A division's P&L is a system. Before you act on a loss, make sure the reporting tells you the truth.
Renzo Demartini, Chartered Accountant and finance executive · · 4 min read Drafted with AI, reviewed and spot-checked by me.
Exhibit 1Is this loss-making division worth closing?
After allocated overheadContribution
Service−A$40KA$60K
Projects−A$15KA$25K
MaintenanceA$20KA$45K
Small works−A$35K−A$10K
Overhead is spread by revenue. Three of four segments show a loss.
Count only each segment’s own costs. Service pays most; only Small works loses.
Allocated profit shows three losers; contribution shows only Small works fails to cover its own costs.
Illustrative figures, not client data.
In short
Rebuild the divisional P&L around allocations, job costing, pricing and mix before deciding anything.
The gap between quoted and delivered margin is usually where the loss is hiding.
Fix, reshape or exit on contribution, not on fully allocated profit, with a time limit and measures.
At an A$40M renewable-energy retailer, I was Head of Finance. Its commercial division was losing money, and I rebuilt the division’s P&L so the reporting showed where the loss came from.
The first step wasn’t a cost cut or a price rise. It was reading the division properly. A loss is a symptom. Until the P&L tells you where it comes from, every decision is a guess.
Exhibit 2
A re-engineered P&L took the division from loss to profit
Head of Finance
A$300K
Division profit after 12 months
Up from a loss, at an A$40M renewable-energy retailer.
Head of Finance
6%
EBIT margin
Reached in the same 12 months.
Results from past roles aren’t a promise of what will happen in yours.
Rebuild the P&L around four things and the loss shows where it sits
01
Allocations
Spread shared costs by what drives them, not by revenue. Show contribution before allocations.
02
Job costing
Compare quoted and delivered margin on every job. That gap is usually where the loss hides.
03
Pricing
Check what was quoted, what was discounted and who approved it.
04
Mix
Averages lie. Split results by customer type, product, job size and channel.
Illustrative figures, not client data.
Most divisional P&Ls are built for the statutory accounts, not for running the division. They answer ‘Did we make money?’ but not ‘Where, and why not?’
More detail
Allocations.
Shared costs spread in proportion to revenue hide almost everything. A division that sells a lot looks expensive, and a division that uses a lot of shared support looks cheap. Allocate on what actually drives the cost: people, usage, transactions, space. Then split the P&L into three layers: direct costs, costs the division controls, and costs it is charged. Show contribution before allocations. That is the line the division’s leaders can act on.
Job costing.
In a project business, the job is the unit of truth. For every job, compare the margin you quoted with the margin you delivered. Capture labour, materials, subcontractors, variations, rework and callbacks against the job, not in a general pool. Match revenue and cost to the same period. The gap between quoted and delivered margin is usually where the loss is hiding.
Pricing.
Look at what was quoted, what was discounted and who approved the discount. Check whether price reflects the real cost to deliver, or last year’s price plus a bit.
Mix.
Averages lie. Split results by customer type, product, job size and channel. A division making a thin overall margin can be two businesses: one that earns well and one that loses money on every sale.
The questions to ask
Exhibit 4
Seven plain questions show where the loss sits and what closing would change
Where is the loss?
Which jobs made money, and which didn’t?
Is the loss spread across all the work, or concentrated in a few jobs or one type of customer?
How did it open up?
Where does the gap between quoted and delivered margin open up: estimating, buying, scheduling or rework?
Did we win work we priced to lose?
Who sets the price, and on what information?
What would closing change?
If the division closed tomorrow, which costs would go and which would stay?
Is the division growing the work it does well, or the work it does badly?
Once the P&L is rebuilt, ask plain questions and insist on plain answers. The answers usually narrow the problem quickly. Rarely is everything broken.
The levers
Exhibit 5
Six ordinary levers, and the last one holds the other five in place
01
Price
Quote discipline, a minimum margin and clear approval for discounts.
02
Mix
Chase the work you deliver well. Decline or reprice the rest.
03
Cost to deliver
Estimating accuracy, purchasing, scheduling and less rework.
04
Overhead
Size the support to the work the division actually does.
05
Cash
Deposits, progress claims and debtor days. A profitable division can still drain cash.
06
Visibility
A monthly divisional P&L, with job-level margins, that its leaders read and own.
The levers are ordinary. The skill is in seeing which ones matter for this division. When the people who run a division can see their own margin, job by job, they start making different decisions without being told.
Fix, reshape or exit
Exhibit 6
Fix the execution, reshape the mix, exit only when no work can pay
Fix
The core work earns a margin once it is properly costed
The problems are in execution: pricing, estimating, delivery
A management task, with a plan and measures
Reshape
Some segments earn well and others don’t
Narrow the scope
Keep the customers and work types that pay, and step away from the rest
Exit
On true contribution, it can’t cover its direct and avoidable costs
And there is no credible path to doing so
Or the capital it ties up would earn more elsewhere in the business
Make the exit decision on contribution, not on fully allocated profit. Close a division and its share of overhead stays behind, landing on everyone else.
Also ask what the division does for the rest of the business. Some divisions feed others with customers, volume or capability. That value is real, even if it doesn’t show in the divisional result.
Whichever way you go, set a time limit and measures. A fix plan with no deadline is how a loss-making division stays loss-making.
The people side
Exhibit 7
A rebuilt P&L changes results when the division’s leaders own it
01Share it
Walk the division’s leaders through the rebuilt P&L.
02Give accountability
Clear accountability for margin at the job level.
03Give decision rights
Clear decision rights on price.
04Meet monthly
On the numbers, the same way every month.
A rebuilt P&L is a management tool, not a verdict.
What to do first
Exhibit 8
Five steps take you from a loss to a decision with a deadline
01Rebuild the P&L
Around allocations, job costing, pricing and mix.
02Ask the seven questions
Insist on plain answers. They usually narrow the problem quickly.
03Pick the levers that matter
Price, mix, cost to deliver, overhead, cash and visibility. See which matter for this division.
04Choose on contribution
Fix, reshape or exit, with a time limit and measures.
05Walk the leaders through it
Then meet monthly on the numbers, the same way every month.
Common questions
How do you read a loss-making division?
Read it properly first. Rebuild the divisional P&L around allocations, job costing, pricing and mix, and find where the loss sits. Then choose to fix, reshape or exit, with a time limit and measures.
Should you close a loss-making division?
Decide on contribution, not on fully allocated profit. Close a division and its share of overhead stays behind, landing on everyone else. Also ask what the division does for the rest of the business.
What is the difference between quoted and delivered margin?
Quoted margin is what you priced the job to earn. Delivered margin is what it earned once labour, materials, subcontractors, variations, rework and callbacks were counted. The gap between them is usually where the loss is hiding.
Chartered Accountant (CA ANZ) and finance executive, based in Canberra. Practical AI and automation for SMEs, and adviser to owners, CEOs and boards on integration and finance.
Want reporting that shows where a division makes and loses money, built into the systems you already run? That is the work I do.
Renzo Demartini