A scenario
The cost line is easy to read and it is the only part of this decision that is. What the department was quietly holding up does not appear anywhere in the accounts, and it shows up in month three as a quality number and in month six as a churn number.
Run this on your own numbersAll scenarios
Ships as a scenario: one lever, cut_department, landing in month two, with the department named from your own org chart.
Mechanics
Every department in the twin carries three weights: how much capacity it holds up, how much quality it holds up, and how much selling it does. Those three weights are the whole scenario.
A name is matched against a function map. Sales carries 0.8 on selling and almost nothing on quality. Support carries 0.5 on quality and 0.3 on capacity. Product and engineering carry 0.7 on quality. Production, delivery and logistics carry 0.75 on capacity. Quality and compliance carry 0.35 on quality and almost nothing else. Finance, human resources and IT carry 0.15 on quality and 0.25 on capacity. Anything unmatched gets 0.3, 0.3 and 0.2. Check the mapping in the graph before you trust the run.
Headcount falls by that department's headcount. Severance goes into the one off line at headcount times cost per head times the severance months assumption, which defaults to two, and comes out of cash in that month.
Operating cost is headcount times cost per head times the wage index. So the monthly saving is the department's headcount times your company average cost per head, not the cost figure on its line in your org file. The event quotes the org file figure. The arithmetic uses the average. If the department you are cutting is much cheaper or much dearer than your average, that gap is the first thing to fix in the ledger.
The department's quality weight times 0.22 is added to the quality drag and it is never removed. Unlike a supplier switch or an integration, this one has no end date, which is the model taking the view that the work does not come back on its own. Cutting customer support at Harborline, three people with a quality weight of 0.5, adds 0.11 to the drag for the rest of the horizon.
Effective salespeople are multiplied by one minus half the department's selling weight. Cutting a sales department at weight 0.8 removes 40 percent of your selling power in the month, before anybody resigns. A morale shock of 0.16 is booked, subtracted from morale that month and then keeping 55 percent of its size each month until it falls under 0.01, which is about five months.
Quality moves 22 percent of the way to its target each month, so a drag applied in month two is only about half expressed by month five. Satisfaction moves 18 percent of the way to its own target each month behind that. Churn takes the difference through the deviation term. This is why the damage from a department cut almost always lands a quarter after the saving.
Attrition is the staff attrition assumption multiplied by an exponential in morale, so a morale shock raises it for months. Leavers are backfilled at 70 percent by default, at full hiring cost, which means part of the saving is spent replacing people you did not intend to lose.
Utilisation is revenue over headcount times revenue a head can serve. A department cut lowers headcount without lowering revenue, so utilisation rises, which lowers quality again through a separate term, and above 1.12 the strain counter starts and the operations lead hires about five percent more heads if there is runway to do it.
The ecosystem
Every department in your org chart is an agent with its own objectives, which is what makes this more than a subtraction.
Objectives: get the work done without drowning at 0.4, be paid fairly at 0.3, work somewhere that still looks stable at 0.3. In the detailed log it says the obvious thing: the work does not stop when we do, it moves to people who were already full.
Same objectives, more work. They absorb the headcount arithmetic proportionally and they feel utilisation before anybody reports it.
Objective: get the outcome they bought, weight 0.35. They do not know a department was cut. They experience relative quality, which is your quality index over the strength weighted competitor quality, and they act on it at renewal.
Objectives: deliver what was sold at 0.5, hold cost per unit at 0.3, keep people from burning out at 0.2. Hires back into strain after two months over 112 percent utilisation, if runway allows, on a four month cooldown.
Objectives: protect margin at 0.45, protect cash at 0.4. Usually the agent who proposed this in the first place, and the one whose objective the scenario satisfies immediately.
Objective: hit quota at 0.55. If the department carried any selling weight, their effective number falls in the month of the cut and pipeline follows about a quarter later.
An invented example
Honestly
This scenario is a good argument against a bad decision and a poor guide to a good one. It is built to show you the second order cost, so it will rarely tell you that cutting a department is fine.
Where the numbers came from
The one somebody has already proposed. This scenario is most useful as a rebuttal or a confirmation of an argument that is already happening, because the thing it measures, the delay between the saving and the damage, is exactly the thing the argument is missing.
Because a disrupted supplier settles down and a deleted department does not. The engine treats a supplier switch and a manufacturing move as a disruption window with an end date and takes the drag off when it closes. A department cut has no end date on it. That is a modelling choice and it is arguable, so it is stated here rather than hidden.
Cut the department and add the outsourcing cost as a supplier cost change in the same month, then set the quality effect yourself by comparing the two runs. There is no single lever for it, and pretending there is one would hide the judgement you are actually making.
Only as a rate. The morale shock raises attrition across the whole company for several months and the leavers are backfilled at 70 percent. It does not know which person, and the person is usually the whole point, so treat this scenario as a floor on the cost rather than an estimate of it.
Next
The saving is in month two and it is certain. Everything else arrives between month three and month nine, and a run with a band on it is a better argument than a strong opinion.