Brainfood Live On Air - Ep397 - Calculating Employee Lifetime Value (in Era of AI)

0:05That sounds like 10 days. I was 14. If we started from Southampton. So it was 14 days. Yeah. So classic place to exit from a crew is it is probably one of the few places you can do it, isn't it? Southampton. The main one in Britain. Yeah. And then how long did it take you to get to Italy from there? Oh, I think it was probably Italy would have been the halfway point. Italy would have been a week. so it was like two days from um two days from Southampton to Cadiz and then probably then we were Barcelona Toulon La Spezia and then Alicante and back again yeah basically first stop was in Cadiz right nice yeah it's lovely tapas brilliant I love you know what I've never been on a cruise and it just struck me I should do that.

0:57It's kind of, yeah, a good way to see a bunch of places and to just chillax and not have to worry about sort of logistics of getting there and all that type of stuff. I mean, I'd highly recommend, like, I've obviously got two kids. I'd highly recommend if you're not taking kids, go on an adult cruise because kids monopolise the swimming pool and everything. But, you know, there's lots of adults-only areas in it, though. I loved it. It's just terrific. yeah brilliant how how are you thinking about cruises folks are you are you involved in cruises do you do cruise holidays let me know in the comments what you think of a cruise is it a good thing to do i wonder whether i could do it and just continue working imagine doing some brain food lives from a different port and all that you know maybe take three months on a cruise you know some people just ridiculously block book a massive amount of time continue a lot of people a lot of people just live on cruise ships now right the problem is the internet connection is not brilliant if you've got a broadcast it's not good enough yeah this is where you need basically a portable starlink type of thing um so a mate of mine's bought one of those literally carries it around in his backpack um and apparently that does the job so perhaps that's uh that's the pattern um all right listen welcome everybody to brain food live on air bringing it to you every friday no fail um we're delighted to have this conversation with you today uh simply because It's something that it's kind of been on my mind for a little while.

2:22I saw an update on this, and it was from our sponsors from Greenhouse. It was calculating employee lifetime value. So how do you calculate the lifetime value of an employee? You hire the person. What kind of value return are you going to get from that person, not from year one? We're talking about the absolute amount of time this person is going to spend with your company. And there's all kinds of weird and wonderful ways in which that can be done. And this is the reason why we've got Tim Sackett on the show later. We've got Jessica Swan on the show. They're going to tell us how it's been done in the past.

2:58But the big curveball for me is obviously we're in the AI era, Adam. And AI is changing jobs as we see them morphing in real time. So how are you even going to calculate the lifetime value of a person when you don't know what that person is going to be doing in the next month? you know plausibly you could say right a recruiter coming into my company a lifetime value of this they're going to be around for five years they're going to make like 50 hires a year however many it's going to look like this i can calculate a number uh but is it even legitimate to try that now that we don't know the impact that ai might have on the productivity of that person or indeed over the the value of the job itself like did you need to hire that many people over a course of time.

3:44So how do we actually do calculation of employee lifetime value in the era of AI? That is the topic of today's conversation. If you're interested in this, you're in the right place. I'm just going to do a few sound checks here to make sure that you are in the right place. So folks, if you're watching this on Crowdcast, let me know if the audio and visual is fine. I think you are because you're talking about cruises there, which is good. We should be live streaming this on LinkedIn. I think several LinkedIn's are on the go. My LinkedIn, I think Adam, your LinkedIn as well. Rob Walker's blasting this out as ever.

4:17Hats up to you, Rob. If you can hear and see me on LinkedIn, let me know if the audio and visual is fine, folks, because if so, we should be ready to go. Let me just quickly check. Am I actually live? Yes, I am. Okay, that's fine. Okay, good stuff. Very good stuff. Thank you, Michael. Thank you, Tim. Thank you, Rob. All good. Right, we're going to get on with it. We're going to talk about greenhouse in a second because they actually came out with this uh report which i'll share in a moment one of our sponsors of course brilliant ats make sure you check them out but more importantly they're just churning out brilliant bits of content which is super valuable for us and conversation style for us so i'll share some of the links of those uh in the chat in a minute uh but whilst we have it we have adam gordon back from this holiday great to see you adam um and we were just talking about cruise ships and why that's a wonderful place to take a family anyway Adam let's talk about the newsletter did you read it have you read have you have you been reading it um and if so what was interesting if anything yeah I have been reading it absolutely so let's start with the article which is all about um the state of open source um was it Mozilla who put this together yes brilliant I mean my my takeaway from that is um AI models are becoming cheaper, easier to swap out, they're becoming quite commoditized.

5:41And actually, the organizations that win are not necessarily going to be the organizations that have the best AI model, because that's not really competitive differentiation anymore. To me, the big shift is the ability for an organization to be able to capture all their nuances and their data and their context and to be able to enable their employees or to be able to enable their tasks with the systems that they use between the AI, the large language model and their actual operational outputs. So I think things are changing slightly in terms of what is going to give you the biggest competitive advantage.

6:26Just as an aside to that from my own knowledge, this isn't part of part of that particular brain piece of brain food but i am noting the irony of china launching so many really great open source models the us saying we want to ban these and the chinese at the same time saying we want to ban selling them to the us right so i mean who's who's actually like the cold war around ai is hilarious yeah but here's the thing there's two levels to the cold war so i would say both governments are actually quite incompetent um like they're not necessarily knowing what they're doing um i mean trump doesn't know anything about ai he's just listening to people he's like whatever xi jinping he's 60 he's similar age bracket you know i mean he doesn't know anything about ai either um and so you've got those two sort of people who are kind of like making big decisions and then you've got a different type of conflict in sort of at the bottom level between the different companies in both of those ecosystems.

7:28Some of them are open, some of them are closed. US, unfortunately, have gone massively to the closed source. This is where it was very disappointing that Mark Zuckerberg ultimately turned Meta's sort of path from open to closed, because at one point I had hoped that he would be the open source champion for AI in the US, but he thought okay let's not do that let's go closed um and right now open source is definitely dominated by ai from chinese labs however uh the one of the things about open source is that you can just download the model learn from it and just roll out your own um so it should it should have a distribution sort of effect yeah i'm actually sorry we do also need to differentiate between a lot of the models that are like we are referring to as open source are not actually open source they're what's called open weight correct which means you can see certain aspects of it but you don't actually ever get to see the training like data and you can't download them um they're just really allowing you to understand like how they're working and why they're working rather than the exact ingredients like they may be telling you what the macros are not precise ingredients for example so that is an another interesting distinction in that there's this kind of pathway house type of model option yeah but the bottom line is for an entrepreneur or a business owner right now i think the future is going to be one where you're going to have a multi-model system you'll do a lot of routing around between different models to get the best price for what you get you don't need the frontier for 99 of the things that human beings do currently with ai you do not need the frontier model um a frontier model probably is only required really for people in r d um and in software um if you need someone to process your inbox that's not something you need like um a very expensive uh uh ai to handle uh anything bog standard will probably do the job so bottom line is you've got a route between them in other words switch models as you go forward uh and your point with regards to you know maintaining the integrity of your business cultivating that cultural intelligence within your company how you specifically run your business all of those things I think will be the big differential going forward but I do think that's going to lead to things like on-premise AI as well so basically AI in on hardware that you manage a new air gap so it doesn't go anywhere you don't go and train sort of anthropic in terms of how to do it but anyway when we introduce the main topic for this session around employee lifetime value I'm going to come back to this because there's a specific point I want make which links what we just talked about there to the concept of eltv right brilliant um i think it is linked that way i look forward to your comments um i've just shared the link by the way to mozilla who by the way obviously open source champions um their report on open source ai and where its adoption is in the enterprise brilliant um sort of bit of research very accessible as well beautifully designed go ahead and read that okay give us a couple more all right So let's talk about Glenn Cathy's respectful employee with candidate rejection email playbook.

10:44Like the very worst case scenario is like 80 % of what happens. Apply for a job, hear nothing back ever. That's the absolute worst case scenario. Probably slightly better than that. In fact, I don't know if it's better than that or not. is 15 seconds after you send the apply, you get an email saying, no, you're rejected. And that happens a lot. Yeah. That's the workday, Mobley versus workday scenario. Overly fast response. Yeah. Yes, absolutely. Absolutely. There is not a chance that humans reward that, right? Despite talent acquisitions cry to say, AI doesn't reject candidates. It definitely, definitely does.

11:25This is utter nonsense. So the next sort of best thing is you get an email from an individual to say, hi, sorry, you don't have enough experience. That's really great. The next best case is you read Glenn's playbook and you have got maybe stored on like your folders or something. Here's five different intros and five different mid sections, five different down bits. And you go and copy and paste those bits into your email and send it to each individual. That is good. That's really good. But the problem is that Hung gets 300 applications. He is not going to do that. He doesn't have time to do that or he'll get fired because he's not actually filling any jobs.

12:05So I honestly believe there is a next step on this. And I'm job smacked every single day that applicant tracking systems haven't just made this one of the most fundamental things they offer, which is recruiter gets my rejecting hung not enough experience click hung gets a message saying hey sorry you're not getting this um opportunity because you don't have an experience here's four things you can do that our company would really value so that when you apply again we you will get to the next stage right you've got to give the people the feedback because what he said was it's not you're not getting rejected from the company you're getting rejected from that specific opportunity at that particular time and we can 100 % put do personalization at scale and rejections and it's just mind-blowing that we're not doing it yeah you know what Adam I don't even think you need to personalize it I think you just need to not give people a stone wall um I remember a conversation with our friend Stan Vasovich shout out to Stan by the way I haven't chatted to him for a long time hope he's all right uh but I'll get into sharing a bit Stan um but um he mentioned that you know what best thing you can do is yes you can deliver the straight facts up front say listen we're not going to go forward but then give the individual some other path that some other thing to do rather than just pee off uh you know what i mean um here's a community you can join here's some resources you can work with um here's some sort of additional pathways you might want to explore um give them something else um so that they can feel that actually there's another forward path and it's not just a complete blanket stopper um now i reckon you can do that without personalization it'd still work quite as well um because i don't think it's personalization that's the main thing even though that's probably a marginal position i just i think it's just the provision of just some forward progress some options to go forward um all right anyway glenn kathy um obviously follow him um and obviously have a look at his scenarios on rejection emails i think uh you could just educate yourself on how to just uh uh communicate a little bit better by reading some of that stuff and you know i'm sure you could automate a lot of that stuff with ai these days as well using glenn's templates all right give us one more mate before we get into this okay um andreas and horowitz's talent engineer fellowships um you know we've seen this type of job emerging in talent acquisition teams?

14:31Is it going to take over the job of an executive search professional? No. Is it going to take over the job of early careers and potentially some of what is it going to take over early careers? Yes, 100 % or very, very high amount. Is it going to take over experience higher? Yes, for evergreen type stuff. Is it going to take over the job of like someone who's doing 20 roles at one time from a part-time receptionist through to a financial controller probably not a bit higher to bit hard to kind of engineer solutions in that but the broad concept is look in sales we've had sales engineers for a long time we've had marketing engineers for a long time we've had finance engineers a long time fine-tuning every process um within those areas talent acquisition is simply or talent or even hr is simply a smaller business function, a less valuable business function in terms of money terms and lower down the pecking order.

15:33So is it natural that we get to the point where you've got roles or even whole teams whose job is to engineer solutions? Yeah, absolutely. And good on Andreas and Horowitz for creating a central vaccine sense. You know what? I'm interested to know anybody who's applied to that. fellowship because it looks like it's open for anybody to apply fellowship I don't even know what the fellowship is but presumably you get involved in some sort of program maybe there's a course or what have you and there's a list of people that Andres and Horowitz have actually put together as being like senior fellows there and a lot of them are actually in recruiting so it's interesting to know what they're going to teach what they're going to show what the direction travel is my big takeaway from their spec though was that they were looking for people that were from a non-classical recruitment background.

16:25Coder first turned into recruiter. That's how they saw that job. And I think given the fact that obviously they're a VC that invests in technology businesses, I suspect that they would kind of prefer to hire that background to almost any job now. So if they're looking for a sales engineer, they would look for someone who's a coder who then works on that on sales. Or they're looking for someone doing marketing, the work with someone technical skills who then can apply it to marketing so it's almost like leaning very heavily someone who can basically be very comfortable um with producing automations producing scripts producing uh ai agents um so yeah very very interesting i'd love to hear as well from anybody who thinks of themselves as a talent engineer what is that job how is it different from a recruiter, let me know.

17:18I'd love to hear about it. Okay, cool. Let's get on with the show. Do we have a poll I need to sort of address people to? We do. Folks, if you're watching this on Crowdcast, I just want to point you to the right-hand side of your screen. There's a poll there, and the poll is quite simple. In your current organization, do you calculate employee lifetime value? Why don't you give that a vote while you are sure at it? and I'll bring on our guests to have a chat about it. We've got Tim for sure. Hang on. What's happened there? Tim and Jess, let's bring those two on. And by the way, if anybody else is fancying coming onto the show, just stick your hand up.

18:01I think there's a way in which anyone can do that, and we'll get you on. I see no reason why we should not go back to the spirits of Brain Food Live, which is, in fact, just to bring randoms onto the show to have a chat with us. right I've lost Tim here there he is cool Tim's getting moved on how are we doing with the poll wow no is overwhelming overwhelming response on no's here that's really interesting Jess I can see you but your audio is off is there any way you could have a quick fiddle with that and see whether you can get off mute

18:45Tim's here hello There you are. Brilliant. Hey, everybody. Tim and Jess on. There we are. Wonderful. How beautiful is all this arrangement, I have to say. Let's do some introductions here. Jess, why don't you introduce yourself? Who are you? What it is you do? Hi, everybody. So nice to be back. My name is Jessie Svan. I am the VP of People Strategy and Operations at LeapSim. prior to that was the chief operating officer of talentful uh recruitment consultancy based in the us and the uk um and the author of two books about how to apply product management principles to your people operations yeah and the new book is recently out wasn't it um did you have your recently i did i had a launch in a couple actually a whole bunch and um yeah it launched in New York with a big party and a grubby nominated jazz band only like last week so it's very exciting good stuff good stuff I'm going to be in New York as I think you know in a couple of weeks and I want to sign copy Jess all right um okay so we have we have Tim Sacker here as well Tim wonderful to see you can you quickly introduce yourself who are you what it is you do yeah hey do I do anything besides just come on podcast and stuff like this is probably the only thing Also two books like Jess, but not as recent.

20:08I'm the CEO of HRU Tech, an analyst in the space, investor, advisor, blah, blah, blah, all this stuff. Good stuff. And Tim must follow person, Jess must follow person, share their links on the Crowdcast there. Go and follow both of those folks, really good commentators on industry. Tim Sackett, I'm going to go to you first because I do believe it was actually your contribution on a greenhouse blog which prompted this sort of topic. And I remember scheduling this topic way back when. We didn't get to do it for whatever reason, and I wanted to look back on it. So brilliant to get you back on here.

20:47Can you just explain a little bit about sort of your contribution to that conversation and what it is, the message you were trying to communicate there? Well, and I actually like stole the concept. Like, I mean, so like this is like a 10 year old thing that Greenhouse did, right? So Maya Joe Bacavelli, who's the current head of revenue for Stripe, was previously like the head of like, like marketing and like strategy for Greenhouse created this. And I thought like she created out of thin air. Actually, about 10 years before that in 07, there was an academic paper that mentions like employee lifetime value, but not the ELTV kind of like thing.

21:25but for me it was like i when i first saw this like i was just totally taken aback because when i show this to like executives that go what the hell do like does hr do what does ta do what does all this do and you go hey here's a visual that kind of explains really what the roi is and so if you think of a human as a machine right we buy a machine it costs us all this money up front we put it in the factory and then little by little we start to make money on this machine and eventually we go positive on the ROI, really humans are exactly the same. And so they think, okay, so from a TA perspective, when we go in, we go out and post jobs and screen and assess and blah, blah, blah, we finally hire somebody.

22:03We're actually negative ROI in terms of the value of that employee until they come up to speed, right? So now great onboarding, great manager, leader, motivation, training, great benefits and pay, all of that stuff potentially has an impact to this employee lifetime value. And you can see then also how negative pieces that we do in our company, having a really bad manager, right? Potentially could bring down that value because the employee's demotivated or they leave faster than they maybe should be or whatever that might be. And so everything we do has an impact to this value of this employee.

22:39And so for me, it's just a great visual to tell people every single thing we do should have an impact to the value of this employee across our entire life cycle of this employee. And what we do or don't do potentially has a positive or negative impact. So like, to me, I just think it's a great way to kind of explain what we do to people that kind of like, you know, HR is a cost center. Well, maybe if you think of it that way, I think of it completely different. I think we're this engine that potentially increases the value of every single employee. Yeah, yeah, really good. So basically it is a way to communicate with the rest of the business on the primary value that HR, people ops, or how a talent, you know, whatever term we want to call it.

23:23Like, Hunk, think about like, again, go back to the machine example. I put a machine in a factory. If I just run that machine to death and it breaks down, like I'm wasting that machine. Same with a person. But if I do preventative maintenance and, you know, I change out parts or potentially like AI, like you guys had mentioned, like the AI piece. If I give a great tool to this employee that makes that machine, that person much more valuable to the entire organization or should, it should, right? And so like, to me, I just think like how we actually are treating our employees, how we're training, developing them, the tools that we surround this person should be increasing the value of that person overall.

24:03Yeah, really, really good. And this kind of helps us speak to the CFOs, doesn't it, Jess? Because they're used to things like amortizing the cost of something over several years, the CapEx spend. I buy a computer, the value of the computer is this, the value of it's going to go down over time because it degrades. We have to swap it out, but we all calculate all of this. Anybody who's ever filed any kind of tax return also knows about this, right? So from your experience, I remember you writing something about this, in you about employee lifetime value uh tell us a little bit about did i dream that or did that actually happen um no i did i actually wrote quite a lot about this for a while i am this is kind of like my white whale it's like a thing that i want so desperately to try and work out it's interesting right like eltv is when you think about marketing um cap to uh ldv to cap ratio like i've done some work with uh professor peter and dan mccarthy peter is uh wharton and dan mccarthy is uh was at emory at the time um and they are kind of like the preeminent experts on marketing ltb to cac and they've done a huge amount of work around how to calculate um those uh formulas um and what's really interesting is we have kind of like the opposite problem that marketing has so let me try to give you a little bit of an explanation here it's like attributing brand marketing is really hard right if you buy something say i like you know i buy this ipad right it's really difficult for apple to work out how much of is how much of it was bus ads that drove past me or my neighbor that said you need to get the new ipad it's really great or whether it's targeted advertising i'm getting on my laptop or my computer um so they try to like kind of aggregate and attribute all these things together and say like we think it approximately cost jesse like this uh it cost us this much to get Jesse to buy this iPad.

25:54But to some degree, it's kind of a fuzzy number if you haven't got like really, really specific performance marketing driving all of your inbound sales. And then the lifetime value is very easy because they know exactly how much I'm spending based on my name, my card, my Apple ID, all these things. We have that kind of opposite problem a little bit in HR. It's actually pretty easy to calculate how much it costs to have an employee. We kind of know how much it costs to have a recruiter on the role. We know how much time and effort we spend interviewing for a role. We know how much it costs to pay somebody, obviously, that benefits what they're using.

26:28What's really difficult is to attribute their lifetime value because it's really difficult to say, well, what did this person achieve? Now, it's kind of easier in sales and marketing, but even then you never really know, right? So a great salesperson could bring a deal in. That deal could go on to grow, to refer other customers, to have a really great effort on marketing because their logo is able to be like encourage people to join your company and buy your product. So it's kind of the opposite problem. But just like brand marketing, it is possible to kind of at least fuzzy attribute things towards a team, right?

27:03So one of the pushbacks I've been getting from people when I've been talking about ELTV is like, well, it's almost impossible to get ELTV to cacked out to the individual level. And yeah, I broadly agree. But what it's very good at being able to do is directionally push you, like Tim said, to be thinking more like your commercial team when making decisions on the kind of aggregate level. And I think that's what it's really there to do because ultimately your company is 60 plus percent of your spend probably, maybe less now with AI is going to be headcount. And your job as the leader of the recruitment team or the HR team is to optimize that massive amount of your budget as much as possible through really great org design, through great capital expenditure into like benefits and offices, what's working right and if you just pretend that these things are impossible to understand and it would never be able to get it done at the individual level therefore we throw the whole metric in the bin I think what you're actually kind of doing is alienating yourself from the commercial realities that your executive team and board is expecting of you in order to be successful yeah really good so in many respects this is underlined Tim's position that you know there is a very strong communication value to using this.

28:19We're also a great way to align the, what we're doing with the rest of the business. Often big criticism of HR is that we're offering our own thing. Recruitment's often its own thing. And it's like not paying attention to what the company is actually doing this way. So it's a, it's a, it's a stepping stone for us to get closer to the business. And indeed start speaking to the language of the company rather than to drag them to us. quick question though um do you think that sort of this is a metric that hr produces or do you reckon sneakily that cfos have actually got a secret not a secret but an undisclosed metric themselves i mean that or does it come from them um where does who came up with this and uh you know what direction of where's the genesis of this uh particular metric is it in hr or is it elsewhere?

29:11No, Jesse actually hit it. It came out of the customer lifetime value, right? So I mean, that's like we said, okay, hey, if we can measure a customer lifetime value, why can't we measure an employee lifetime value? I think for every company though, it's going to be independent. I don't know if there's a benchmark metric that you could actually build. This is like quality of hire, right? We were trying to build a benchmark quality of hire metric forever. The reality is, is it's really difficult. Like when someone says, oh, well, our quality of hire is great. You ask them how they measure. They're like, really what they're measuring is 90 day turnover or retention.

29:41You're like, that's not really quality of hire, but like, they're just trying to like build something right to kind of give them that. So that's, to me, I'm like, if you can build something that everybody within your leadership team agrees to, cool, measure against yourself, benchmark against yourself. This is, it's going to be valuable to you from that perspective. But I think, you know, I, again, like one of the reasons I go back to this, like foundationally, I think I've run into so many C like heads of HR, heads of TA. They have a struggle to define what the job of HR TA really is. And for me, it's just, are you increasing talent?

30:16That's it. Simple period. Are you increasing talent? We can do that by great TA. We can do that by great benefits, great pay, great employee relations, great training. Are you increasing talent in the organization? Period. You know? Yeah. Very, very good. So how do we do that? Go ahead, Adam. When Jesse was last on a few weeks ago, we talked a little bit about the concept of considering employees as balance sheet items. And I don't think we have ever been in a better position to be able to do that. We didn't entirely agree on whether we should actually position that towards everybody or not. But the fact is, we absolutely should be thinking about that behind the scenes.

30:59yeah and i think we've never it's never been easier to really measure somebody's input and i really believe that in the age of ai you've got let's take for example data labeling companies right like surge and scale and marcor and these businesses they are i mean they know that the people who are working for these organizations they know we are getting paid to give our information once as a one-off disposable piece of knowledge transfer and then that's going to be able to use that knowledge over and over and over again right so that's great for the large language models and where I was going to link back the earlier discussion we had around like the state of large language models as a state of open source is we can today easily measure somebody's input into an organization based on how much are they equipping our own ai resources to be able to repeat their job again and again next time and repeat all their tasks again and again next time so we should be thinking about as we have got people in our company we should be thinking about them exactly the same way that surge and scale and marco are thinking about it which is we take their information we take their information we take their information next year we'll be able to do their job right that doesn't mean that we don't then put them into more jobs and there's other jobs that emerge there needs to be but that's got to be the cfo's objective around how do we measure the value of in nearly every company around other than like i don't know some examples like people serving starbucks or whatever like that doesn't quite work in that organization with certainly any knowledge business.

32:40Yeah. I would even say about the line-out sort of P &L thing. Like, I think there's, like, a kind of slightly more empowering way to think about it that I've been kind of noodling on lately, which is, like, every single person is an investment. It's a contract you're investing in. And what your job is as the VP of HR is to look at the portfolio of that investment. You're kind of building a vanguard, right? And how do you make that portfolio worth the most? And when you're thinking about ELTV, like one of the things that sometimes people speak to me about is like, well, if I really wanted to have the highest ELTV to cap, what I would do is hire a billion VP sales, right?

33:20Because they're the ones that produce the most revenue. They're the ones that produce the most growth. And I'm like, well, that doesn't make sense, right? Like there's a reason why we don't do that. There's a reason why we have one VP sales. And that's because we want the portfolio to have one person that's job is to do this within the portfolio. And maybe the best investment we can give to that person to be successful is hiring an executive assistant or three AEs. And that's where it comes down to, like, how do you balance ELTV and these investments that you need to be taking care of with org design, right?

33:47So org design is not just making people feel happy. Org design is optimizing your investment portfolio so that all of those things work well together and you've got the absolute maximum capital expenditure to output. And I think that's why, like, there's a much bigger push lately, and AI obviously is driving this, to think about things like ARR per head or growth per head, like growth revenue per head now. And these are the metrics that I think really do lead us to start, if you really start picking through them, you start all of a sudden asking org design questions and talent density questions to Tim's point as well.

34:20Jesse, here's one of the issues I think we face, especially as we think about traditional HR as a function that we struggle with. And I'll give you the example of like when I ran TA at Applebee's, I had 2 ,500 restaurants. So I knew I basically had 25 separate business units, right? every single restaurant ran their own P &L. And I had, and I could force rank them from like the highest profitable restaurant to the lowest. And I would have a general manager. So if our highest to lowest was literally one restaurant could make us 1.5 million a year, or it could lose us 500 ,000 a year. So a$2 million spread from worst to best.

34:53And I would have one of my worst GMs say, put their notice in to leave. And HR would be like, well, there's not, you know, not a lot we can do. And I'm like, no, no, no, no. You don't understand. This person's making us a million dollars a year. There's all kinds of shit I can do. And they're like, well, no, but we would have to do that for the lowest person. And so we have this concept in HR that we have, well, we have to treat everybody in the same position the same way because of legal, because of government regulation, because of whatever. I'm like, actually, no, I don't have to treat everybody the same way.

35:24Could I put myself at risk, right? If I go, oh, hey, every white male in my company that's making us a million dollars is getting paid more than every female that's losing us a million like yeah i could but i could also argue that in court right because of the financials now will i get myself in trouble if i'm paying somebody more that's a white male at a million and a woman that's being paid less at a million for sure but like but we have this concept we have to we struggle where we have to treat everybody the same and there's where we go if we are going to start to put resources on P &L of people, we're going to have a really clear indication that there's certain people in our company that are more valuable than others.

36:04Yeah. I mean, this is like, it's funny how we try to shield ourselves. I think sometimes from those realities is like, and I do actually think like, look, there's many things about kind of the discourse around like high performance culture that I find a little like cringy. But the ones that I don't think are cringy are like the reality, again like you know adam and hung we spoke about this last time that we were on here chatting together is like i'm an adult i know that my employment contract is me working for money that's like what it is it's why i'm here and for my employer it's me working so that they can make more money than they're paying me like i'm aware of that reality right and i think all of us are adults and we can all figure that out um yeah i think we've kind of like tried to obfuscate it a little bit in hr we try to like well i don't know why don't make people too aware of that it's like well why not like I'm an adult I pay a mortgage I pay a mortgage I vote I like I'm smart enough to work out that like they need to make money off me and I'm into some degree like that like there's an investment on my side too right like I could take a massive big whiskey bet and throw all my time and energy and to try and become a millionaire and start my own company but I'm also making a calculated decision to not do that like I'm I'm getting something out of this too my job is doing a job for me um and I think like I really kind of find that like kind of wishy-washy soft kind of discourse from HR teams.

37:25And I think it also, you mentioned compliance, right? It's like, I feel like HR teams that lean on compliance as a way to garner power is because that's the only way that they're able to really have like a seat where they can make a decision, right? So their job then is just, well, we don't want to do that because that's compliance actually. And compliance now is becoming commoditized. Like you can look that stuff up and prove or disprove very quickly whether it's going to be an issue. And I think if you're a HR person out there that's been using that to wield your authority, rightly or wrongly, and from a good place or not.

37:55I think now's the time to be like, maybe I should wake up to the reality that that's not going to work anymore. That's really interesting commentary, everyone. And I love this direction. It is, because sometimes HR, as functions go within business, it's fair to say the area we're in is probably quite low esteem within other parts of the business. and in large part that's because we are perceived pretty much as a compliance function and therefore we're almost like agents of the government so to say you know it's like we're there to go in and say no to stuff and it's like yeah it's sourced in the government it is it's like the government has basically outsourced itself to and we've forced us to hire HR people to do compliance from but anyway the point is that's not the direction that we want to go in and it's not the direction that we have to go in um particularly if we start um looking at sort of uh employees as investments i love this concept of a company as a portfolio of investments jess i think that's a wonderful analogy um because though the stock price is going to go up and down on all of those people and all of those investments and then it's up to you what you want to do with them like um do you want to double down on it is are you buying cheap do you want to sell high i don't know but the point is um ta has never gone up and down there's nothing that's gone up and down as much as ta in that stock market right it's true enough we've been like a massive roller coaster but actually the stock market analogy actually might be a launching pad to the next thing we're going to talk about which is this intrusion of ai into the world of work um so before we get there i want us to quickly go to how we calculate employee lifetime value pre-AI.

39:42So Jess, I'm going to lean on you a little bit because you wrote your piece, I think two, three years ago, and that was pretty much pre-AI. So how can you just run us very quickly through how you think you can, a company can calculate a workable definition of this? And then the next question is, how does AI then impact the validity of that method of doing it? Yeah. Yeah, so I think kind of like calling back to the answer I gave before around things we already know. We already know the cost. We've got things like Xero and Gusto and Leapsim, HRAS and Bamboo and all of these different things, Greenhouse.

40:22So we can understand how much is it cost to bring someone in. The training, the offices, the payroll, the benefits, like the finance team can help you get that cost number pretty easily. What we're missing is the attribution of value. So I think the easy part of this is to understand, like, how much money is your company making overall? So you can, again, your finance team can give you, like, this is our current annual recurring revenue, or this is our MRR, doing this on that basis. But then it comes down to, like, okay, well, how do we can start splitting that through into different teams? Like, where are teams being effective or not?

40:56One of the kind of theories or ways that I think this is possible to start like getting a bit more of a fuzzy ELTV attribution is through organizational network analysis. So that is looking at certain behaviors that you believe genuinely and hopefully with some evidence drive revenue. And the behaviors that you don't believe drive revenue should not be the ones that you're measuring. So, for example, it might be something like your company's measuring impact, your company's measuring like KPIs or goals hit. If you believe there is a causal relationship between somebody hitting their goals or being rated highly for impact in their performance review, you can then theoretically attribute to them that they are contributing to your overall growth.

41:40And then organizational network analysis comes into it because one of the big areas of BLTV that I think is really fuzzy is like you might have someone like Jane in BI, right? Jane and BI is hitting her goals, but she hasn't really got a close relationship with revenue. A lot of her goals are things like support on this deal by creating this dashboard. But Jane has also trained everyone in the sales team to understand the metrics that drive close rate success. And everyone in their performance reviews or through organizational network analysis attributes to Jane part of their success and says, we wouldn't be here without Jane.

42:13We wouldn't be as successful. So then you're able to kind of say, OK, well, let's like weight that score a little bit. and all of a sudden Jane becomes a more profitable employee in your org design. Now, again, you're never going to be able to say like, well, Jane drove exactly$915 ,000 of revenue over the course of her employment, but you should be able to say, look, Jane is above and beyond this other person, this other team, a person that if we continue to invest in her, we will see outsized returns based on the fuzzy attribution logic that we've got for our performance reviews and our organizational network analysis.

42:45Is this too much, is the lemon worth the squeeze, Adam? Does this sound like too hard work? That is like a great high level analysis of how you get there. So like for like a really mature company, like you can do that. That'd be amazing. I still think like, if you just want to go down and dirty and get it started, the revenue per employee thing could work. Is it really that accurate? No, but is it directionally correct? Maybe. But like, again, you can do this in a lot of ways, I think is like the one thing, but Jesse just laid out like that's, if you're a mature organization and you can build that, like that's exactly the way to do it.

43:18And actually, just go back on Jess's point. Nice, high five to the screen right now. I do remember one of the things on Jess's piece was that it did kind of segment as to which type of company you were, like the maturation of the company, the size of the company, all of that is relevant. And it's another topic to talk about, okay, when is it particularly relevant and what type of company is particularly sort of interested in this number. um moving on i posted this dumb thing on linkedin yesterday really about no doubt like it's yeah no doubt like everything i post i'm like that's dumb but um like i know the algorithm really well um it's so it was about like i'm like i literally i just got gotten a text nudge from my dentist about cleaning right and it's like i get like seven nudges about making sure i show up for my appointment but then so my question was why don't we do this for interviews and again i said it knowing the fucking answer.

44:13But like everybody came back to me like, oh, Tim, like interviews aren't, they don't drive revenue. And that's where you go. There's why it's disconnected is you have literally people in TA that don't understand that hiring better, faster actually equals revenue to their company. And yes, we should be nudging interviews. We should be eliminating ghosting. And like, we still have this huge disconnect that they don't understand employees equal revenue. Yeah. I won't have to read that post, Tim. I think how we know but how we connect it is super super important um because it obviously does have this impact i think jess you mentioned it right it's very fuzzy the attribution is very difficult to do i disagree with jacob by the way saying you give up on it don't do it um because it is something that it's worth trying to at least map forward even if you just settle on the absolute basics uh which is revenue per employee um which we know is inadequate but step one maybe that um is this not the purpose of OKRs in the first place?

45:13So the top level is we want to make a billion of revenue. So what does that mean for finance and for sales and for products and for whatever? And how do they all contribute the right building blocks? And so therefore your key results have to all, is that not, I don't want to dumb it down too much, but is that not the entire purpose of that? Yeah, you're just reversing hearing the math, right? Yeah, exactly. You should be able to go down to these people's contributing backup terming. I think that one of the reasons why I kind of stray a little away from being like too heavy on like kind of drilling down goals into the individual level is because I pretty much every single person that I know that's tried to implement OKRs has had a nightmare with it.

45:53Like I've done, I've had OKRs work successfully. But if I'm really honest with you, it's like they're just smart goals at a team level that we do before the quarter starts. And then we reflect on them when the quarter's begun. and i think whenever we start we start talking to hr leaders about like you need to have like drilled down rolling up team specific objectives and key results i think people just start getting a little bit cold because they're like no i don't want to have to do it because it's like not a it doesn't always feel like the juice is worth the squeeze on that if i'm honest with you yeah yeah and we let's move on to the next point because it's relevant to this to the point we're just making ai ai number one tells me hey it could actually be really useful to help us calculate employee lifetime value?

46:36Because suddenly we've got this amazing sort of technology that's going to help us crunch all this information that might have been just totally not worth the squeeze. If you had to do all this manually, it's like, oh my God, I can't do it. Okay, now AI can help. So number one, it can do that. But number two, AI is also potentially transforming the jobs of people that we're measuring. So doesn't that significantly complicate the possibility of calculating employee lifetime value when we don't know the value of the skills tasks or even the job itself over a period of time, given that AI may make this transformation.

47:13So the two things separately there. Jess, firstly, do you see AI as an analytical support to the calculation of employer lifetime value? What's your view? Yeah, absolutely. I think actually the way that AI can be helpful is that I think it does take the edge off that OKR point that Adam just raised. Like in the past, the reason why OKRs I felt were really, really painful is because they were manual. It required somebody to sit there and type in every single person's goal and count. Did they all make sense? Do they all sequence well together? And I think you can kind of do that pretty quickly now.

47:45You can say, here's our company goal. Let's set top line goals for each of these teams. Distribute them, check them. That kind of work can actually happen pretty quickly now. And the other thing is organizational network analysis. Like I have recently built in the back end of Leapsim, I actually shared an example of it today on LinkedIn. a full organizational network analysis on all previous peer feedback that's ever existed in our company by just running like a dashboard tool now. So I said, I'd want you to build me an organizational network analysis of all previous peer feedback so I can see where the strength of connections are between people.

48:19And also then I would like to start seeing as a potential measure of quality for hire, how quickly does someone start receiving positive feedback from their peers and And who are the outliers that it starts happening quicker than others? These kinds of questions, you can just run your old data to run tests of like, does this actually indicate something useful? If this is now pulling out that a really great quality of hire was Jane Austen in sales, I can then look at Jane Austen in sales, like actual performance record historically and say like, is this hypothetical analysis actually telling me what I want it to tell me?

48:50Or is it telling me something that's actually useful? So I can dive bigger into it. And I can just do that in like a two day kind of sprint on figuring out like what might be a useful metric to look at. Whereas before you had to implement a special tool that had a special algorithm, run it for a year, then look at the data, then assess whether it was helpful or not. I can just do all that stuff on historical data now, which is like really fun. Yeah. So organizational network analysis, folks, people don't know. I had like a big moment about 10 years or so ago when everyone's saying, you know what, let's just track all the metadata.

49:22Let's track everyone's messaging from all of that stuff. if we can then start making inferences on what's going on in the business, it was brilliant. But it was also one of those examples of the lemon not worth the squeeze at the time because it was just too much effort. But now AI may bring OANA back because suddenly we can figure out what's going on. I mean, I love your example, Jess, when you're talking about, you know, the speed of positive feedback. Is that something worth paying attention to? I think maybe so. maybe that person is making immediate impact with their colleagues. Very, very interesting.

49:58I think there's so many dots to connect that I think it is difficult for a human to comprehend ELTV. Just like it's hard for us to comprehend quality of hire. But I think with AI, can you imagine having a private model within your company that you say, hey, we're going to give you access to everything. All of the emails, messaging, all of the voice calls that happen, every system interaction, all the teams, the Slack messaging, blah, blah, blah. And now you come to us and tell us, and here's the concept, right? That we want to try to figure out the value of every single employee, who's the most valuable to the least valuable, blah, blah, blah.

50:31And in between, we're going to add in network analysis. So we know those informal kind of connection leaders. So someone who might have a number of value that's lower potentially is greater because of the network effect that they have. And all of a sudden then see what AI comes out with, right? Because there's so many signals of noise and as humans we can only comprehend so many at a time that makes sense to us whereas ai potentially could connect hundreds thousands millions of signals and come back to us and go oh by the way hung is your most valuable employee here's why yeah really good i remember sorry go ahead it was organizational network analysis was the right it was the right solution at the wrong time so that was we were trying to incorporate that in an age of like trust and like people were you know trying to make themselves an employer of choice and a caring employer and a you know one that reflects respects the individual and none of that exists anymore there's never been less trust between employee and organization so you can put it in now that would probably work um wonders but just on the ai bit i wanted to add that if you've got a company with like basically no ai no data capture or like knowledge sharing or anything like that versus another company with great ai compounding organizational knowledge and workflows and um jason um averbrook's put it very well in the chat on here like lifetime value could be dramatically different according to are we capturing what that person knows and is doing or are we not and that is all just like single use stuff that's just getting thrown out the window and has to be repeated every single time somebody performs that task, you know, that company's doing badly.

52:10That company is going to compound and, you know, really grow considerably. And for an Applebee's, it's maybe not as useful as it is for a, you know, a recruitment company or for some other type of organization. But to a lesser or greater extent, the information security or its information, whoever the chief information officer, the IT team, those people, they've got a brilliant brilliant opportunity to work with hr on how do we do this you know what i'm reminded of ants guys i'm a romantic you know there's certain species of ants that basically cultivates um fungus um and they're basically one of the first farming non-human farmers out there so they actually cultivate this fungus to eat the fungus to take care of the fungus um and i'm just thinking we may just be ants guys like we may be cultivating the intelligence of the business we just got to keep the thing alive the thing is going to keep us alive but we've got to keep the thing alive and this is going back to jason's point yeah when you're hiring this person coming in in future this person is not just a a like a task uh uh executor uh the person that's going to feed the intelligence of your organization um and therefore their value need their employee life the lifetime value of that person is significantly advanced um more than just this tax task executor in either direction, by the way.

53:32So if this person is a negative influence, they could pollute your intelligence or they could really degrade it in some way. And it could really be deleterious to your system, your culture, your institution, and so on. Like we've seen this in politics, correct? Look at the clowns that we're electing globally. We're electing clowns into politics. Now, I think these people have a corruptive and corrosive influence in the institutions that they inhabit. They're going to leave, but they're going to leave the institution much weaker than they were before they found it. And I think the same effect happens with human beings.

54:10So we're going to have to scrutinize our hires a lot more significantly as we go in. But it leads me to another point, which I've got to raise before we finish. Does AI's impact on the job itself and how it changes unpredictably, whether it's valuable or not, the composition of the job, the skills that you're doing, is that actually going to defeat our ability to come up with the metric? Because again, I'm thinking about recruitment as a classic example. We were five years ago, only five years ago, premium skill recruiter, technical sourcing. Can this person have the wizardry to be able to interact with all these chaotic systems and just pull out identities of candidates that were very hard to hire and get a hold of them?

54:55These days, that skill is useless. right AI can do that it can it can discover this information um and therefore I guess an an employee that was highly skilled in that previously would be high employee lifetime value might have declined now because AI has replaced it so anyway Jess is upset with this so let's hear no no it's it's funny I'm not upset with it I think what I think I was talking to Matt McFarlane And if you all know him, he's kind of an expert on compensation and pay. And we were having a conversation about like, okay, there's another attribution problem that's emerging now. We previously had this human attribution problem, and now we have the attribution problem of like, who is really responsible for driving the performance of a person?

55:39Is it them or the tool? If someone is spending a bunch of money on tokens and able to do 10x what they were before, is that attributed to them? Is that attributed to this spend that we've introduced? Do we include token usage as part of the acquisition and like retention cost of an individual? Do we exclude that entirely and just focus on growth less token spend? It's kind of introduced a new complexity to how we calculate the success of someone in your team. And actually I've got one other little thing that I think is like an article worth sharing, which is an article from someone in Salesforce.

56:16I'm going to find the exact name of the person. it's this idea of stranded okay stranded ROI Shabani at Salesforce as we use data talk to this idea of like people are building these really fabulous things that really like could be potentially massively efficient facility improving but it's not just their ability to build it it's about the organization's ability to capture that ROI through things connecting really well so it doesn't matter how effective someone is at creating efficiency within their their own role anymore. It's actually more about, I actually kind of like this, like my cell and like mushroom roots analogy here.

56:53It's like, it's their ability to connect deeper into the organization and capture that ROI through like, again, like the organizational network, right, of their work. So I think that that's the thing that I would say, actually, AI starts really disrupting is like, it changes the way we think about an individual's performance. It's now actually more than ever about the ability to build systemic thinking and like ROI capturing systems. That's really interesting, Jesse. Tim, what are your thoughts on this? Like, how do we even do it now? I guess is my fundamental question on this. I mean, there's a part of this, if we stick within this ELTV kind of concept is that the AI is going to be the tool, just like if we're giving them a sourcing tool or an ATS or something like that.

57:42Like there's additional tool that should make them better. Now we also know there's certain employees that will leverage it and be super users of the tool and gain more value. And there's going to be other employees we have that will not use AI because whatever reason they've decided AI is not for them, and they're going to be less valuable. And so there's always that. And I think at enterprise, we'll figure out how to scale and we'll figure out how to measure. and so we'll kind of force the usage. It's this kind of mid-enterprise SMB where you're going to have super employees come up that are actually having the AI backpack on that are going to be like, oh my gosh, we can't lose Tim because he's so valuable because he's figured out how to use these tools and make us a 10x better organization.

58:26I don't think you'll see that at enterprise as much as you'll see that in smaller orgs. And then those people will be like, again, however much they want to be paid, you're going to have to pay them because they're giving you that much value to your colonization. So again, their increase of value is going to be exponentially better than potentially other employees who are doing the same job. Do you know what? So that makes me... Go on. I was just sorry. I was going to say that makes me think because I do agree with that. However, I also believe that whatever they contribute, that's the property of the employer, right?

59:01I believe that every thought somebody has when they're getting paid to think needs to be the property of the employer so if that person is becoming so super valuable that you know they have to get paid higher and whatever i mean i can see that that's where that's going to go but does it mean that almost nobody's job is going to be the same in next year or the year after or the year after are we all going to be moving jobs pretty much moving jobs every six months every six months every six months and is this going to create a massive amount of opportunity for generalist, brilliant problem solvers, of which there's lots of those people, and zero opportunity for people who just can't work like that, which is 75 % of people.

59:45Yeah, but you also have to put in the fact that most people don't want to change jobs. It's a huge stressor, right? They might have to because their job doesn't need it. We see people all the time that are underpaid, that are super valuable to their employer, that just stay there because they're comfortable. But their job shouldn't exist, though. I mean, maybe a comfortable job isn't a thing in the future. That's not a thing. The job itself might not be comfortable. They're comfortable because they know the job. They know how to do it. They know how to utilize AI, blah, blah, blah. But they also don't want to leave their town, their friends, their system, their whatever to go someplace else that's going to pay them two or three times more.

1:00:24That's not true, Adam, because the market's not efficient in that way. I mean, there's going to be companies out there that are mega inefficient that still just trick on. I mean, just simply because there's a huge amount of legacy there. There's a huge amount. I went to a restaurant the other day. It's like, listen, mate, there's no way that restaurant should still be around. But it's, do you know what I mean? It has the legacy. It keeps trucking forward. It's going to be around for another three years. No doubt about it. Because there's too much built up. So I think there's a lot of companies that corporate-wide are that way.

1:00:54They're just locked in. They've got contracts. They've got this, that, and the other. It doesn't all go sort of the way you think. But I think you're right, though, Adam. The future direction, the companies are going to have this huge problem because you're going to start encountering these super workers emerge unexpectedly that add times 10x the value, but they're not getting 10x the comp. And your compensation can't dynamically move in that direction unless we figure out a way to somehow directly attribute their contribution to profit, let's say. um which i by the way gentlemen but like here's what we know too about technology is it's also a white man driven thing so you want to talk about compensation issues think about like all of a sudden because of the ai driven it's more technology more male driven become more valuable and we're going to have an hr system and complex setup that says wait a minute we can't do this because of the inequality which is a true but you're going to go well wait a minute we get back to the original thing Jesse talked about, which is like this balance sheet of like, there's going to be different values of individuals on your balance sheet that are doing the same job.

1:02:02And this becomes problematic across the board from a legal perspective, from a compliance perspective, from a lot of things. Like this isn't, it's not an easy problem to solve. No, I think it's a very interesting comment. It's given me a bit of an inspiration for another topic um how do you how do you handle compensation when performance might just dramatically change uh unexpectedly um and you've just you know you've you've agreed to do this just your questionnaire and where everything this person is doing is already signed sealed delivered that's great but what if they over delivered by times 10x mate um like you get you as a boss you're gonna want to give them something but you'll give them some off-the-cuff bonus or whatever it's not sufficient keep their job I love that retain their place they retain their luck count themselves lucky count themselves lucky you work for Adam Gordon that's it alright listen we're out of time folks this has been a brilliant conversation obviously we need to continue having this chat we might immediately do the comp one in fact we're doing something next week which both of you guys might need to come on onto it is exactly this topic about how you, sort of the essential guide for increasing talent density.

1:03:17So in other words, okay, what does talent density mean? Is it another term for A player? Is it another rehash of warfare talent? You know, how seriously do we take it? But I think the idea of AI now coming in, really disturbing, really shaking up all of these concepts, which only a couple of years ago were very fresh concepts. We've got to look at them anew. So anyway, we're doing that next week. but thank you all for watching and participating this week, it's been a brilliant chat Jesse Zwan, great to see you Tim Sackick, great to see you as well Adam Gordon of course great to have you back on the show I hope you enjoyed it everyone if you have, follow the channel on Crowdcast, if you're watching this on Crowdcast do it every week, if you're watching this on LinkedIn, follow my page on LinkedIn because I live stream from there as well every week alright, that's it everyone, we'll see you next week thanks for watching Hey guys.