The machine

Leontief, applied to the enterprise.

  • Skills are economic inputs.
  • Workforce decisions are financial decisions.
  • Impact can be quantified.
  • The same logic applies inside companies.

Wassily Leontief’s input-output framework showed how the sectors of an economy depend on one another, so that a change in one could be traced through all the rest. GWI adapts that framework to the organisation. The enterprise is treated as a dependency system rather than a collection of isolated jobs, and workforce capital is analysed with the same discipline as financial capital: value, exposure, redundancy and opportunity on a single analytical surface.

Economic impact

We already measure how industries, spend and jobs contribute to a nation’s GDP.

Three components traced through national input-output tables to output.

Corporate impact

Why couldn’t we do the same thing for skills, capabilities and dependencies to understand how they contribute to a company’s revenue?

Companies are systems too.

The four layers

A repeatable model, built in four layers.

  1. 01

    Verified intake

    Corporate structure, workforce composition, organisation design, compensation and financial data, extracted from the documents the organisation already holds and reconciled against each other. Every figure carries its source, and anything the data cannot corroborate is confirmed by a person before it enters the model.

  2. 02

    Dependency mapping

    The enterprise as an input-output system. Each unit of work consumes the output of others and supplies its own onward: sales depends on product, product on engineering, all of them on finance, legal, security, operations and institutional knowledge. The matrix records those flows, so the model can say what a given body of work economically enables, not only what it costs.

  3. 03

    Scenario and forecast modelling

    Decisions as levers on the model: automate a task, remove a role, redeploy a team, deploy a tool, acquire a unit. Each scenario propagates through the dependencies and returns the capacity it frees, the capacity it consumes elsewhere, and the revenue and risk at the end of the affected chains. The forecast and the scenario are kept apart so a what-if never silently becomes the plan.

  4. 04

    Attribution and return

    The question financial statements cannot answer: did the combination of technology and people create more enterprise value than the process it replaced? The output is a fixed-format assessment with an executive summary, unit economics and the evidence behind every number, reviewed through a sign-off gate before publication.

Where it applies

One architecture, three contexts.

Enterprise

“What does this body of work enable, and what happens to that if we change it?”

Operating-model design, AI transformation, capability transition and restructuring, decided on attribution rather than headcount.

Transactions

“What workforce am I acquiring, and where is the value or the redundancy?”

The same layers applied to a target, so the value-creation plan is tested against the dependencies it would cut.

Regions and policy

“What does this programme, site or instrument contribute to the economy around it?”

Economic Impact Assessment on published national input-output multipliers, documented to a standard a funding body can audit.

Why GWI

GWI is a neutral economic methodology and measurement layer. We endorse no technology or advisory solution. LEON includes technology and change-management ecosystem partners who can support clients through workforce redesign, but those relationships do not influence our assessments.

We do not share assessment revenue with ecosystem partners. There is no pay-for-play and no vendor favouritism.

LEON is not built to argue that more people are always better, nor that technology should replace work by default. It is built to give clients the data before they make the decision.

The method is young and we want it challenged: by economists, CFOs, workforce leaders, technologists, investors and the people doing the work. If the thesis is wrong we want to know where it breaks.