Case studies / Broadspectrum
Management control & financial architecture
Rebuilding contract profitability visibility across 20 mining-service contracts
A mining-services business running 20 contracts for seven major clients was losing money without being able to say where. The fix was not more reporting. It was changing the unit at which the business was measured.
Headline figures
- Basis
- Across seven major mining clients
- Basis
- Three days to one hour, measured on the accounting close cycle
- Basis
- Solution built internally instead of procured
- Frame
- Avoided replacement cost, not a cash saving.
- Basis
- 2014 to 2016
- Frame
- Company-level outcome delivered by the whole organisation.
The starting point
Broadspectrum — formerly Transfield Services — was an Australian infrastructure and industrial services group operating across Australia, the United States, Canada and Chile, with A$3.8 billion of revenue and more than 25,000 employees in FY2015. Ferrovial acquired it in 2016. The Chilean business served the mining sector.
By 2014 that Chilean operation was running at a sustained loss. Twenty service contracts for seven major mining clients were managed through fragmented systems: the ERP held the accounting truth, but everything needed to actually run a contract lived in static Excel exports that went stale the moment they were produced.
Contract managers received results weeks after the period they described, at a level of aggregation that told them the contract was underperforming but not which part of it. Forecasting was not a discipline — budgets were set once and thereafter referred to. Finance, Operations, Supply Chain and HR each optimised their own numbers, and the areas rarely met around the same figures.
Decisions were therefore reactive: corrective, after the fact, and frequently aimed at the wrong cost.
Diagnosis
The obvious reading was that the business lacked reporting. That reading was wrong, and pursuing it would have produced faster versions of the same useless numbers.
The real problem was the unit of analysis. A mining-service contract was being measured as a single block, when in practice it contained four businesses with different economics: the base service, priced and staffed against a fixed scope; additional services, requested ad hoc at variable margin; bonuses, tied to performance and settled on their own cycle; and investments, the start-up and closing costs that distort any period they land in.
Consolidated, these four offset one another. A healthy base service could mask sustained losses in additional services for months. No amount of reporting speed would have surfaced that, because the information was being destroyed at the point of aggregation, not at the point of delivery.
Separating them was not a reporting change. It changed the unit of accountability — and therefore who could be asked what.
Mandate and scope
- Reporting lineCFO and CEO
- OwnedContract-level financial architecture and the information model behind it; the ERP–SQL–Excel integration; forecasting and budgeting method; the monthly review routine.
- Worked withFinance, Operations, Supply Chain, HR and IT. The redesign only held because the areas adopted a shared set of figures.
- Not ownedCommercial negotiation of the contracts, and operational execution on site.
The operating model
Implementation
Before
- Manual data entry — every change made one record at a time
- Static Excel exports: frozen data, no ERP connection, each change means a new export
- Financial statements assembled by hand, with limited detail below contract level
- No contract-level review meetings; managers had no feedback loop on their own numbers
After
- Validated Excel tables integrated into the ERP database — change and adjustment cycle ≈20× faster
- Direct database query into Excel and Power Pivot — report refresh ≈10× faster
- Data cube with drill-down to transaction level by contract, with actual, forecast and budget in a single view
- Monthly review cycle: 20 contract managers in two days
The technical integration was the smaller half. What made the change hold was the review routine built on top of it: every month each contract manager reviewed their own numbers against forecast and budget, in a format they could produce themselves. Forecasting stopped being a finance exercise and became the contract manager’s own statement about the rest of the year.
Improvement work then ran cross-functionally — Finance, Operations, Supply Chain, HR and IT — using Lean and Six Sigma, and prioritising the company result over the area result. That was the part that could not be bought.
Results
| Measure | Before | After | Basis of measurement | Result |
|---|---|---|---|---|
| Month-end close | 3 days | 1 hour | Accounting close cycle across 20 contracts | −99% |
| Report refresh | Manual re-export | Query refresh | Database query into Excel / Power Pivot | ≈10× |
| Changes and adjustments | One record at a time | Batch, validated | Integration of validated tables into the ERP database | ≈20× |
| Contract review cycle | No routine | 20 managers in 2 days | Monthly review with contract managers | 20 / 2d |
| External development | Procurement path | Built internally | Avoided external development and software cost — not a cash saving | ~€100K |
| Company result | Sustained losses | Profitability | Company-level outcome over 2014–2016, delivered by the whole organisation; the management-control workstream supported it | 24 mo. |
“Broadspectrum wishes to recognise your willingness to embrace change, your ability to seek new ways of carrying out tasks, the integrity with which you perform your duties, and your readiness to work as part of a team.”
The letter is dated May 2016 and records expanded responsibilities — the same period in which the operation returned to profitability and Ferrovial completed its acquisition of Broadspectrum. The document itself is not reproduced here.
What I carry forward
- Diagnose the unit of analysis before building anything. Most reporting problems are accountability problems wearing a reporting costume.
- Give the number back to the person who owns the outcome. A forecast produced by finance is a document. A forecast produced by the contract manager is a commitment.
- Design for adjustment, not for the first run. The systems that survive are the ones where changing an assumption is cheap.
- Integration buys speed; only the routine converts speed into result. The meeting cadence did as much work as the data model.
Scope and measurement note
Figures reflect internally measured operational and financial reporting cycles during 2014–2016. The ~€100,000 refers to external development and software cost avoided by building the solution internally; it is not presented as a cash saving. The return to profitability was a company-level outcome delivered by the whole organisation — the management-control workstream described here supported it. Contract identifiers and client names are omitted.