Strategic workforce planning: the gap between the plan and the market

We took a year of workforce plans from centres we support and laid them against what the market could actually supply, month by month. For most of the org, the annual plan was a good instrument. Deep talent pools, familiar functions, seats that reprice slowly: time-to-fill landed close to assumption, comp held, the ramp behaved. The plan earned its keep across the bulk of the headcount.

It came apart in a predictable minority: the scarce, high-demand roles where a handful of firms compete for the same short list. There, the plan's assumptions aged inside 2 quarters. The availability the plan counted on had thinned. The comp the plan budgeted had already moved. They planned once, annually, for roles whose market resets several times a year.

So the plan is two documents stapled together. The stable majority, where an annual forecast is close to right and cheap to produce. And the volatile minority, where a number set in the planning cycle is stale before the first requisition opens. The trouble starts when a centre treats both halves as if they age at the same rate.

The centres that closed the gap forecast at two speeds. The stable majority stays on the annual cadence, because that's where the annual plan is accurate and where re-reading it more often would just cost effort for no gain. The volatile minority moves to a quarterly re-read, with live availability and comp checked against the market before each commitment rather than once at plan time.

This is a division of labour. Identify the roles whose market changes inside the year and put a shorter clock on those. Leave the rest alone. A centre with 40 open positions might have 6 that need a quarterly read and 34 that don't. Spending equal planning attention on all 40 is how the 6 get under-served: the volatile roles need a fresh look precisely when the annual cycle has gone quiet.

What the shorter clock buys you is early warning. A quarterly market read on a scarce band surfaces the repricing while the requisition is still a draft, not after 3 candidates have declined the budgeted number. The plan becomes a hypothesis you refresh where refreshing is worth the cost.

Most planning functions are built to execute. Once the plan is signed, the work becomes delivery: open the requisitions, run the pipeline, hit the ramp. Execution rewards discipline and consistency, and for the stable majority that's exactly the right posture. You don't want a planner second-guessing a role the market supplies reliably.

Re-reading is a different discipline. Its job is to notice when the ground under an assumption has shifted: when a comp band the plan set in January no longer clears offers in May, when a pool the plan sized has been drawn down by 3 competing centres nearby. Re-reading rewards doubt. It asks whether the plan still describes the market, and it's most valuable on exactly the roles where execution feels hardest.

The failure we see most often is a planning function that only knows how to execute. It treats the signed plan as fixed truth and pushes harder against a market that has already moved, reading the resulting misses as a recruiting problem rather than a planning-input problem. Nobody owns the question of whether the plan is still right, so the plan degrades silently while everyone works overtime to hit its now-wrong numbers.

The two-speed gap plays out slowly on a scaling charter, and it has a familiar shape. Drawing on centres we've watched scale, some that broke through, some that stalled, the timing of the wall is remarkably consistent. Year one goes well. The first cohort is hired from the most accessible talent, the pitch is fresh, the roles are the ones the market supplies easily. Momentum is real, and everyone reads it as proof the plan works. Then year two arrives, the same plan is executed the same way, and it starts to grind.

What changed is that the plan assumed the easy supply would keep flowing at year-one rates. It doesn't. The most accessible slice of the pool gets hired first, by you and by everyone else building nearby. The second cohort is drawn from a thinner, more contested, more expensive market than the plan modelled, and the ramp that looked linear on paper bends. The wall is a supply-depletion assumption the plan never pressure-tested.

The centres that scaled past it out-planned the wall. They treated year one's easy supply as a depleting resource and started widening the funnel before it ran dry: opening a second location, building a returner or adjacent-skill pipeline, growing capability internally rather than only buying it in. The diversification went into the plan while year one still looked effortless, which is the only window in which it's cheap to do. This bites hardest in pharma and other regulated domains, where the specialist talent the charter needs is concentrated and slow to replenish, so the second cohort hits scarcity fast.

The sort is worth doing deliberately, because guessing puts stable roles on an expensive quarterly cadence and leaves volatile roles on an annual one that fails them. Three questions do most of the work. How many firms within commuting or remote range are hiring the same profile right now? How fast has the comp for this role moved over the last 18 months? And how deep is the qualifying pool relative to the number of seats the market is trying to fill?

Where competition is light, comp is steady, and the pool is deep, the role belongs on the annual plan, and refreshing it more often earns nothing. Where several centres are chasing the same short list, comp has climbed sharply, and the qualifying pool is a fraction of open demand, the role belongs on the quarterly clock. The sort isn't permanent: a role can migrate as its market changes, and a band that was stable for years can start repricing when a new hub opens nearby. Re-run the sort each planning cycle so the two lists stay honest.

Most centres find the volatile list is short, a handful of bands rather than the whole org, which is what makes the quarterly discipline affordable. The cost of two-speed planning is small precisely because the roles that need the shorter clock are few. The cost of ignoring it is the year-two wall.

This keeps the annual planning cycle and adds a second, lighter loop that runs against the volatile list between annual cycles. Once a quarter, someone checks live availability and current comp for those few bands, flags any that have drifted from the plan, and brings the deltas to the planning owner before the next set of requisitions opens. That's the whole mechanism, a short, scheduled re-read on a named list.

The reason it works is that it changes when you learn the plan is wrong. Without it, you learn from declined offers and slipped ramps, after the cost is sunk. With it, you learn from a market read while the number is still editable and the second supply source can still be built cheaply. For a centre with an aggressive growth charter, that early warning is the difference between scaling through year two and stalling against a wall the plan could have seen coming.