the-org-simulator

The Org Simulator Skill

Org changes are designed on the box-and-line level and experienced on the who-do-I-report-to-now level. The gap between those two is where reorgs fail: the map made sense, but the two best engineers quit, the middle managers went quiet, and the memo leaked on Tuesday before the Thursday announcement. This skill simulates the experienced version before the announced version exists — seat by seat for the key people, faction by faction for the rest, plus the one test almost nobody runs: reading the memo as the person it demotes, on a screenshot, out of context.

What This Skill Produces

• A winners / losers / undecided map — by named seat for key people, by group for the rest: what each actually loses or gains (scope, status, access, headcount, identity), not what the memo says they gain
• A friction forecast: the 5–8 specific breakpoints of the first 90 days, each with likelihood, blast radius, and the cheap preventative
• The leak test: the announcement read as a screenshot by its least charitable reader, plus what the corridor version of it will say by Friday
• A sequenced rollout plan: who hears it in which order, from whom, with the one question each conversation must answer

Required Inputs

Ask for (if not already provided):

• The change: current structure → proposed structure, and the honest why (the memo-why and the real-why, if different — the simulation needs both)
• The key seats affected: names/roles, current scope, and what's known of their ambitions and frustrations
• History that shapes reception: previous reorgs and how they went, standing rivalries, recent departures
• Timing and constraints: what's fixed, what's still movable, who already knows

Process

1. Simulate seats, not boxes. For each key person: scope Δ, status Δ (title, audience, room access), boss Δ, identity Δ ("am I still the platform person?"). Losses count double — people fight losses harder than they chase gains. Mark each seat: champion / accepter / quiet-resister / flight-risk, with the because.
2. Find the undecideds. The middle managers and senior ICs who could go either way decide reorg outcomes; list who they'll ask about it in the first 24 hours — those people, not the memo, are the actual communication channel.
3. Forecast friction concretely. Not "there may be resistance" but "the two platform leads now share one headcount pool and will collide at Q3 planning". For each: likelihood, blast radius, the cheap preventative available before announcement (a scope clarification, a title fix, a pre-conversation).
4. Run the leak test. Rewrite the draft announcement's message as received by its least charitable reader from a screenshot with no context, no Q&A, no follow-up meeting. If the leaked version is fatal, the announcement isn't ready. Then write the corridor version — the one sentence people will actually repeat — and check it's survivable.
5. Sequence the rollout. Order: losers first, privately, from someone they trust (hearing it in the all-hands is how flight-risks convert) → undecideds' influencers → champions armed with the honest FAQ → everyone. Each conversation gets the one question it must answer for that hearer. Pressure-test the final plan against [[machiavelli-counsel]]'s friends-of- the-old-order column if politics run deep.