Diagnostic record · Diagnosis published, results pending
Four conversations located the constraint: capacity had been acquired faster than the system that staffs it.
Industrial fabrication group · ~$500M revenue · multi-plant. Identity, brand, locations and individuals withheld. This record publishes the diagnosis only — no result is claimed, because the measurement window has not opened.
← All evidenceCondition 01 · Growth has outrun structure
The business had become larger than the system that runs it. Backlog and plant capacity were secured; the acquisition of the people to run them was still being handled as a series of campaigns.
See the eight enterprise conditions →How the constraint was found
Four stakeholder reads, one question each: what does the system produce, and what does it cost when it does not?
Chief Financial Officer
“Capacity and backlog have been bought and won. The cost of not staffing the plants is now larger than the cost of the hiring system itself.”
What it surfaced: The constraint is economic, not administrative: unstaffed capacity is already priced into the plan.
Former head of human resources (now in a finance seat)
“Every prior fix was a source, a vendor or a campaign. None of them changed how the work of hiring is designed.”
What it surfaced: Repeated interventions at the tactic layer, none at the architecture layer.
Human resources directors, multi-plant
“Requisitions arrive without a defined standard of work, and each plant runs its own sequence, forms and decision points.”
What it surfaced: No single acquisition system exists; there are as many systems as there are plants.
Compensation & benefits lead
“Offers are competing against a market read that is not current, and pay steps are not tied to a certified skill level.”
What it surfaced: Yield is lost at offer because economic truth is absent at the decision point.
No survey, no data warehouse and no discovery phase preceded these reads. The architecture of the diagnosis — a defined present-reality question asked of each stakeholder who holds part of the system — produced a single agreed constraint statement in a matter of hours.
The binding constraint
Talent acquisition had never been architected as an operating system. Sourcing, selection, offer economics and onboarding sat in separate hands with no standard of work, no shared measure and no cadence — so hiring performance could only ever equal the effort applied to it that week.
Proposed architecture
SEE → FOCUS → ARCHITECT → INSTALL → TRANSFER
One governed acquisition system: a current market read at the point of offer, a defined standard of work from requisition to certified skill level, decision rights and pay steps tied to that standard, one console of measures, and an ARC™ renewal loop the client's own leaders run. Compass leaves business architects, not dependency.
Measure baseline register
Six measures, their baseline status today, and the direction the architecture must move them.
| Measure | Baseline status | Target direction |
|---|---|---|
| Time from requisition opened to accepted offer | Not measured enterprise-wide | Reduce |
| Offer-to-acceptance yield | Reported by plant, not comparable | Increase |
| First-90-day retention of direct labor hires | Not tracked to hire source | Increase |
| Unstaffed capacity (open positions against plan) | Known in aggregate only | Reduce |
| Cost per hire, fully loaded | Partially captured; agency spend outside the read | Reduce |
| Share of hires filled through one governed process | No single process defined | Increase |
No after-figures are published for this engagement. The measurement window opens at INSTALL; results will be published only once measured and only if the client permits anonymized disclosure.
Four spend gates
Every stage is approved on evidence. Nothing beyond the current gate is committed.
Gate 01 · See
- Opens when
- The present-reality read is complete and the constraint is agreed in writing by the executive group.
- Funded scope
- Diagnostic only — stakeholder reads, process observation, baseline definition.
- If not approved
- The engagement stops with a written diagnosis the client owns. Nothing further is committed.
Gate 02 · Focus
- Opens when
- The vital few interventions are named, sequenced and costed against the measures above.
- Funded scope
- Design of the acquisition architecture and the measurement baseline.
- If not approved
- No build begins. The diagnosis stands and can be re-opened at any time.
Gate 03 · Architect & Install
- Opens when
- Baseline measures are instrumented and named client owners accept the standard of work.
- Funded scope
- One governed acquisition system, standard work, decision rights, cadence and console.
- If not approved
- Design is transferred as-is; the client may install it with internal resources.
Gate 04 · Transfer
- Opens when
- Client leaders have run the cadence unaided for two full cycles with evidence recorded.
- Funded scope
- Capability transfer, ARC™ renewal loop, Compass exit.
- If not approved
- Support continues only by explicit extension — dependency is never the default.
Client identities, brands, and locations are withheld under confidentiality. Metrics are client-reported and measured over the window shown; they describe those engagements and are not a projection of results for any other enterprise. Entries labelled "Composite example" are illustrative composites drawn from patterns across multiple engagements, not the record of a single client. This entry is labelled a diagnostic record: it publishes an agreed diagnosis and a proposed architecture, not a measured result.
Begin a conversation →Not sure where to start?
Three questions, and we point you to the right instrument.
Under a minute. From this page, most leaders begin with The organization.