EXPOSED
The analytical core of this job — benchmarking against salary surveys, building job-evaluation grids, modeling merit-increase budgets, drafting benefits summaries and open-enrollment comms — is exactly the spreadsheet-and-text work current AI does at usable quality. What survives is the accountable part: owning a pay philosophy that executives and the board will defend, negotiating carrier and broker contracts, and signing off on FLSA classification and ERISA/ACA compliance calls where a wrong answer becomes litigation. No license protects the role; the modal worker manages a small team inside HR and will spend less time producing analysis and more time defending decisions.
Mixed — a routine tier and a judgment tier. Salary-survey regression, grade-and-band construction, merit-matrix modeling and open-enrollment collateral are all reproducible from structured inputs, which pins the bulk of the workweek low — the 9 rather than a 5 reflects carrier renewal negotiation, union or works-council pay discussions, and live executive-comp committee work that has no dataset to draw from.
Fully desk- and screen-based. The job runs on HRIS, survey portals, and spreadsheets; the 3 rather than 0 covers benefits fairs, on-site open-enrollment sessions, and walking plant or field locations to see the jobs being evaluated.
Certification preferred, not legally required. CCP, CEBS, or SHRM-SCP is common on the resume and never legally required — FLSA exempt/non-exempt determinations and ERISA 5500 filings expose the employer and the plan fiduciary, not you personally, so there is no license a model has to route around; the 5 rests on credential expectation and named-fiduciary designation in some plan documents.
Some relationship component. Employees deal with the benefits portal and the broker's call center, not with you, but the CFO, CHRO, and comp committee rely on your read of the market and your history with the carriers — a relationship that matters at renewal and merit-cycle time without being the deliverable itself.
Exists to be accountable for ambiguous calls. You decide where to sit against market, which incumbents get exception approvals, whether a role clears the duties test for exemption, and whether to self-insure — calls with pay-equity litigation, DOL audit, and ACA penalty consequences that no procedure manual resolves; 15 rather than 18 because the board and general counsel co-own the largest of them.
Has AI actually changed your work?