This story takes place in the near future at the HQs of most companies. It's about you or someone you know.
Natalie Ferris had been an HR Business Partner for fourteen months. Before that, talent strategy at a mid-cap tech company. Before that, management consulting. She moved fast because she thought fast, and right now something was bothering her. This was more true after attending a briefing on Paul Sackett's remarkable study that pretty much puts every existing hiring system and HR tech stack at risk.
One of her business leaders, Tom Hadley, had just lost his senior analyst. Tom did what everyone does – sent Natalie the req to backfill. Same title. Same JD. Same level.
But Natalie had been watching Tom's team. Half of that analyst's work – reporting, data pulls, formatting exec decks – was already being done faster by AI tools two junior people had started using. The other half – strategic synthesis, cross-functional storytelling – was really being done by Tom himself because the analyst had never been strong at it.
So why were they hiring the same role again?
She brought it to her CHRO, Diane Colbert with an idea for rethinking work. The idea turned out to be profound.
Diane listened for twenty minutes. Then she called the CEO's office. "Rich, I need thirty minutes today. Natalie on my team has developed something we need to move on."
That afternoon, Natalie sat across from Rich Kessler with Diane beside her.
"We don't have a hiring problem," Natalie said. "We have a work design problem.
Every tool we use – job descriptions, reqs, interviews – is built to fill one seat at a time. But nobody's designing work at the level where it actually happens: the team, the project, the department. That's where performance lives. That's where AI changes things. And that's where we should be starting."
Rich leaned forward. "What does that look like in practice?"
"We tie it to the operating plan. Instead of writing a job description for a person, we write a performance-based job description for the department. First!
We first ask: What measurable outcomes connect this department to the business plan we've already committed to? We define the high-level work first – no headcount, no titles, no org chart."
She pulled up an example from Tom's analytics department that had this major key performance objective from the department-level description: Deliver weekly revenue intelligence briefings that enable the commercial team to adjust pricing and pipeline strategy within 48 hours of market shifts.
"That's the department's job," Natalie said. "You'd never get there writing individual JDs in isolation. That briefing isn't any one person's job – it's the team's job. How the work gets distributed across people and AI is a design decision that only makes sense at the department level."
"So what happens to the individual role?" Rich asked.
"It gets shaped by the team design. Today, the senior analyst spent 70% of her time pulling data and building dashboards. AI handles that now. The real human work is interpreting the data, knowing which signals matter to which leaders, and presenting a point of view that drives action. So the individual objective becomes: Synthesize AI-generated market and revenue data into strategic recommendations, and present actionable briefings to commercial leadership weekly. Completely different hire."
It's comparable to telling our controller to "Get out of the numbers and make a difference!"
"Yes, and you only get there by designing the team first," Rich said.
"Exactly. Here's the process:
- Step one, define the department's outcomes from the business plan.
- Step two, assess your people by real capabilities and interests, not legacy titles.
- Step three, filter every task – can AI do it, assist it, or is it irreducibly human?
- Step four, identify what's missing and build a plan. Maybe a hire. Maybe training. Maybe a contractor or a better tool."
Rich sat back. "I've been asking for two years how we get more from our headcount. Every answer is 'hire more' or 'cut costs.' Nobody's said design the work at the team level to match the plan we already have. This is such a great idea, Natalie!"
Then Diane threw in a little surprise. “Natalie’s already developed a Work Analyzer tool using AI that actually does much of this work. It creates the team and individual performance-based job descriptions by analyzing everything it knows about the people and their abilities in the context of the department objectives. With your okay she’ll begin testing it out with Tom’s department.”
"You got it!," Rich said without hesitation.
That's the disruptive idea here and how to think beyond the job description by changing the unit of design. Stop designing individual jobs in isolation. Start designing teams, projects, and departments as integrated systems – then let the individual roles emerge from that design.
One way to get started is asking AI how to rebuild your company's entire hiring process.
Most important, never ask AI how to hire the same people faster.
