In brief
- When AI enters the picture, many companies start by asking which jobs they can cut. That’s the wrong first question.
- The first question is what your company is here to do, and what work moves you toward it.
- Examine the work before you examine the people. Then decide where AI belongs.
Every few months, another well-known company announces it’s cutting thousands of jobs while spending more than ever on technology. And in boardrooms everywhere, leaders start asking whether they should do the same.
Investing and downsizing aren’t opposites. A company can do both in the same year and be completely clear about why. The risk isn’t the company making the announcement. It’s following its own plan. The risk is every company watching and copying the move without the understanding underneath it.
The wrong first question
The question in almost every leadership room right now is “What can AI do?”
It’s a fair question. It’s just the wrong one to start with. What AI can do is a capability question, and vendors are happy to answer it. What AI should do in your company is a business question. A vendor will answer that one too, and they’ll answer it in their own favor. Only your company can answer it in yours.
That’s what purpose is for. If the work isn’t aligned to what your company is here to do, what is it aligned to?
How to redesign the work first
Step 1: Start with what you’re here to do
Before you look at a single role, get clear on your company’s purpose and identity: what you exist to do, for whom, and how you do it. Every decision about work flows from that. It’s also the part AI can’t supply, because it’s who your company is, not a task.
Step 2: Analyze the work, not the job description
A job description tells you what you’ve said the job is. A job analysis asks what the work actually requires. They’re often different, and the difference is where the real decisions live.
For each role, start with one question: why does this job exist? Not its tasks or its title. What outcome is this work here to produce?
Then ask of the work itself:
- Does it follow rules, or require judgment?
- Is it repetitive, or does it require interpretation?
- What happens when it’s done wrong?
- Where is a human presence essential?
- What outcome does it protect or produce?
The people doing the work should help describe it. Leadership decides what happens to it. And the work gets examined before the people do. That order is the whole point.
Step 3: Decide what happens to each piece of work
Once you understand the work, you have real options. Each part of it can be automated, augmented by AI, redesigned, kept as it is, or stopped altogether. Some work should never have existed. Some work should never be handed to a machine.
Step 4: Then bring in the technology
Now the AI conversation changes. Instead of “Where can we put AI?” you’re asking “We understand this work. Where can AI improve it?” That’s a much better question, and it leads to much better decisions.
Step 5: Start small
Pick one department, one workflow or one group of roles. Learn what your analysis missed before you scale the decision across the company.
Step 6: Keep doing it
Jobs change. Technology changes. Business priorities change. So this can’t be a one-time exercise. The line between human work and machine work will keep moving, and it needs to be revisited.
Listen to the reason behind the decision
You can hear how ready a company is for AI in the reason it gives for a decision.
- “It moves us toward what we’re here to do.” That’s a purpose-led decision.
- “The CEO wants it.” That’s a leader-led decision.
- “We’ll fall behind if we don’t.” That’s a fear-led decision.
Same technology. Very different readiness. Only the first one holds up when the next new tool arrives, because it’s anchored to something that doesn’t change every quarter. It’s the same question that tells you what your culture really runs on.
AI didn’t take the decision away from you
It exposed that nobody had made one. Companies that can’t say what they’re here to do end up letting the technology, or the headlines, decide for them.
As I wrote in AI Won’t Fix Your Culture, AI gives you more of whatever you’ve already defined. Define your purpose. Analyze the work. Design the work. Then decide where AI belongs. And when you’re deciding who will lead that work into the future, the same principle applies to succession.
It all starts with knowing who your company is.
Read next: AI Won’t Fix Your Culture. It Will Give You More of What You Already Measure.
Start here
Know what you’re here to do before AI decides for you
It starts with naming your culture’s personality. The Culture Quiz gives you a first read in a few minutes.
Take the Culture QuizFrequently asked questions
Should we cut jobs because of AI?
Not before you’ve analyzed the work. Start with what your company is here to do, then look at what each piece of work requires. Some work can be automated, some augmented, some redesigned, some kept and some stopped.
What is a job analysis?
It’s a look at what a role actually requires, not just what the job description says. It asks why the job exists, what outcome it produces, and where judgment and human presence matter.
Where should AI go first in our company?
Start with one department or workflow you understand well, after you’ve analyzed the work. Learn what you missed before scaling.
Who should decide where AI is used?
The people doing the work should help describe it. Leadership decides what happens to it, based on the company’s purpose.
