Bonus Episode: Ask A Headcount Expert - July Edition
Podcast Overview
-
Eric Guidice Headcount experts, special episode. I think we're at 24. I don't know if these count. You know, when I asked Claude to summarize these, I don't know if the bonus episodes are counting when I'm making my table of contents for our future book. But nonetheless, episode 24, special reaction episode. Before we get into it, Chris, what are you working on? What's on your plate this month, this week? What's going on with Meander? What's going on with LearnDot? Tell me about what's going on with you. I'll give you a quick summary, then we can get right into the content.
Chris Mannion Yeah, big summer really, a huge project going on right now where I'm launching this new AI and HR accelerator. So it's an eight-module series, taken self-study, comes with a bunch of SHRM PDC credits so that you can requalify. And the idea is to onboard, you get access to all the platform or the former courses that we did. And then this is a new progressive course that takes you about 20 hours to get through and will take you from zero to hero in AI and really give you a lot of credibility in doing it. I'm actually pairing that with a new thing that I'm testing where we're actually creating simulated cases. So there's one thing to sit through a course and answer quiz-based questions, but a lot of this you actually have to put in practice and it can be quite hard to practice on real HR data. So we've developed these cases, the same things that we used to do at business school. If you've seen anything from Harvard Business School, they do these cases all the time. We actually work through a case, takes you 30 to 45 minutes, and you get live feedback based on an answer key, which you'll never see. And Claude, which actually looks at the answer key and looks at your answers, will give you feedback on where to go and then direct you to the courses, videos, and free content that you need to upskill. So a lot of work going on in really making the Academy best in class, but really excited about that. The only thing that's getting in my way is spending a little bit of time on the beach. We've got a little bit of good weather here in Massachusetts. What's new with you, Eric?
Eric Guidice Yeah, you have to make the most of an East Coast summer. That's the difference between, you know, I'm from New York, New Jersey, and I'm now in Los Angeles. Every day is an eight in Los Angeles, but there's never that moment where everyone's like, "Hey, we all have to come out because it's been 38 degrees for the last six weeks". Anyway, jealous. Jealous of the beach time.
On my side, I'm doing implementation stuff. I've been really diving into how to get our customers onto the platform. I am building an approvals AI. Our approvals product, if someone makes a headcount request, is so advanced that it slows down the implementation when people learn about what's possible. It's normally like my manager, then finance, then TA, then good, which was our first approvals product, funny enough. But as it grew in popularity, we developed it so if the salary is within this range and it's this department, route it to this person only if it came from this user.
So we want people to be able to use natural language to say, "This is what I want," and then have it create the prompts for our implementation team, or identify if you can't do that, or if there's circular logic, or if it's conflicting. And I think that'll really reduce the implementation time. But that's my vibe. I'm not allowed to vibe code because we have so many different microservices that have relationships that I can't add code to the code base without all this additional knowledge and training. Stick to marketing and sales, fine, but this little app I can do because it's kind of independent and doesn't really affect our code base. So I'm excited for that, but really just trying to nail that process down for the Q3, Q4 busy season.
And one of the main articles I'm writing, which maybe we'll talk about on a future episode, is about headcount IDs. One of the Headcount 365 selling points, or at least I'm going to market with it, is that we create this universal ID. One ID for what a headcount is across all three systems. And again, through the implementation work that I'm doing, people call it budget ID, finance ID, seat ID, position ID, requisition ID, job ID, or opening ID. Each individual system has its own ID system as well as a use case for someone to game that system for the benefit of their job, like recruiters adding the same job twice in the ATS under different titles to see which one gets a better inbound response rate. Maybe that's pre-candidate fraud, who knows? And so we run into all these examples. So we have to tell that central business systems person how to organize all this stuff, how to select an ID, what the difference is between a display ID and a UUID that a system reads, and which ones are reused versus which ones are unique.
So I'm going to release that article soon. Just waiting on my designer who did your four-box image, we just posted it, to make the visual representation of all the IDs in one. So that's what I'm working on.
Chris Mannion That's very cool. I don't know if we've spoken about this, but I've spent a lot of time designing approval workflows. Because it seems to be one of the things that catches up headcount plans. And if you add too much friction, the company doesn't really grow, everyone gets frustrated, people actually start to backchannel approval requests, and then you lose control over it. Make it too easy, and then you actually lose your headcount run rate and wind up having a RIF six months down the road.
And so it's like, how do you get to that sweet spot? We could actually do a whole episode on that. Maybe we should at some point. I'm actually going to do a module on that in the accelerator program. But as a systems engineer, that is my bread and butter. I really enjoy when something works and you connect all the boxes and you can actually see things flowing through and it makes life better for everyone. So I'm really excited to see the launch of that. I hope I get a preview before everyone else.
Eric Guidice Yeah, no, I'll show you approvals. Approvals is the most interesting because it's the first step of what we call Headcount 365 orchestration, where multiple people are involved, but the plan is what everyone has approved. But purgatory is like the pending approvals. And there's a couple of purgatories. There's backfills or attrition that is or should be a backfill, but isn't or might be. How many vacant seats do I have? There's that problem that we solve. There's also, "I backfilled this senior person with two junior people, what is the ID?" There's someone that we hired this year who is already getting backfilled, so the ID or the person already occupies space in an existing system, and then you're doing it again. There's accepting an offer that was a backfill, but they never actually started. So there are all these little purgatories.
Figuring out how to capture all of them and organize them was the first challenge. And as we develop the product, it's more about who needs to know the information and where do we show it. Like, is it part of the budget yet, or is it not part of the budget? It's the most interesting problem. I've actually been talking with another one of our competitors, Carol, the founder there, and we just bounce ideas back and forth. And the thing that we always land on is this is the most frustrating but fun part of solving headcount: the sheer complexity of not only what the process is, but how each individual customer has strung together systems or workaround tools and put in base processes to fix it. And then how do you create a tool that can ingest all of that, give them the specificity that their very custom tool wants or not, and then navigate the whole thing?
It's my favorite part of the job, but also the most frustrating. So there's probably some psychologist who's going to tell me that the hardest is the most rewarding, but that is what we feel and it permeates the headcount industry. So if you're an expert, shout out to you, Kathy. Anyway, that's life as a headcount founder.
A lot of good posts this week. I want to dive into these. We only have a few minutes to bust through some of the most interesting content. And for those of you who haven't seen this segment before, I try to find the inflammatory, interesting, statement-making content about headcount planning, workforce planning, and workforce analytics, and try to get Chris to react to it so we can discuss it a bit. So I have six today from some interesting profiles around the space.
Starting first with Antonio Labruzzo, based in Switzerland, who is the Senior Vice President of People at Arcara Life Sciences. He has a post showing the ranking of HR priorities, even though the topic list stays stable. Let me pull it up.
From Antonio Labruzzo: "HR priorities don't change as much as we think, but the weight of each topic does. What's going up, what's going down, and strategic workforce planning is top of the list". I don't know if these are in order, Antonio, but if they are, you're speaking to the right crowd. This is data from Boston Consulting Group and the World Federation of People Management Associations. Have you heard of the WFPMA? Let's start there. Have you heard of that organization? That's probably deserving of a follow.
Chris Mannion I haven't come across them yet, actually.
Eric Guidice Okay, well, you earned two new follows, WFPMA. Anyway, strategic workforce planning is on the up and up. Chris, what's your reaction?
Chris Mannion I think it makes a lot of sense. This is what we're seeing in the market. Everyone's looking at their organization and thinking, "Do we have the right people in the right seats?" I think what's interesting is what's moving down: employee engagement, rewards, and recognition. I think that's implying that retention is no longer a concern; it's more about finding the right people and onboarding them. Maybe this is a result of the RIF culture of the last few years, where companies now probably still have more people than they need, or the skill set's not right, or they don't know what skill set they want. And so the focus is now on how you hire more people with the skills that you need.
I think it makes sense. I don't really have any strong opinions against that. I would argue that if you overlook employee engagement for too long, then you're going to need to do a lot more recruiting and onboarding, and strategic workforce planning becomes less predictable because attrition is going to go up. And if attrition goes up, it's probably going to be the best people who are going to get poached by others who are also prioritizing recruiting and onboarding. So does this then become a fight for the best talent, and how does that actually manifest in the market?
Eric Guidice My new global philosophy on recruiting applied to every position I hire for, or that I think people should hire for, is that we are hiring for critical thinking. If an AI can do it, great. I'm buying the brain of the person who interprets that output and customizes it based on a message we're trying to send or a product we're trying to build. And in my space, at least, a lot of what we're doing hasn't been figured out yet, so it requires this critical thinking. Looking for how that fits in the org chart and how to associate what I have today with what I'm trying to hire has been one of the biggest things I look for.
In our app, we're seeing utilization of the org chart go up. We've developed some new features where you can organize it by different work statuses and incorporate the hiring plan. But we see a lot of people going in there and trying to visualize what their workforce looks like. I anticipate that that will go up significantly between now and the end of the year as everyone figures out 2027. So we've done some work to try to improve the visual scenario planning process for that purpose, but we're still in the learning process from our customers about what specifically they are looking for, why it's important to them, and how they want to visualize it with the data they have today. What is the proper outcome, and how do we optimize for that? So I'm not surprised about it.
Here's my other theory. As AI makes productivity expectations go up, we are forcing the quality of work down. People are accepting AI outputs as good enough if they need to be productive, and they will just improve at the level of AI's improvements. That not only drives token usage up, which will then impact the tokens versus people conversation, but it's changing the difference between who is good and who got the work done with a great outcome they can stand by. To me, that is the balance, but that's what I think is happening in the workforce planning conversation. Is that inflammatory, or what do you think about that?
Chris Mannion That matches, actually. I'm in the middle of a four-part framework that I'm presenting on my YouTube channel, and I'm recording part two today. It's about how to think about AI-based org redesign. The way a lot of people are thinking about it is, "Okay, I've got a hundred people in my org, AI can do 20% of the work, so the future org needs to be 80 people. Which of the 20 people am I going to let go and replace with AI?" We all know that's totally the wrong way to do it.
Those companies then find they have to rehire those roles, they don't get the efficiency savings they were expecting, and AI transitions everything to the mean in terms of quality so you don't actually get quality output. The key part is the critical thinking piece: how do you make sure that the right people are in the right roles to interpret the AI output so that you get the productivity gains without dropping quality down?
The framework is really interesting. I picked this up at the Leading Edge Consortium last year with SIOP. I'm presenting it back this year with operational headcount planning as an overlay of how to do this continuously. The framework is essentially that you look at the tasks everyone currently does, figure out how much of those tasks can be automated and to what extent, and then figure out what additional tasks you now have. Once you've automated those, you need to assess the output of the automation, rebuild the roles, and figure out who is best fitted for those positions. That's a pretty big chunk of work. Going from "20% of people are going to be let go because of AI" to "it's probably a three-month project to get the right people in the right roles" is a big deal. I think that's why Antonio has strategic workforce planning as number one, because if you don't get that right, everything else falls apart.
Eric Guidice That's the Recruiting Analytics YouTube channel for people who aren't following along. Where can they find you at the SIOP presentation? When is that?
Chris Mannion The SIOP Leading Edge Consortium is a subsection of the annual SIOP conference focused solely on what's new in people analytics. September 30th and October 1st are the two days that I'll be there, hosting a session on operational headcount planning and AI role redesign as it relates to what's happening in the market right now. That's off the back of a number of different presentations I've given this summer to VC/PE portfolios around how this new framework is going to impact the way they think about headcount planning. It's such a new theory and framework, everyone's trying to figure out how to apply it.
Eric Guidice If you're new here and haven't followed Chris, he runs Meander HQ, a learning academy where you can take free or membership courses and apply that knowledge to your business. He is also a public speaker and runs the Recruiting Analytics YouTube channel. Give him a follow.
Next up, we have Sarika Lamont, Chief People and Enablement Officer at Technomile. Sarika says HR is leading the AI transformation at her company. She advises some of the AI startups in our network. What she's saying is that she doesn't view it as a technology problem as much as a human problem, so she is taking on the role of AI transformation as the HR leader. Chris, what's your take?
Chris Mannion She's totally right. The way I think about it is like sales enablement; AI is talent enablement. With sales enablement, you buy sales software like Salesforce or HubSpot, your IT team implements it, but your sales enablement team gets the sales team to use it and measures success. Applying AI to a whole organization has to go through a team that understands how people work and how to manage change so that it gets adopted, utilized, and measured successfully. I think HR is the team to drive this.
The big gap I'm seeing is that people are trying to understand how AI actually works so they know what to expect from the implementation, and what that change management process looks like for a successful launch. We don't want data points like the recent MIT study which said 95% of AI projects don't return a positive ROI. If you just switch on a tool and tell everyone to use it, no one gets the benefits or measures it successfully. HR driving this is definitely the right thing to do and what I'm seeing in the market.
Eric Guidice On August 13th, we'll be hosting a people leader doing AI transformation, culminating this series on what AI is doing to the org chart, and having a live discussion about how a people leader rolls out AI.
Someone needs to architect. Someone has to have the critical thinking of what this is doing to our organization and how we invest in it based on a skills map and our goals. In every successful tech company I've been in, there's been a lab, an experimentation think tank, and there should be that. But rolling it out broadly without testing reminds me of JCPenney rounding up to the whole dollar instead of 99 cents and killing sales across all stores instead of testing in one. AI needs to be handled strategically. Shout out to Sarika Lamont for the interesting post.
Next up, we have Paul Walsh, an advisor in the security space. He wrote an article about Meta being sued by 26 former employees who claim AI was used to dismiss people with disabilities or those on medical leave. It names Eightfold AI and claims it improperly discriminated against certain employees so management decisions were made without considering medical leave. Chris, what's your reaction?
Chris Mannion For years I've been trying to figure out what Eightfold AI does. I've attended conferences where they presented and didn't tell us what they do, so it's fascinating to read that they're creating hidden profiles of every employee. There is an issue with AI where you set the criteria to train it, and it makes selections based on that. AI is not inherently biased; it's biased based on the information you give it and how the algorithm runs. If you're running algorithms based on token usage, you discriminate against people who aren't trained in token usage, who are absent, or who are on medical leave. If you don't have a human check the output, you make terrible mistakes.
This is one of the dangers of relying on AI without keeping a human in the loop. We're going to see a lot more cases like this. I've seen new products where founders build a digital twin of employees. There's a question around who owns that digital twin. If a company trains an AI on my work and I leave, does that AI twin stay with the company with my IP? It's an interesting legal and ethical framework. I side with the people bringing that lawsuit; we need transparency so that if we use AI, we do it ethically and without discriminating.
Eric Guidice The digital twin thing is super interesting. There's probably a whole episode in what AI will do to the employment contract if things like this become prominent. Follow Paul Walsh on socials for great content and a different perspective on workforce management. As these AI cases unfold, like the Workday discrimination case, it will be interesting to see the outputs.
Next, we have Anna Maria Senkovici, Chief Talent Officer at the Royal Caribbean Group. She wrote an article talking about the price to move the needle and token costs. She advises workforce planners that the cost of replacing headcount with AI is becoming cheaper in two ways: the capability of the model and the ability to use smaller models for menial tasks. Software development is the canary, but this will come for a lot of routine knowledge work. Chris, what is your reaction?
Chris Mannion Every time I come across a problem where I spend a lot of time doing something, I think about whether I should hire someone or automate. When comparing the costs of screening, hiring, and employment contracts versus building an automated workflow, at my scale, it always makes sense to use AI. It expands my personal capacity in a way where hiring would cost 10 to 100 times more.
Token cost is critical, but it's cost and quality together. If you can get 80% of the way there and the remaining 20% is your own capacity to improve the output, that's really feasible. The models being released now are so effective that the need to increase team size is going away. We're seeing fewer people enabled by AI, and reduced token costs will drive that further.
Eric Guidice If you're looking at things that are repetitive, predictable, and require an interpretation that can be trained well in a model, you're going to see that happen. As a founder, AI makes me better to a degree. But when you try to reduce headcount and increase the workload on the remaining people, you will be fed AI results more often and will have to figure out how to keep quality up. You'll have to determine the cost of that last 10% for a person versus a token.
Chris Mannion AI is an enabler, not a replacement. You shouldn't replace a good designer or recruiter with AI, but you can increase their capacity if they use AI for low-value work. If you replace a recruiter with AI completely, you get poor quality hires.
Eric Guidice I couldn't agree more.
To close, we have Jackie Simon, an executive coach and former SVP of talent. She posted that if you're an executive, you better learn how to do headcount. Strategic workforce planning is moving up in priority, and there are basic questions every company must answer to understand its workforce. The difference between an executive and someone who just has headcount is their ability to answer these questions and show that headcount choices achieve business outcomes. To be a great executive, you have to know headcount. Chris, what's your reaction?
Chris Mannion I've seen this play out many ways. When executives try to preserve their own team size or empire build, they ultimately end up leaving the organization because they aren't doing what's best for the business. Great leaders think broadly about reallocating headcount dynamically to deliver against new initiatives. At the director level, the question is how you use your team productively to deliver success for the business and help other teams. Leaders who protect their team by refusing to contribute outside their core job ultimately hurt the business and themselves.
Eric Guidice The difference between a leader and a player is individual stats versus team outcomes. When I was at Uber, we had distributed city teams, and when it made sense to centralize to leverage economies of scale, you saw which leaders held onto their team size versus those who embraced regional ops for the business. The closer to the top you get, this becomes the difference in who a founder or leader wants to have around.
My name is Eric, Headcount 365 founder, joined by Chris Mannion. We are the headcount experts. Follow Chris on Meander, LinkedIn, and the Recruiting Analytics YouTube channel. Check out the Headcount 365 blog for our four-part series on AI and the org chart. Stay tuned for more upcoming guest episodes. Goodbye!
July’s most interesting headcount posts
In this reaction episode of Headcount Experts, Eric Guidice and Chris Mannion dig into six posts making noise across the HR and workforce-planning space, from Antonio Calco' Labruzzo, Sarika Lamont, Paul Walsh, Ana Maria Sencovici, and Jackie Simon. The through-line this week: strategic workforce planning is climbing to the top of the priority list for a reason. As AI gets cheaper and more capable, the hard question stops being "how many people can we cut" and becomes "who are the right people in the right seats to interpret AI's output without letting quality slide." Eric and Chris react to each post from the operator's chair, where the token math, the org chart, and the human-in-the-loop all have to reconcile.
Featured Headcount Content Creators | Episode 24
Antonio Calco' Labruzzo - SVP of People, Arcara Life Sciences | Are HR's priorities actually changing?
Fresh BCG and WFPMA data shows the topic list stays stable but the weighting shifts strategic workforce planning, recruiting, and succession up, engagement and rewards down. Eric and Chris land on why planning tops the list: get the right-people-right-seats question wrong and everything downstream (attrition, quality, rehiring) falls apart.
Sarika Lamont - Chief People & AI Enablement Officer, TechnoMile | Who should own AI transformation?
Sarika frames AI adoption as a human problem, not a technology one, and owns it from the HR seat. Chris's analogy: AI is talent enablement the way Salesforce needs sales enablement IT can implement the tool, but only a team that understands change management gets people to actually use it, which is why 95% of AI projects (per the MIT study) fail to return ROI.
Paul Walsh - Online Safety & Privacy Expert | Should AI be scoring who gets hired and fired?
Paul's post on the Meta lawsuit 26 former employees alleging AI and token-usage metrics were used to select people on medical or disability leave for dismissal, with Eightfold AI namedsparks the episode's meatiest tangent. Chris digs into the "digital twin" problem: if a company trains an AI on your work, who owns that twin when you leave? A legal and ethical framework the industry hasn't built yet.
Ana Maria Sencovici - Chief Talent Officer, Royal Caribbean Group | What actually moves your headcount math?
The number that moves headcount isn't model capability it's price. As token costs fall and smaller models handle menial tasks, the automation timeline workforce planners have been budgeting against just got shorter, with software development as the canary. Eric and Chris agree the catch is cost and quality: AI gets you 80% there, but that last 10% is where the value — and the expense — lives.
Jackie Simon - Executive Coach | What separates a department-builder from a business-builder?
A post-and-article two-parter arguing the tell for C-suite readiness is how a leader talks about headcount — can they answer the basic questions about their workforce and tie it to business outcomes, or are they empire-building to protect team size? Chris's take from experience: the leaders who hoard headcount ultimately stall out, because they never grow from manager into executive.