The Skills Adjacency Playbook: How to Actually Redesign Work for the AI Era
The most common question I’ve gotten since my last piece on AI and “Franken-jobs” isn’t about the data or the challenge ahead. It’s this: “Okay Victoria — but where do we actually start?”
Fair question. Diagnosing the problem is the easy part. Telling a CHRO that their job architecture is built on brittle task lists while AI is quietly rearranging the furniture is useful. Handing them a flashlight so they can see what they’re working with — that’s the harder, more important job.
So here’s the flashlight.
First, a quick reset on why most organizations fumble this
The instinct, when AI enters a workflow, is to update the job description. Add “AI literacy” somewhere between “strong communicator” and “proficient in Excel.” Congratulate yourself on being future-ready.
That’s not job redesign. That’s job decoration.
Real redesign starts one level deeper — at the task. Not the role. Not the title. The actual work being done, day to day, and what’s changing about it. When you skip that step, you end up with roles that are either bloated (too many disparate responsibilities crammed together) or brittle (defined so narrowly they snap the moment AI shifts the workflow). Neither serves your people. Neither serves your business.
The Four-Step Skills Adjacency Framework
This isn’t a theoretical model. It’s the approach I’ve seen work — across all industries navigating serious AI adoption. The specifics vary. The logic doesn’t.
Step 1: Task Audit — What’s Actually Changing?
Before you redesign anything, you need to know what’s moving. Pull a role apart at the task level and ask: which of these tasks are being automated, augmented, or eliminated by AI? Which are expanding? Which are new?
In one financial services organization I worked with, an analyst role that had historically been defined by report generation and data compilation was quietly becoming something else entirely. The AI was handling 70% of the data work. What remained — and what was growing — was interpretation, client communication, and escalation judgment. The job description hadn’t moved an inch. The actual job had moved substantially.
You can’t map adjacencies if you don’t know where the work is actually going.
Step 2: Adjacency Mapping — What Skills Naturally Sit Next to Those Tasks?
Once you know which tasks are growing, look at the skills that logically neighbour them. Not skills that sound impressive. Skills that genuinely sit close to the work being done.
An operations manager whose role is shifting toward AI-workflow oversight doesn’t need to become a data scientist. They need stronger analytical reasoning, risk awareness, and structured problem-solving — all of which are already latent in good operations talent. That’s adjacency. It’s not a leap; it’s a step.
The discipline here is resisting the temptation to over-reach. Skills that are too far from a person’s existing capability set create performance risk and burnout. Skills that are genuinely proximate create growth.
Step 3: Human Capability Layering — Where Does the Human Value-Add Live?
This is the step most job redesign processes skip entirely — and it’s the most important one.
As AI takes on more of the “how,” humans become accountable for the “why,” the “should we,” and the “what happens when this goes wrong.” That shift needs to be explicitly designed into the role — not assumed.
In a services organization I worked with, AI was deployed to handle the majority of tier-one contact centre interactions — routing, FAQs, basic troubleshooting, account inquiries. On paper, it looked like a headcount story. In reality, it was a capability story.
The agents who remained weren’t doing less important work. They were doing the hardest work — the calls where a customer is frustrated, confused, or in a genuinely difficult situation that no chatbot is equipped to navigate. Complaints that carry legal or reputational risk. Emotionally charged interactions where the wrong response compounds the problem. Complex, multi-part issues that require both system knowledge and human judgment in the same breath.
What the organization quickly discovered was that their existing role design — built around call volume metrics and script adherence — was completely misaligned with what the work had actually become. They weren’t running a contact centre anymore. They were running an escalation and resolution function that required empathy, critical thinking, and situational judgment as core competencies, not soft-skill afterthoughts.
When we redesigned the roles around those human capabilities — and restructured hiring, training, and performance measurement to match — both customer satisfaction scores and agent retention improved. Not because the technology changed. Because the role design finally caught up to the reality of the work.
Step 4: Role Resilience Testing — Will This Hold Up?
Before you finalize a redesigned role, pressure-test it. Ask: if AI capabilities advance meaningfully in the next 18 months, does this role still make sense? Does it have enough human judgment and relational complexity built in to remain relevant? Or have we accidentally designed something that’s one model update away from being automated itself?
Resilient roles are anchored in human capability, not task lists. They flex as the work evolves.
The Missing Role in the Room
Here’s something nobody wants to say out loud: most of the people redesigning roles right now have no business doing it.
That’s not an insult. It’s a structural problem.
When AI transformation hits, organizations scramble to respond. Strategy consultants map the business case. Tech leads redesign the workflows. HR generalists update the job descriptions. The occasional well-meaning executive arrives fresh from a red-eye with a McKinsey deck and a lot of energy. Everyone means well. Almost nobody in that room is an Organizational Design practitioner — and it shows.
OD isn’t HR-adjacent. It isn’t a strategy workstream. It’s a discipline. And when it’s missing from the table, you get exactly what we’ve been seeing: Frankenstein roles stitched together from incompatible parts, accountability gaps wide enough to lose a team through, and burnout curves that spike six months after the “transformation” is declared a success.
The problem is partly structural, partly a pipeline issue. Strategy consultants understand the business but not the human system. HR generalists know the people but not the structural mechanics. Tech leads grasp the workflow but miss the organizational implications entirely. And none of them — however talented — have spent years learning how to hold all three together under pressure. That’s not a gap you close with good intentions and a tight deadline.
A skilled OD practitioner — with real authority, not just an advisory mandate — brings something the rest of the room genuinely doesn’t have. Job architecture methodology that goes beyond titling conventions. Spans and layers expertise that accounts for how humans actually operate, not just how they look on a slide. Capability framework design that survives contact with reality. And critically, the ability to hold the tension between what leadership wants on paper and what’s actually executable at the human level.
That last part is chronically underrated. Anyone can draw an org chart. Very few people can tell you whether it will actually work — and why it won’t if you push it too hard.
If your organization is serious about redesigning roles for the AI era, the first question isn’t “what does the new org look like?” It’s “who in this room actually knows how to answer that question?”
If the answer is nobody — that’s where you start.
The Leadership Imperative
None of this happens by accident — or by committee. Skills-based role redesign requires deliberate, senior-level ownership. CHROs need to be driving this with the same urgency they’d apply to a compensation restructuring or a talent acquisition crisis. CEOs need to be asking for it. Boards need to be holding both accountable. The organizations that are getting this right aren’t waiting for a workforce crisis to force the conversation. They’re treating job architecture as a strategic capability — something that gives them a competitive advantage in attracting, retaining, and actually deploying talent effectively.
The ones that are getting it wrong are still simply updating job descriptions.
The bottom line
Skills adjacency isn’t a buzzword. It’s a methodology. And like any methodology, it only works if you actually use it — not as a one-time exercise, but as an ongoing discipline built into how you think about roles, talent, and organizational design.
AI isn’t waiting for your job descriptions to catch up. The post-mortem on organizations that got this wrong will say the same thing: they saw it coming, made cosmetic changes, and called it transformation. Don’t be the one explaining to your board why that was the strategy.
