Total Workforce Baseline: Counting the Whole Crew
- Maturity
- L1
- Domain
- Plan & Cost
- Analytics
- descriptive
- Permanent headcount
- 38,412
- Sponsor
- VP, Workforce Planning & People Analytics
- Confidence
- Moderate
The situation
How many people does it actually take to run Meridian — and what does that whole workforce cost — once contractors are counted the way the operation actually uses them?
The recommendation on the table
Adopt one governed total-workforce definition — head, FTE and contractor FTE-equivalent — and report it to the board every month.
Ends the five-systems-five-answers problem and makes the ≈30% of cost that is contractor labour permanently visible at portfolio level.
Trade-offBusiness units must give up their local headcount definitions and accept a built-in ~30-day lag on the contractor figure.
The evidence
Meridian could not answer a deceptively basic question — how many people does it take to run the company — because its 14,000–22,000 contractors lived outside the HR system and five source systems disagreed on the permanent count. This project built one reconciled total-workforce view: 38,412 permanent staff and roughly 18,000 contractor FTE-equivalent, and it showed that contingent labour is about 30% of total workforce cost yet appears in no board headcount report. This is descriptive work by design — you cannot forecast a cliff you cannot count.
Meridian — Total Workforce View
Give leadership one trusted view of the whole workforce — permanent and contractor — and make the ≈30% of cost that contingent labour represents visible for the first time.
- Permanent38,41268%
- Contractor FTE-equiv.18,00032%
Key takeawayNearly one in three of the people delivering the work is contingent labour.
Key takeawayUpstream and Downstream carry almost all of the ≈$1.4B contractor spend — click a unit to drill in.
Key takeawayThe workforce is genuinely distributed — which is what justifies a geographic view here, unlike a hub-based population.
Key findings
Once contractors were counted as workforce rather than as procurement, roughly 30% of Meridian's total workforce cost — about $1.4B a year — turned out to be invisible to the people accountable for workforce strategy. The organisation had been planning around the two-thirds it could see.
What we can’t claim
The reconciled total-workforce figure is a best estimate, not a precise count. It rests on a contractor-hours feed that lags by about 30 days and an FTE-equivalence assumption, which together give the contingent number a band of roughly ±2,000 FTE. Presenting it as exact would repeat the error this project set out to fix.
Recommendations
Adopt one governed total-workforce definition — head, FTE and contractor FTE-equivalent — and report it to the board every month.
high priorityEnds the five-systems-five-answers problem and makes the ≈30% of cost that is contractor labour permanently visible at portfolio level.
Trade-off
Business units must give up their local headcount definitions and accept a built-in ~30-day lag on the contractor figure.
Bring contractor spend under workforce governance, starting with the ≈$1.4B in drilling, turnaround and EPC contracts.
medium priorityGives a single owner accountability for roughly a third of workforce cost that today no one manages as workforce.
Trade-off
Procurement and HR must share decision authority each has historically guarded separately.
Analytical framework
How we reached this
Descriptive workforce accounting — reconcile the true count and movement of the whole workforce, permanent and contingent.
ConfidenceMedium-High
Analytical framework
How we reached this
Descriptive workforce accounting — reconcile the true count and movement of the whole workforce, permanent and contingent.
Methods applied
Statistical techniques
Algorithms
Data sources
Outputs generated
Why this confidence
Permanent headcount is reconciled and solid; contingent is FTE-equivalent with a stated ±2,000 band, which caps confidence below High.
The reasoning
Business context
Meridian is a ~$48B integrated operator running upstream, midstream, downstream and trading across offshore platforms, onshore fields, refineries and terminals. Workforce data sat in five systems that had never been reconciled: the core HRIS, two contractor-management systems, the SAP-based cost ledger and a refinery rostering tool. Analytics maturity was about 2.5 — competent reporting on permanent staff, but no integrated view and no governed definition of the contingent workforce at all.
Expected value
A reconciled baseline is the prerequisite for everything downstream: succession risk (MER-02) and forecasting (MER-05) are impossible without it. Concretely, the project reconciled a 6% discrepancy between systems on permanent headcount and surfaced roughly $1.4B of annual contractor spend that no single owner managed as workforce, converting an invisible third of workforce cost into something governable.
Workforce landscape
The permanent population splits across operations and field technicians (34%), engineering and subsurface (22%), maintenance and turnaround (15%), corporate functions (17%), commercial and trading (6%) and leadership (6%). The contingent population is structurally different by type: drilling and well-services crews, turnaround and maintenance labour that spikes on a cyclical calendar, EPC project contractors and specialist consultants — concentrated in upstream and refining, and almost entirely outside the HRIS.
The analytics journey
Deliberately Level 1, descriptive. The work began with definitions, not dashboards: separating head from FTE from contractor FTE-equivalent so the organisation could stop arguing about which number was right. It then reconciled the five sources into one governed taxonomy and published a single total-workforce view. It makes no prediction and draws no causal claim — and it says so plainly, because a baseline that pretends to be more than a baseline is how analytics loses credibility before it starts.
Under the hood
Sources: core HRIS, two contractor-management systems, the SAP cost ledger and the refinery rostering tool. Contractor FTE-equivalence is derived from billed hours against a standard annual-hours basis, which is the source of the ±2,000 FTE band. Known caveats stated up front: the contractor hours feed lags by roughly 30 days, so the contingent count is always a recent estimate rather than a live figure, and the FTE conversion assumes standard hours that overstate part-cycle crews slightly. Governance: one definition, one owner, monthly publication.
Confidence & evidence
Why you can rely on this
The inconvenient truth
The reconciled total-workforce figure is a best estimate, not a precise count. It rests on a contractor-hours feed that lags by about 30 days and an FTE-equivalence assumption, which together give the contingent number a band of roughly ±2,000 FTE. Presenting it as exact would repeat the error this project set out to fix.
Method
Confidence is a deterministic read of KPI strength, target and benchmark coverage across this project — shown on an illustrative reference dataset, computed the same way it would be on live data.
Take this further