person company doing twice what you were doing before. Whereas most people are going to get to the 200. And I think that's the challenge we're going to have to walk. Well, I'm also interested though, because I mean, with 7 ,000 customers, you must be seeing a huge spectrum in AI adoption, right? So we live in the world where everyone's like on the latest models, experimenting with the latest tools the week they drop.

53:58That is not the common experience of most organizations worldwide. There's a lot of fear that drives that. There's a lot of inertia. There's a lot of lack of access or lack of understanding even of what the tools enable, lack of imagination to use the language we were using earlier. What's your prediction actually on this speed of change. Because my sense is it's going to go really, really, it's going to go like wildfire through our world because of that kind of access and imagination. But it's going to be much slower and arguably more tricky for large-scale enterprises that are just not seeing this change come.

54:30Do you agree?

54:31Didier Elzinga:Yes, I agree. I mean, there's the full spectrum. It'll spread out. And the way to think about it, we're on an exponential curve. So once again, people only have to be a little bit further ahead to get way more value. But I think when you're on an exponential, the bit that you keep coming back to is that things are fundamentally different in a short period of time. But this idea that coding fundamentally changed between, say, November and February. The models that we're using today were largely written by the models themselves. That wasn't true four months ago. That's a fundamental shift in what's happening.

55:05Didier Elzinga:So I think it's going to go faster than people think. So the other day I saw something that I had yet to see. I'm interested whether or not I was just, you know, head under a rock. I saw someone in a global business win the title of chief people and agent officer. I had never seen the job title written as such. And it's so curious to me that, you know, it was clearly an intentional decision to signal that I'm not saying agents are the same as people or nor that they have the same value to me as an individual. I might think quite differently about how I quote, manage those people and support them or support those two different entities.

55:41But this idea of like, to think of Justin in your own organization at CultureAmp having agents as part of his role type and like, what should we be thinking about for that? Like, do you think that's just the future of all organizations that we're just going to see, you know, we're going to see people be responsible. Like your core customer has always been the chief people officer in some senses. You know, what are you seeing in that universe?

56:03Didier Elzinga:So, I mean, I love that you asked that question. I think part of it is what we were talking about earlier, which is people are just like, ah, agents are kind of like people, so let's put them together. There is a big question going on in terms of who's driving this transformation and who's helping organizations. And I think there's a pretty good argument that chief people officers, if they're prepared, have as good a shot at that as anybody and absolutely need to be part of the conversation. And so that's part of what we're talking about when we go out to CPOs and we go, who's driving your AR transformation?

56:31Didier Elzinga:And they're often like either no one or maybe the head of IT or maybe the head of engineering. And we're like, well, they have a really important role to play, but so do you, because this is ultimately a people problem. Well, not even a people problem, this is a people opportunity. Where I think that the mistake is just to think about it as like, cool, now it's HR's job to manage a whole bunch of agents. Because that reduces HR back to what it was before. The bigger opportunity, and we think about this concept of a context layer. And so for anyone who's listening, doesn't know what that is. Like, AI is a super powerful, but they're best when, like, even though they're trained on all of this information, to make them useful, you kind of have to help them understand what's the relevant information for the problem you have right now.

57:12Didier Elzinga:And a lot of the power is in taking a very powerful frontier model and then giving it the right context. And so in an organization, what's interesting now is we have the ability with our AI systems to be able to, first of all, update, synthesize, and understand that. So if you think about the values of an organization, if you think about the performance culture that organization has, up until that's been largely implicit and probably not done everywhere. So the culture isn't consistent. The alignment to strategy isn't consistent. What we can do now with what we're talking about was a concept of a culture OS is we can start to identify that.

57:49Didier Elzinga:And so the head of people's job goes from being a reactive firefighter to being somebody going, this is almost a steward of this is what our performance culture is. And this is how we want the agentic systems to be imbued by the culture that we have at this organization. And that's a reflexive conversation. It's not them writing something once. It's them working with something like the culture OS to go, what is our culture? Not just what we want it to be. What actually is it? How do we shift it in the places we need to? Where are the leaders we need to help? How do we drive this? And then how do we step back and look at it as a whole and go, oh, this is what's going on in our culture right now.

58:23Didier Elzinga:Same thing for strategic drift. It's like fascinating time. It's actually going to be a lot of fun to run companies over the next few years. Oh, totally. And I think there's also something there about like, what is the data that you're picking up? Because traditionally CultureAmp has been like surveys, like started out with surveys. You've experimented with lots of different models about, you know, pulse surveys and structured surveys and how all of that is done. But it's largely been based on the presumption that at large enough scale, asking people their experience of work has value. In a world where it's not all people anyway, but also where actually the context that matters lies in the format of Slack conversations or the way that pull requests are reviewed by others, including other agents.

59:05And like the kind of the data layer to inform what is the culture and what's working and to use your language, you know, where do we need to grow into? So maybe this is a part of the culture that's perfectly matching, perfectly aligned with what we say our culture is. and then we've realized, well, that's recursive and we need to grow. So I want tension in that culture model. You know, I'm really interested in like what data do you imagine, again, not the 10 years but 10 months version for CultureAmp, like what data layer are you going to be bringing in?

59:34Didier Elzinga:This is a question as old as time. Like ever since I started CultureAmp, everyone's like, oh, surveys are dead. And the thing is we didn't create a survey company and I don't consider us a survey company. Surveys are an active way of listening. Passive listening is super powerful, but it also comes with its own challenges, not least of which the privacy element. Like how, you know, everybody goes, well, all this stuff happens in Slack. It sort of does and it sort of doesn't. And do people show up differently if they believe that this is being, you know, watched and interpreted and then who owns the summaries and all that sort of stuff?

1:00:09Didier Elzinga:All sorts of interesting questions. But the future is absolutely both. And the two work hand in hand. So on the one hand, you use active listening like CultureAmp to periodically go, let's go get a snapshot of how everyone feels, because it's actually a very efficient way of capturing a very high signal knowledge. And now with agents, we can actually take that to a whole nother level too, where maybe you use a survey first, and then off the back of it, we want to deep dive with something. You can actually go do a much deeper dive. And so something we support in the platform is 360s. And if you think about the way 360s used to be done, that were only done for execs and you'd have a psychologist that would go and interview all of the people, some peers and ups and downs and so on.

1:00:49Didier Elzinga:And then they would write a detailed report and then they would sit down with the person and they would go through that report with them. We can do all of that with an agent now. So an agent can go and interview people, doesn't just have to be a question. So we can get really good active listening. Yeah. And then we can say, okay, cool. Here's our picture. Now let's use passive picture to track for drift and to see things that might happen and to follow along and figure that out. or you do it the reverse way and you go, you use passive listening to identify something. And then you go, Oh, something looks like it's going on here.

1:01:18Didier Elzinga:Let's grab a group of people and go find an active, um, active listening on this so that we can bring that to bear as well. So the future is going to be both. Super interesting. It's pretty obvious to anyone listening to this, how reflective you are as a leader, both about yourself and what role you play. You've also experimented over the years with lots of different ways of spending your time. I'd love if you're willing to share a little about some of the practices you've experimented with over the years that you think listeners might also get something out of, particularly probably for founders, but not exclusively for them.

1:01:49Yeah.

1:01:50Didier Elzinga:So I think as a founder and as an executive or CEO, one of the first things is just realizing that often the very things that have driven you to be successful are the things that are making you unhappy. So, you know, that strive, whether it's perfectionism, I don't actually think I'm a perfectionist, but that deep competitiveness, all of those things are why you are successful. They're also why you have that restlessness, why you don't feel connected. And so it's coming back into which of those bits serve me, which of those bits don't, who is it that I want to be on the other side. But if I think about actual practical things that I've done, um, that I've got a lot of value out of.

1:02:34I think breath work as a, as a spectrum has been super valuable.

1:02:39Didier Elzinga:So I went down the rabbit hole of, of cold plunges, which I know a lot of people have. And I actually do it because I really enjoy it. Like whether it's, there's physical benefits or not, I don't, doesn't actually bother me. I do it because it changes the way I feel. And so I can go do a, you know, two minute or three minute plunge in six degree cold water. And I come out of it with a mindset that I find really valuable. But the reason I think it works is it's basically all about breath. So you get in and then you're super, super cold. And so you have to get your breath under control. And so I use a four, seven, eight breathing.

1:03:11Didier Elzinga:And it's the process of getting, you know, having a stress response, invoking the breath control, using the breath control, getting my mental model to where I want it. And that itself has been super powerful. And so I've now gone down the rabbit hole and I do breath work two, three times a week and learning all sorts of different modes. And, you know, if you've ever done any meditation, there's always this thing like, you can always come back to your breath, you know, no matter what's going on in the world, no matter where you are, the breath is always there. It's such a powerful tool. And, uh, I, I learned this from my, from my wife, Greta, uh, the simplest version of this is if you make your out-breath longer than your in-breath, it invokes the parasympathetic nervous system, which puts you into the, away from the fight flight model.

1:03:55Didier Elzinga:And so just that simple model of make your outbreath longer is probably the most useful thing I've ever learned. And maybe to be super tactical, like, do you intentionally do it before stressful meetings? Like, is that how you think about it? Or is it more like I do it enough times during the week that it becomes more, you know, habitual for me as a mode? It's less about, I mean, I do occasionally use it like, I'm feeling stressful. I want to do this thing. I'm going to do it now. Did you do it before you came onto the Wild Hearts podcast? I've done enough podcasts now, but I don't do it. But it's more, someone once said to me, you meditate, you meditate so you're ready for the worst day of your life.

1:04:32Didier Elzinga:You don't meditate on the day. And so it's actually the practice. And that's why I like, you know, doing the cold plunge and getting my breath under control is teaching me and my body that when I have that stress response, that I can manage it. So that when I'm stressed, I'm like, it's okay, I know how to manage this and I can deal with it. and what interestingly what I found was that when I was doing some of the breath work after stepping out of the CEO role is I actually felt quite emotionally unstable like and I realized that was just me unblocking some stuff that I you know as any founder or CEO is listening probably feels it's like you just push everything right down like I don't have time for that I'm not interacting with that I've just got to keep going and so I had that moment to be able to kind of go well, if not now, when?

1:05:19Didier Elzinga:And so it was actually interesting in that way too. It's about sort of connecting back to the more human version of yourself. Okay. Didier, final question. What's something you've changed your mind about? I think this is a lesson I've had to, I'm still learning. I'm very motivated by potential and seeing the potential of what I think people could be. It's not necessarily that I've changed my mind, but what I've realized is that as a CEO, I wasn't serving the organization as the CEO by constantly trying to get people over the line. And what I mean by that is that I'll see somebody who's not quite where we need them to be.

1:05:58Didier Elzinga:And I'll think I can help them. I can help them get to the level of performance that we need or we require. And I'm a big sports fan. And one of the things I'm always fascinated by is you see the top level coaches. They don't have the time to do that. They see someone's potential, but they're like, Either you perform and then I'll give you more minutes or you don't perform and then I'm not giving it to you. And I've had to learn the hard way to be more like that. So the conditions don't intervene. And it's up to them to get to the line. And if they can get over the line, then Didier can go help that person try and get to the line.

1:06:33Didier Elzinga:But the CEO can't. The CEO has to basically spend all of their time on people that are over the line and helping them be successful. it's it yeah so that that's not I didn't change my mind on it but it's the thing that I've had to constantly fight myself on coaches they don't have the time to do that they see someone's potential but they're like either you perform and then I'll give you more minutes or you don't perform and then I'm not giving it to you and I've had to learn the hard way to be more like that to be able to conditions don't intervene and it's up to them to get to the line and if they can get over the line, then Didier can go help that person try and get to the line, but the CEO can't.

1:07:14Didier Elzinga:The CEO has to basically spend all of their time on people that are over the line and helping them be successful. It's, yeah, so that's not, I didn't change my mind on it, but it's the thing that I've had to constantly fight myself on. For it time. Thank you. Thank you. This was fun. It was really fun.

1:07:35Thank you so much for joining us for another episode of Wild Hearts. If you want to learn more from other ambitious people building, designing and creating the world that we all want to live in, then please hit the subscribe and follow button. This podcast is a labour of love from the Blackbird team and Day One. The show is produced by Camilla Herring from Blackbird. Our marketing genius is Laura Cofford. And our editors are from Day One. Andy Jones Sanjay Chabaria and Georgia Catalan thank you all so much for listening and I'll see you all next week