Before You Cut Jobs for AI, Use This Matrix

Is Your Team Doing Work That Actually Matters? The Value of Work Matrix Says Check First

TL;DR

Short answer: No, probably not, and almost nobody stops to check before they start cutting headcount in the name of AI. In more than 20 years sitting in HR planning rooms, I never once heard someone ask the one question that actually matters before a restructuring decision: where does the value of this work actually sit? Everyone argues over who stays. Almost nobody maps what the work is even for.

Key stats you need to know:

• In one organization I worked with, two full-time employees spent every single day producing a weekly report sent to roughly 700 to 800 managers. When I polled those managers, 2 out of 700 said they actually read it.
• The World Economic Forum's Future of Jobs Report 2025 found 40% of employers expect to reduce headcount specifically where AI can automate tasks, and estimates only 47% of work today is performed mainly by humans (World Economic Forum).
• Gallup's latest global data shows manager engagement fell from 31% in 2022 to 22% in 2025, the sharpest drop in the entire workforce (Gallup).

The leadership takeaway: Over the next 12 to 24 months, the companies that cut headcount without first mapping which work drives the business are going to rehire a third of it back at a premium. The ones that map first will make the cut once.

📺 Watch the full breakdown on The Human Capitalist: Value of Work Matrix on YouTube

🎧Listen on Spotify: The Human Capitalist Podcast

What Is the Value of Work Matrix, and Why Do Most Companies Skip It Before Automating Anything?

The Value of Work Matrix is a simple two-by-two exercise I built to score every task, role, or department on two axes before anyone touches headcount: value and focus. I've run it inside organizations for years, and it starts before the matrix even gets drawn.

You have to agree on your beacons first. Beacons are the two or three metrics that actually define enterprise value for your organization this year. In the workshops I run, I use three as an example: increase revenue about 20%, move client retention from 75% to 85%, and push the AI-driven efficiency ratio per employee to 87%. That agreement step alone takes 30 to 45 minutes, and it's the hardest part of the whole exercise, especially with a room full of C-suite leaders who each think their own department is the one that matters most. Skip it and everything downstream is a guess dressed up as a framework.

Here's how it shows up when companies skip that step. I keep watching organizations look at a department and decide they can automate 60% of the work, so they eliminate the whole team, only to rehire 30% or 40% of those people back within a year. That's not an AI failure. That's a company that never separated the busywork from the work actually driving the beacons, so they cut blind and paid for it twice.

The macro data backs up what I see in the room. McKinsey's research on automation makes the same point at scale: "the right level of detail at which to analyze the potential impact of automation is that of individual activities rather than entire occupations," because every job bundles together tasks with completely different automation potential (McKinsey). You can't score a department. You have to score the work inside it, which is exactly the discipline I laid out when I asked whether headcount planning itself is broken.

What Are the Four Quadrants of the Value of Work Matrix?

The matrix plots every task, role, or department on two questions: does this drive one of your beacons (value, high or low), and how much human judgment does it require (focus, high or low). That produces four quadrants, and I always work through them in the same order.

• Why (low value, low focus): The lower left box is the one I call Why, as in why are we even doing this. The report example above lives here. Two FTEs, every day, low value, low focus, almost nobody reading the output. This quadrant exists in nearly every organization I've ever run this exercise in, and it's usually bigger than leadership wants to admit.
• Automate and Scale (high value, low focus): This is where you want to be moving work toward. A group of recruiters I worked with kept a "hit list" of top talent they'd built relationships with, and once a month they sent tailored check-in emails to stay warm with those candidates. High value, because that relationship drives revenue and retention the day one of those people joins. Low focus once the relationship exists, because a well-built nurture campaign can carry most of that outreach without losing the personal touch.
• Protect (high value, high focus): This is the bullseye, and I want roughly 80% of organizational work living here. Offer negotiations. Employee investigations. Real manager coaching. I don't think there's ever a point where I want a machine negotiating an offer. That's human interaction, and it's where the human side of HR creates the value a system can't replicate.
• The Danger Zone (low value, high focus): This is where the fights happen. Employees doing high-focus work often equate that effort with job security, and it isn't always true. My position here isn't elimination. It's redeployment: take the experience someone built doing this work and point it at a role where that same judgment drives one of your beacons instead of protecting work nobody asked for.

In a boardroom, I'd say it this way: comp, attrition, and risk all live inside these four boxes, and the Danger Zone is where you'll do the most damage if you handle it like a spreadsheet instead of a conversation. That tension is part of why manager engagement has fallen further than any other segment of the workforce in Gallup's most recent data (Gallup), the same split I traced in whether the workforce is dividing into two tracks.

How Should C-Suite Leaders and CHROs Actually Run This Exercise?

C-suite leaders and CHROs should run this as a workshop with sticky notes and peer review, not a survey that gets summarized into a slide. Here's the playbook I use.

1. Lock your beacons before you touch the matrix. Give the 30 to 45 minutes it takes, and expect resistance, especially from the C-suite. Every department believes it's more valuable than the one next to it. You need agreement on what actually drives enterprise value before a single task gets scored.
2. Score the work, not the job title. Ask each division or department to list every activity they're responsible for as pre-work, and bundle similar tasks together. You're not asking "should this role exist." You're asking what specific behaviors and tasks that role performs, one at a time.
3. Put peers in the room, not just the team being scored. I guarantee you that on the first pass, almost nothing gets categorized in the Why quadrant and almost everything lands in Protect. People rate their own work generously. Peer review is what gets you an honest map instead of a flattering one.
4. Treat the Danger Zone as a redeploy signal, not a cut signal. If someone's high-focus work isn't tied to a beacon, the answer is moving their experience into a role where it does drive value, not eliminating them. This is also where the toughest conversations in any restructuring happen, so plan for it.
5. Use the finished map as your AI baseline, not an afterthought. Companies that skip this step look at the Protect quadrant and decide they can automate it anyway because the technology exists. The Harvard Business Review put it plainly: AI transformation requires redesigning work, not cutting roles wholesale, and plenty of layoff announcements are ordinary restructuring dressed up in AI language (Harvard Business Review). The matrix tells you which work needs the judgment a system can't supply before you make that call, the same case I made in asking whether HR itself was built for a workforce that no longer exists.

What This Means for You as a Leader

AI is rewarding the organizations that can tell the difference between busywork and value, and it's exposing the ones that never bothered to find out. The Danger Zone and the Why quadrant have always existed inside every company I've worked with. AI just made the cost of ignoring them a lot more visible, a lot faster.

Winning here doesn't require predicting exactly which tasks a model will handle next year. It requires mapping the work you already have, protecting the 80% that should stay human, and redeploying the people whose experience got stranded in the wrong quadrant instead of cutting them. Leaders who do that mapping keep authoring their own org chart. Leaders who skip it end up rehiring a third of the department they just eliminated, at a worse price than the first time.

I break the full Value of Work Matrix down with more examples in an entire chapter of my book, "The FutHRist," dropping September 14, 2026, on Amazon. If you want to see where you land across the five mindsets that make up the book before you buy it, the eight-minute self-assessment is at trentcotton.com/futhrist-start-here. Watch the full episode this post is drawn from on The Human Capitalist, or listen on Spotify.

Word count: ~1,890 including the FAQ (roughly 1,510 in the main body, excluding meta and schema blocks)

FAQ

What is the Value of Work Matrix in simple terms?

It's a two-by-two exercise that scores any task, role, or department on two questions: does it drive one of your core business metrics (value), and how much human judgment does it require (focus). The answer sorts every piece of work into one of four boxes before you decide what to automate, protect, or cut.

What are "beacons," and why do I need them before I do anything else?

Beacons are the two or three metrics that define enterprise value for your organization this year, things like a revenue target, a retention goal, or an efficiency measure. You have to get agreement on your beacons first, because every other decision in the matrix gets measured against them. Skipping this step, which usually takes 30 to 45 minutes of real debate, means every task you score afterward is a guess.

Is most of my team's work supposed to land in the Protect quadrant?

Yes, roughly 80% of organizational work should live in the high value, high focus quadrant I call Protect. That's the bullseye: work that drives your beacons and genuinely needs human judgment, like offer negotiations, employee investigations, or real manager coaching.

What happens to work in the Danger Zone?

Low value, high focus work is where employees often equate effort with job security, and that's not always accurate. The right move isn't elimination. It's redeployment, taking the experience someone built in that role and pointing it toward work that actually drives one of your beacons.

Is the Value of Work Matrix about cutting headcount?

No. It's about knowing what to automate, what to protect, and who to redeploy before you make any headcount decision. Companies that skip this step and automate by department instead of by task tend to rehire 30% to 40% of the people they cut, at a higher cost than if they'd mapped the work the first time.

Where can I learn more before The FutHRist comes out?

The book drops September 14, 2026, on Amazon, and an entire chapter walks through this matrix with more examples. Before then, you can take the eight-minute self-assessment at trentcotton.com/futhrist-start-here or watch the five mindsets breakdown that the book is built around.