Set a restructuring on Marden, a synthetic 270-person company, then change the pace or the support and see where the modeled load travels.
Real engagements go far deeper. This public view is a simplified read on a synthetic company. Full work adds scenario sensitivity, role-level network structure and decision alternatives, on your own organization.
How the Decision Room is built, in ten short statements.
Marden is fictional.
Marden is a synthetic 270-person European technology and business-services company. Every figure on this page is read from a fixed, versioned result set precomputed for that synthetic organization. Nothing is simulated in your browser, and no text is generated on the fly.
A range, not a point.
Each headline number is a range: the middle half of the modeled outcomes, 25th to 75th percentile. A trajectory describes how the model expects the change to unfold across a 26-week window, not a guarantee.
A pattern, not a score.
A Changeprint is the pattern of modeled change a WhatIF produces: four headline ranges, an executive read, what changed and where load traveled.
Load belongs to roles, never to people.
Demand load and delivery trajectory describe departments and roles under the plan. The model projects how work is allocated; it never describes a person.
Calibrated on 1,102 studies.
The model's response parameters are calibrated on 1,102 published studies of how people responded to organizational change. Calibration bounds the model; it does not turn any single run into a forecast.
200+ calibrated coefficients.
In a full engagement, the engine runs on more than 200 calibrated coefficients across its state model, and every run records the engine version and the scenario.
Up to 500 runs per decision.
In a full engagement, a decision is rehearsed across as many as 500 simulation runs, so every output is a distribution, never a single path.
A Loadpath is a route, not a verdict.
A Loadpath is a modeled route along which added coordination, dependency or transition load travels between departments during the simulated window. One to three are shown, most consequential first.
A Switchpoint is a position in the network.
A Switchpoint is a role whose position connects functions that are otherwise weakly linked. It is a property of the network structure under this Base State, not a judgment about the person in the role.
Same inputs, same bytes.
Each WhatIF resolves to one entry in a versioned result set generated offline, so the same inputs always return the same output. What is modeled: structure, dependencies between departments, demand load on roles, dependency concentration, delivery effects, and how these travel over 26 weeks.
How these figures were produced. Each WhatIF range comes from 100 engine runs, each paired with a Base State run that shares its random seed. For this demonstration the engine runs with a small Decision Room adapter: conflict and peer influence are averaged over each role's connections so large departments stay in a realistic range, and each change reaches only the roles whose reporting line changes and the managers who take them on. Pace staggers when each team switches across a 30, 60 or 90-day window. Demand load and delivery are measured over the departments in scope; Loadpaths show where load travels beyond them. Bands move from moderate only when the modeled change exceeds 2 points. The model does not include span-of-control workload or how people anticipate an announced change. The figures are demonstrative: the model is calibrated on published research, and no real organization's outcomes have been compared with it.
Full engagements use masked, role-coded positions, never names.
Explore a real decision