The planning cycle that's missing a layer
Your workforce plan for next year is mostly done. Headcount by department, attrition assumptions, a hiring budget, a handful of backfill-versus-growth decisions. It has all the pieces workforce planning has always had — and none of the pieces that answer the question your CFO or board is now asking alongside it: which of these roles carry meaningfully more AI exposure than others, and does that change what we do with the hiring budget?
Most workforce plans don't have an answer, because most workforce planning processes were built before that question existed. Gartner's 2024 HR Priorities research found that 86% of HR leaders have not implemented strategic workforce planning at all, and among those who have, 66% say the practice is still limited to basic headcount planning and struggle to demonstrate its return. Adding an AI-exposure layer on top of a planning process that's already thin in most organizations is not a small ask — but it's also not optional for long, given how fast the underlying skill base is moving. Lightcast's 2025 analysis found that 32% of the average job's skill requirements changed between 2021 and 2024, and that a quarter of jobs saw 75% skill turnover in that window.
This article covers how to fold AI exposure data into strategic workforce planning without turning the practice into something it isn't — a forecasting tool that tells you who gets let go. Exposure data is an input to planning judgment, not a substitute for it.
What "AI exposure data" actually measures
Before it can inform a plan, exposure data needs a definition precise enough to survive a budget review. At WorkforceAnalysis, exposure scoring works from O*NET's occupational task data and scores each task a role performs against four dimensions, each rated 0–100:
- Cognitive routine — how standardized and rule-based the reasoning involved is
- Physical routine — how repeatable the physical or procedural steps are
- Social/judgment — how much the task depends on relationship management, negotiation, or contextual judgment calls
- Creative — how much original synthesis, design, or novel problem-solving the task requires
Each task in a role's O*NET profile is weighted by its documented importance to that occupation, and the four dimension scores roll up into a role-level exposure profile. A role doesn't get a single "exposure number" so much as a shape — a role heavy on structured data reconciliation looks different from one heavy on client negotiation, even if their headline exposure scores are similar.
Worked example. Say a role's O*NET task list breaks into three weighted tasks: a 40%-importance task scoring 80 on cognitive routine, a 35%-importance task scoring 55, and a 25%-importance task scoring 20. The importance-weighted cognitive routine score is (0.40 × 80) + (0.35 × 55) + (0.25 × 20) = 32 + 19.25 + 5 = 56.25. Run the same weighting across the other three dimensions and you get a full profile, not a verdict. The output sorts roles into Monitor, Review, or Redeployment Candidate — categories that describe where attention is warranted, not what should happen to the person in the seat.
That distinction is the whole reason exposure data belongs in planning at all. Planning is where judgment happens; the data just tells you where to point it.
Where exposure data changes hiring and staffing decisions
The most immediate planning use of exposure data is in decisions the workforce plan already makes every cycle: which backfills to approve, which new requisitions to open, and which roles to consider consolidating or redesigning before recruiting against them.
A role landing in the Monitor category on your exposure heatmap doesn't mean the requisition gets frozen. It means the hiring manager and HR partner have a specific conversation before the req is posted: is this the same role it was eighteen months ago, or has the task mix already shifted enough that the job description should change first? That conversation is cheap. Discovering the answer after a six-month search and onboarding cycle is not — SHRM's 2024 research puts the fully loaded cost of replacing an employee at 50% to 200% of their annual salary, and SHRM's benchmarking data separately puts the average cost per hire at $4,129. Exposure data doesn't reduce those numbers directly, but it gives planning teams a reason to ask the redesign question before the spend, not after.
For a granular look at how exposure shows up differently across role families — which functions see the heaviest concentration of Review and Redeployment Candidate flags, and which stay squarely in Monitor — see our breakdown of the staffing and hiring impact of AI by role. It's the natural next read once you've decided exposure belongs in the hiring conversation.
Redeployment as a planning output, not an afterthought
The part of strategic workforce planning that exposure data changes most is the one most plans handle worst: what happens to the person in a role that's shifting, not disappearing. The WEF's Future of Jobs Report 2025 frames this at the macro level — 170 million jobs created and 92 million displaced globally by 2030, a churn rate of about 22%, netting to 78 million new roles overall. That net-positive number matters, but it hides the harder operational fact underneath it: the same report finds that of every 100 workers, 59 need meaningful training before 2030, and of those, only 29 are expected to be upskilled into their current role and 19 reskilled or redeployed internally — leaving 11 unlikely to receive the training they need at all. Separately, 63% of employers cite skill gaps as the single biggest barrier to workforce transformation.
Redeployment only works as a planning output if the plan has somewhere concrete to redeploy people to. This is where task-level exposure data earns its keep over department-level guesses: because the scoring is built on ONET occupational data, a Redeployment Candidate flag comes with a matched set of adjacent ONET occupations sharing overlapping task and skill profiles — not just a flag that says "at risk," but a starting shortlist for where a person's existing skills already partially transfer. That's the difference between a plan that identifies a problem and a plan that produces a next step.
If your organization hasn't yet built the case for treating exposure data as a planning input rather than a compliance exercise, our company-specific AI workforce exposure guide walks through how to run that first assessment end to end, from role mapping through department-level rollup.
Building the exposure layer into your planning cadence
Most strategic workforce planning runs on an annual or semi-annual cadence tied to budget cycles. Exposure data works best layered into that same cadence rather than run as a one-off:
- Assess. Map your current role list to O*NET occupations and run the four-dimension scoring at the start of the planning cycle, before headcount and hiring-budget conversations begin.
- Segment. Roll individual role scores up to department-level heatmaps. This is where patterns become visible — a department with a cluster of Review flags concentrated in one sub-function is a different planning problem than the same flags scattered across every team.
- Plan against the segments, not the individuals. Use Monitor-tier concentrations to flag roles worth a job-description refresh before backfilling. Use Review-tier concentrations to prioritize training-budget allocation, informed by the reality that Josh Bersin's research finds external hiring costs three to five times more than internal placement — a strong argument for weighting the plan toward internal mobility wherever a Redeployment Candidate flag has a viable internal match.
- Monitor and re-run. O*NET's underlying occupational data updates regularly — the Department of Labor's ~277-descriptor dataset receives its primary annual update in the third quarter, with quarterly refreshes throughout the year — so an exposure layer that was accurate at the last planning cycle can drift out of date. Re-running the assessment each cycle, rather than treating it as a one-time audit, keeps the plan current.
For teams evaluating what kind of tooling should sit underneath this cadence — a spreadsheet-based process, a full enterprise people-analytics suite, or something purpose-built for exposure scoring — our workforce analytics software guide lays out the tradeoffs by organization size and planning maturity.
Where the data stops and judgment starts
None of this changes the central constraint on exposure data: it describes task composition, not futures. An exposure score is not a prediction that a role will be automated, and a Redeployment Candidate flag is not a notice. What the data does is narrow the set of questions a planning team has to ask from scratch every cycle down to a defensible, task-level starting point — which roles need a closer look, which departments should get training-budget priority, and which internal moves are worth exploring before an external search opens.
McKinsey's 2023 analysis estimated that 60% to 70% of work hours across the economy could theoretically be automated with current technology, up from roughly 50% a few years earlier — theoretical technical potential, not a company-specific timeline. The gap between that macro estimate and what's actually happening inside your organization is exactly the gap strategic workforce planning with AI exposure data is built to close: replacing an industry-wide statistic with your own role list, your own task weights, and your own department heatmap.
Getting the exposure layer into your next planning cycle
If your planning cycle is coming up and you don't yet have a role-level exposure picture to bring into it, that's the gap worth closing first — before the hiring-budget conversation, not after. Our guide to future-proofing your workforce covers the broader practice this fits into, and the HR AI Strategy Toolkit packages the frameworks in this article into a working document you can bring straight into a planning meeting.
WorkforceAnalysis runs the assessment itself: map your role list, get the four-dimension scoring and department heatmap, and see where Monitor, Review, and Redeployment Candidate flags concentrate before your next planning cycle locks its budget. Compare pricing for your organization's size and start a trial before your next planning meeting, not after it.