Workforce management capability models are usually drawn by vendors and always end at the rung the vendor sells. This is a four-rung ladder with the cost of each rung stated, and an opinion about where most teams should stop. The rungs are cumulative and skipping one does not work, which is the most useful thing about the model.
Rung one: measured inputs
Shrinkage, handle time and volume measured rather than assumed. Cost: an afternoon and access to your own data. Almost every team that thinks it is at rung three is actually here, because nobody agreed the definition of shrinkage.
Rung two: a computed requirement with visible assumptions
The arithmetic done in the open, with the inputs stored beside the answer. Cost: a worksheet and a record, which is roughly free. This rung ends most staffing arguments, which is why it is worth reaching before spending anything.
Rung three: a scheduling engine against real constraints
Availability, skills, rest rules and multi-site, generated rather than assembled. Cost: a suite and an implementation. Worth it when producing the roster is genuinely the bottleneck rather than agreeing it.
Rung four: closed-loop optimisation
Adherence, intraday re-forecasting and automated re-planning. Cost: the suite plus a dedicated planning function to run it. Genuinely valuable at scale and expensive to half-do, because the capability decays without someone whose job it is.
Questions people ask about workforce management capabilities
Where should most teams stop?
Rung two, and deliberately. It is nearly free and it removes the most expensive recurring problem, which is disagreement about the number.
Can we skip a rung?
Not usefully. A scheduling engine fed by unmeasured shrinkage produces precise rosters for the wrong headcount.
How do we know which rung we are on?
Ask three people what shrinkage is. Agreement means rung one is done; anything else means it is not.