Walk one company from the basement to the top floor and you cross more distance than an org chart can show.
In the warehouse, decisions are physical: what this shift can move, what breaks, what the truck can carry. On the support floor, decisions are procedural: what the ticket allows, what the script covers, when a human overrides. On the sales floor, decisions are immediate: what closes this month, what the customer will forgive. On the product floor, decisions are bets: what we build while it is still cheap to be wrong. On the executive floor, decisions are existential: what we are becoming, and what we stop being.
Same building. Same payroll. Five different conditions, five different decision logics.
The stack
Each floor answers a different question first, because each floor lives in different consequences.
| Floor | Life conditions | The question that owns the floor | Native decision logic |
|---|---|---|---|
| Operations and production | Physical limits, safety, shift time | What must ship today? | Procedure plus seasoned judgment |
| Support and service | Emotional load, queues, policy | What does this person need without breaking the system? | Rules with human overrides |
| Sales and revenue | Quota, competition, customer memory | What closes, and what does it cost later? | Fast bets with personal stakes |
| Product and engineering | Uncertainty, rework cost, discovery | What should we build while being wrong is cheap? | Hypotheses, gates, reviews |
| Finance | Control, exposure, audit | What can we actually afford? | Thresholds and signatures |
| Executive | Direction, survival, legacy | What are we becoming, and what do we stop? | Judgment under incomplete information |
None of these is the best floor. Each logic fits its conditions. A warehouse that decided by consensus would be a dangerous place. An executive team that decided by procedure would be a slow one. The mistake is not having different logics. The mistake is pretending there is one.
Why the uniform system breaks differently on every floor
A company decides to install one decision system. One template, one cadence, one escalation path, one culture slide.
What actually happens is that each floor breaks in its own dialect. Consensus, installed top-down, works at the executive table where stakes are shared and time exists, becomes ceremony on the product floor, and becomes a safety incident waiting to happen in operations. The rulebook that protects finance suffocates sales. The bet-everything velocity that excites the top floor reads as recklessness on the floor where someone can lose a hand.
The uniform system does not fail everywhere equally. It fails where its assumptions are furthest from the floor's conditions, and holds where they happen to match. That is why every consultant can produce a testimonial and every operator can produce a counterexample. Both are telling the truth about different floors.
The deeper pattern is the one from the start of this series: a system works when the ground matches its assumptions. An organization is not one ground. It is a stack of grounds.
The seams are where decisions leak
Floors mostly run fine on their own logic. The failures concentrate at the seams: the handoff where one floor's decision becomes another floor's constraint.
Sales promises what operations cannot physically deliver. Product ships what support must then explain. Finance applies a threshold that quietly kills a product bet nobody upstairs remembers making. Each seam is a place where two legitimate decision logics collide, and whoever owns the seam owns the outcome.
Most organizations own none of them. The seam is everyone's job, so it is nobody's job, and it fails the way unowned edges always fail: ambiguity accumulates until consequence arrives faster than anyone can claim it.
The visitor problem
There is a reason transformation programs keep failing in the same way. The top floor writes the culture deck. The ground floor never reads it. Both are acting honestly.
The top floor genuinely lives in conditions where the deck describes real behavior: bets, vision, permission. The ground floor genuinely lives in conditions where the deck describes fiction: nobody there has ever seen a decision happen the way the deck describes. Nobody is lying. The building just contains more than one world, and the deck describes the one with the nicest view.
A decision system adopted at the top and rolled out downward is not adopted. It is performed, exactly as far as inspection reaches. This is workflow theater with a reorganization budget.
Coordinate the stack without flattening it
The goal is not uniformity. It is legibility between floors that decide differently:
- Let each floor run its native logic. Fit beats consistency.
- Name the seams explicitly, with an owner per seam. Sales-to-operations. Product-to-support. Product-to-finance.
- Share the minimum viable artifacts across floors: a decision log anyone can read, and an escalation path that is safe to use from the bottom floor to the top.
- Translate at the seams. A sales commitment should arrive at operations as capacity math, not as enthusiasm.
- Test the escalation path from below, occasionally, on purpose. If nobody bottom-floor can name what happens when they use it, the path is a diagram.
An organization that respects its own stack decides better than an organization that pretends to be one mind. The stack is not the problem. Pretending the stack is flat is the problem.
FAQ
Is this just saying frontline and leadership are different?
It is stronger than that. Every floor differs from the floor above it in life conditions, not just in seniority. The point is not respect for the frontline, though that helps. The point is that decision logic is grown by conditions, and a tall building contains several conditions under one roof.
Should each floor design its own decision system?
Within limits, yes, and the limits are the seams. Local decision logic should fit local conditions. What must be shared across floors is deliberately thin: how decisions get recorded, how escalation actually works, and who owns each handoff. Thin and shared beats thick and uniform.
Where do agents fit in a stacked organization?
Agents inherit the decision logic of whichever floor briefs them. An agent tuned by the executive floor will enforce executive assumptions on support-floor work, and the mismatch surfaces as friction nobody can name. Match agent contracts to the floor whose work they touch, and put a human at every seam an agent crosses.
How do I find out where our system is breaking first?
Follow the seams. Ask each floor what decisions arrive from other floors that make no sense in local conditions. The first seam named by three different floors without coordination is usually the one bleeding.
Related reading
- A Decision System Only Works on the Ground That Grew It
- The Spiral Is Not a Scoreboard
- Rooms Drift When No One Owns the Consequence
- Unowned Edges Fail the Same Way in Rooms and Systems
- Design the Escalation Layer Before You Add More Agents
- The Architecture of Decisions
If your decision system works on one floor and fails on another, book a discovery call.