Introducing Clu Workforce Risk Models™: Audit-Grade Measures of Organisational Design Efficiency
Clu Labs research release, April 2026
Clu Labs today introduces Clu Workforce Risk Models™, five configurable calculations that measure duplication, bloat, alignment, agility and augmentability across an organisation’s roles.
Each model is deterministic, so the same data and the same settings always produce the same result. Each is built on Clu Cognitive Fingerprinting™ and the Clu Atomic Skill Taxonomy™.
Organisational psychologists and former partner-level accountants from Big Four firms developed the models with us. They are built to an audit-grade standard, so a result can be checked, explained and reproduced by someone outside Clu.
The research problem
Workforce decisions carry financial risk on the scale of any major investment. A restructure, an automation programme, or a merger can change an organisation's cost and capability for years. Yet the evidence behind most of these decisions would not pass the scrutiny applied to a set of accounts.
Measures of duplication or bloat usually come from spans and layers, consultant interviews, or inference models that predict what roles contain. These methods give different answers depending on who runs them and when. Leadership teams cannot easily show a board, a works council, or a regulator how they reached a figure.
Clu Labs set out to build measures of workforce structure that meet the same standard as financial reporting: clear definitions, documented assumptions, consistent methods, and results traceable to their source.
Two disciplines, one method
Organisational psychologists defined what each model should measure. Their role was to ensure each measure captures a real feature of how work is designed and describes roles rather than the people in them.
Former partner-level accountants from Big Four firms shaped how each model is calculated and evidenced. They brought the practices used to make financial figures defensible: every input documented, every assumption stated, every setting recorded, and every result reconcilable to the client’s own data.
Clu Labs combined the two into calculations that run on fingerprinted atomic skills. Because every skill in the Clu Atomic Skill Taxonomy describes one capability in the same structured way, the models can compare any two roles directly.
Clu does not publish the Risk Model rules, but clients can access them under NDA.

The five models
Duplication. Where the same work is held in several roles in similar contexts, and how much overlap truly exists between them.
Bloat. Where structure has grown beyond the work it supports, such as layers or roles whose work is largely held elsewhere.
Alignment. Whether roles actually match what the organisation says it needs.
Agility. How easily work could move between roles, based on how closely their skill fingerprints match.
Augmentability. How much of each role’s work AI, automation or robotics could do, and, importantly, what should always remain human.
None of the models assesses individuals. Each describes how work is designed and distributed across roles.
What makes the models configurable
Organisations differ on what they consider acceptable overlap or a sensible span of control. Clients can set thresholds and weightings to match their own policies and risk appetite. We record every setting alongside the result it produced, so you can reproduce any figure later from the same data and settings.
How the models were tested
We reconciled results to clients’ headcount and cost data in 100% of the 11 organisations tested.
Re-running the same data with the same settings produced the same result in 100% of cases.
Independent reviewers agreed with each model's flagged roles in 92% of 1,400 sampled roles.
Across the organisations tested, 23% of roles had work duplicated in at least one other role, and 34% of the work in tested roles could be partly performed by AI, automation, or robotics.
What it means for clients
Leadership teams can now put the same rigour behind workforce decisions that they expect from financial ones. Before a restructure or automation programme begins, they can see where work is duplicated, where structure has outgrown the work, where roles have drifted from what they were designed to do, how easily work could move, and how much of it tools could take on.
Each figure includes its inputs, settings, and method, ready for review by a board, an auditor, a works council, or a regulator.
Frequently asked questions
What are Clu Risk Models? Five configurable, deterministic calculations from Clu Labs that measure duplication, bloat, alignment, agility and augmentability across an organisation’s roles.
Who developed them? Clu Labs in partnership with chartered organisational psychologists and former partner-level accountants from Big Four firms.
What does audit-grade mean here? Every input is documented, every assumption stated and every setting recorded, and every result can be reconciled to the client’s own data and reproduced by someone outside Clu.
Do the models assess individual employees? No. Each model describes how work is designed and distributed across roles.
Do they use generative AI? No. The Clu platform uses no generative AI. The same data and the same settings always produce the same result.
To see how Clu Risk Models can apply to your organisation, contact Clu's research team directly at labs@getaclu.io.
About Clu
Clu provides work intelligence for regulated organisations. It helps leadership teams identify and reduce the risk of their biggest workforce decisions, including restructuring, AI transformation, M&A and changes to capability models, using the structural data they already hold. Clu Labs is Clu's research arm.



