Josh Bersin says 30% of HR roles are going to go away this year. You're probably not ready to make that call. So the first thing to go is the work that you can't put on a slide.
You've probably run the analysis on operations. I'm sure you did it on finance, on the commercial team too, but never on your own function. You're too busy. And in the next 15 minutes, you're going to find out which piece of your organization will probably go first and what hitting that number is going to cost you.
So we're here to help you move beyond operating as a support function. Become the business-first executive and the right hand that your CEO cannot run the company without. I'm your host, Jackson Lynch, and today is Monday, which means it is a solo episode.
And most of the argument out there right now is about the percentage that Bersin dropped. 30%. Is it 30%? Is he early? Whether the agents are as good as the demos. Like, that argument is one way to spend the one year you had to get ready. Because what determines how this goes for you is the order the work comes in, how you work the problem. And you already know which order you're going to reach for, because it's the one you were trained in. And you're going to reach for it under time pressure, which is exactly when nobody invents a new order.
So if you have spent this year watching people debate whether AI is overhyped, this is the part with your name on it. So stay with me. And you're going to walk out holding a number that you built yourself, and the reason it's the right one. That's worth more than winning the argument about whether 30% is the right number.
So here's what I want to show you today: why the largest redesign of your function in decades is probably going to produce a smaller version of the exact same function. That's not the goal. How the order that you end up working it in is going to decide whether you ever create an effective organization. And what it costs you when somebody else's number lands before yours.
But before we get into today's episode, I want to say a quick thank you. Shout out to Nathan in Arizona. I want to thank you for being a part of this community. And to everyone tuning in, whether you're joining from Manila in the Philippines or Grand Rapids, Michigan, I really do appreciate you being here. All right, let's dive into this.
Let me put you into a conversation you've probably had this year, and if not, you're about to. You're in your CEO's office, nothing on the agenda. They just wanted a minute. You know that meeting. They've read something. Maybe the Bersin research. Maybe a board member forwarded a summary from their other company. And, you know, maybe they just talked to someone down the hall at Rotary.
Regardless, they ask you a version of one question. What's our number? And can I have it before the board meets? And they're talking about your organization.
So now you're in a time crunch. You have three weeks. And here's what three weeks buys you. You can open the org chart, you can find the roles with the hardest cost to defend, and hand back a percentage. Two seats in shared services. The req you weren't going to fill anyway. The coordinator whose work is probably going to be done by the platform now.
Here's what three weeks does not buy you. What each of those roles exists to produce. What the business stops getting the day it's gone. Which handoffs run through the person whose name is now on your list. Because that's work design, and it takes months. And it needed to start last quarter, maybe even last year, when nobody was asking for a number.
So what do you do? You do what you have to. You deliver the financial target. That's what the exercise becomes, because the calendar doesn't allow anything else. You hit the number, but you didn't redesign the work. And you already know the difference. Otherwise, you wouldn't be listening. When they nodded at your percentage, nobody in that room asked what the function would produce on the other side of it, or what they wouldn't be producing. And you were relieved, because you probably didn't have that answer either.
So step back and look at what's moving. In January, the Josh Bersin Company called this the most dramatic transformation of HR in Josh's career. Boom, boom, boom. Core HR headcount could fall 30% or more, according to him. In learning and development alone, 60 to 70% of the current work appears to be automatable, is what he's saying.
Here's a translation. That transaction layer that has justified the size of your function is being priced at close to zero this year. Not in five years. Now.
And then look at what's being done with it. SHRM recently surveyed 1,700 HR professionals. I think it was last December. And the question was where AI is live inside of HR. And it's currently live in resume parsing and interview scheduling. 56% of the functions running it, but not even measuring whether it worked. So tools are pointed at the bottom of the function, and there's no measurement of the result. That's the rehearsal for how this next cut is going to go.
So back in episode 150, we talked about redesigning the work before you hire anybody. The rule runs in reverse: redesign the work before you cut anybody. And the only version worth anything is the one you started long before the number is due.
So why does a leader who has run this analysis for other functions get it wrong on their own? I think there are probably three traps. And guess what? You have agency over all three.
First trap is that you're going to build the list off the org chart, because that's the only shape that the ask has ever come in. Finance asks in heads, managers answer in heads. A head is a bundle, and counting bundles usually hides the contents. The template holds two opinions about any role: keep it or cut it.
Here's the second one. And it has your fingerprints all over it. Your business has always priced this function by its capacity. How much can you absorb? You've known for years that that was the wrong measure. And yet every budget cycle, you defended a role on coverage anyway. We can't lose that capacity. And guess what? It usually works. And every time it worked, you taught them that that was what they were buying. Capacity.
The third trap is the one that costs you, I think, the most. And it starts with something that you never — well, you probably never wrote down. Think about the person on your team who keeps three or four handoffs clean. I'm going to call her Sally. Sally knows which manager needs a nudge before the review cycle stalls, which system lies to you at quarter close, which two functions stop talking every October. None of it's written down, because it's been cheaper every year to let her just carry it on her own shoulders.
Now put yourself in the room where this gets decided. You're not in a conversation about the future of work. You're trying to get a budget approved. And in that room, you can defend anything that produces an artifact. But you can't defend what lives inside Sally's head. Poor Sally.
So you're not choosing to cut the judgment work. Nobody intentionally chooses that. It's the hardest thing that your function has to defend. And the hardest thing to defend is what I've found goes. So you cut Sally. The handoffs she was holding go quiet. And what you've done is scaled the fog across a function that just got smaller. It doesn't work as well.
So here's the force under traps two and three. You don't know where the function creates value in the terms that your business uses for value. So you default and you defend it with the numbers that you happen to have. Time to fill, completion rates, cases closed. Vanity metrics, every one a measure of how much you absorbed, not the value that you created.
So here's the reframe: effective, efficient, defensible. That's the solutions order. Every solution you build has got to pass all three. But if you solve them in any other order than that, you're never going to get to effective. Very rarely, at least.
So you want to test that? Run your own function's redesign through the order that you were trained in, and watch what happens. Defensible, it's going to come first. So you start with what you can prove and what you can protect. Compliance stays. The audit trail will stay. Coverage stays, because coverage is how this function has always answered risk.
Then efficient, which is where the agents go. And they're going to go where the documentation is, even though you probably don't have a full documentation stack. But if you do, or even if you don't, it goes to the bottom.
Then you look up at effective. And that was supposed to be third, but it's not there. It was never going to be. Because the room it needed got spent on risk elimination in step one. Risk elimination and risk optimization might produce similar outputs, right up until the business measures the distance to its competition. Risk optimization is the solve, and that's what you get in step three, not in step one.
Look, I've watched a lot of really good people run this badly, and it's the same setup every time. Short fuse, a number due by Friday, and they start with the financial target, they deliver the financial target, and then it's done, because it was usually never allowed to be anything else.
And I've been in the seat. My last inside job, we had cuts to make and they were unbelievably distasteful. I knew what it was actually going to take, which was rebuilding the work underneath roles. There was no time to do that, though. We had advisors in. And so I did the thing that was available, and I cut heads, because that's what the assignment was. And when the question came around to whether it could hold, I couldn't stand behind my own answer, because I didn't believe it would. That wasn't the solve. I didn't redesign anything.
So just to be clear, one of the challenges we had is we were cutting 80% of our heads. And we had to do it with all these exercises with 24-hour turns. We never stopped to ask whether any of this was going to work. That was not the exercise. But I knew as I went through that that it wouldn't, and started planning what my next steps were going to be outside the organization, because I didn't want to stick around to lead something I knew wasn't going to work.
Now, the only thing that separates you from that version is candidly when you start. Because you can't change it once the timeline started.
So look at the answer that you were trained to give. Are you worth it? Look how much we absorb. That's held for 40 years for one reason. Capacity was scarce and expensive, so having it was the evidence. According to Bersin, superagents can make this capacity cheap. And when capacity is cheap, having capacity proves nothing about anybody.
But I'll tell you, I don't think capacity was ever the value. Capacity was the container that the value was hiding in. Every mature role family is an inverted pyramid. Majority automatable at the base, starved judgment at the top, priced as one blended unit. Your function is the most inverted of them all, I believe. And the judgment layer at the top has been crowded out your whole career by the 60% underneath of it.
So look at what the 30% actually is. It's the base of that pyramid getting priced correctly, probably for the first time. What's left standing is the part you've never had the calendar for and have never actually been measured on. Which makes this the moment in history — the first moment in history of this function — where somebody has to write down what it produces in terms of value to know what to remove.
If you write it down, you're designing the function. You skip it, all you're doing is implementing somebody else's number. And then you spend the next three years working out why nothing gets done anymore. Unless you're lucky and you get out.
But cutting heads and redesigning work are not the same exercise. And you've spent your entire career implementing other people's redesigns. This is the part where the pen is already in your hand. It doesn't feel that way, because it arrived dressed as a budget question.
So here's the climb. Three moves, and all three run on the authority that you already have.
Play one. Take your two largest role families in your function. Write down the work. Not the job spec, the actual work. If you have outcomes, that's great. You should probably do that in terms of blocks of 90 minutes. Put every block against one of the four things that your business already pays for: revenue, margin, speed, risk. Whatever won't go against one of those four moves over to your cut list. Because you're going to be tempted to force the leftovers into a column. But the leftovers are always the answer.
Play number two is, for every role you keep, write down two things. One sentence as to why the role exists, the purpose behind it. And five to seven outcomes it delivers, each one tied to a number the business already tracks. That's where you have to add the "so that" thinking in there. That document is how a role survives a budget review on its impact. Without it, you're defending headcount. And headcount is the cheapest thing in the building to cut.
And then play three is get on your CFO's calendar ahead of budget season with three numbers. Current state cost, future state run rate, and transition cost, including the bridge. That's the step. You stop defending your function's capacity and you start pricing its output. And whoever brings a design first is going to be setting the frame the other number has to argue against.
So if there's one thing I want you to carry out of this episode, it's this. You can hit your number and you can still lose the function. And what separates those two is the work you start before anybody asks.
And let me be really straight with you before we finish, because honestly, I think this one has a clock on it. The largest redesign of your function in decades is going to produce a smaller version of the same function, because the number is going to be due in three weeks and you can only build that one off the org chart. The order decides whether you reach effective, because once the financial target is delivered, the redesign has no deadline and no sponsor, and you're screwed. And when somebody else's number lands before yours, you implement it, and you spend three years learning what it actually cost you to skip the design.
So do the design now, before anybody asks. That's the whole job. You've got to move beyond operating as a support function and become the person your CEO cannot run the company without.
One last thing. If something we talked about today resonated with you, and I hope it did, let's take 30 minutes on the phone. You tell me what's happening. I'm going to help you find the architecture underneath of it, and we can decide what to do next. You can find the calendar link in the show notes, or you can track me down at www.mytalentsherpa.com.
And that's the summit for today. Thank you for climbing with me. Until next time, keep raising the bar, keep writing down what the work produces, and keep on climbing.
