The delivery note was never built for this
Material arrives. Somebody signs for it. The note carries one project code, and into the cost ledger it goes, against one work item.
Then you look at where the material actually went. Shared consumables in particular — paint, fasteners, pipe, cable — rarely go to one place. The same consignment gets drawn down across several work items and several locations over the following days, and the single code on the note stops describing anything real.
This isn't an error. The note is correct. It was produced to prove that goods arrived and were accepted, and it does that perfectly. It was never designed to allocate cost, and asking it to is your problem rather than the storekeeper's.
That distinction runs through this whole week. Almost every record the commercial side gives you is accurate for its own purpose and approximate for yours.
What the commercial side actually holds
Three things reach you from that direction, and they behave differently.
Commitments are money the project has promised: orders placed, subcontracts signed. They exist long before anything is spent, and they are the earliest warning you get that a package is going to cost more than the budget said.
Accruals are work done but not yet invoiced. They are an estimate by definition, and they are the line where your progress figure and the ledger touch each other most directly.
Invoices are what has actually been billed. They are the most certain of the three and the slowest, which is exactly the wrong combination for a monthly report.
Cost & Cash Week 10 teaches what these are and how they behave inside one set of books. This week is about receiving them from somebody else, in their format, on their timetable.
The code is where it breaks
Every cost that arrives is tagged to something, and the tag came from the code of accounts the project set up before any of this started. Cost & Cash Week 6 is where that structure gets built, and it is worth reading again from this side.
The problem isn't that the coding is wrong. It is that the commercial breakdown and the physical breakdown were designed to answer different questions. Commercial needs a structure that maps to the contract and to payment. Planning needs one that maps to where work happens. On a good project these two are reconciled at the start; on most, they are cousins rather than twins, and a cost that lands cleanly in one lands ambiguously in the other.
Shared material makes this visible faster than anything else, which is why it is worth watching. If you can explain where a consignment of cable went, you can usually explain the rest of the ledger.
On somebody else's calendar
None of it arrives when you want it. Finance closes on a financial calendar and you close on a progress one, so a delivery note that lands late or a measurement finished a few days after your cut-off is counted in a different month by each of you.
That is a calendar problem rather than a commercial one and it belongs later in the track, where all three of a project's closing dates can be looked at together. What matters here is narrower: the cost figures you receive were closed to a date that isn't yours, and using them as though they were is the most common way a cost report and a progress report come to disagree.
The cost that has not been billed yet
Accruals deserve a moment on their own, because they are the line where the two sides of the project touch most directly and disagree most easily.
An accrual is an estimate of work done but not yet invoiced, and somebody has to make it. Where it is made from a progress percentage, it inherits everything Week 8 described: if the underlying quantity was estimated rather than measured, the accrual carries that optimism straight into the cost report, where it looks like a hard figure.
Which is a reason to know how your project's accruals are built. If the answer is that they come from your progress number, then your progress number is doing two jobs, and an error in it is now an error in two documents rather than one.
Whose job it is to notice
Finding this sort of thing isn't the planner's responsibility, and the correction isn't the planner's decision. But the planner is usually the one who spots it, for a structural reason: project controls is the only function that sits with the site record, the survey measurement, the store movement and the ledger open at the same time. Nobody else has cause to compare them.
So the pattern that works is the one from Week 8, applied to a different pair of records. Write down which two disagree and by roughly how much. Send it to the commercial team and the store, with the site engineer copied, because they are the ones who know where the material went. Then let commercial decide, because the allocation is theirs to make.
Planning doesn't manage cost. Planning makes the inconsistency visible to the people who do.
System design
The reconciliation isn't about the numbers. It is about the items that crossed the boundary between two closing dates, and it takes twenty minutes if both dates are written down beforehand.
| Record | Produced by | Required quality | Verified against | Feeds |
|---|---|---|---|---|
| Commitments | Commercial | Coded to the breakdown you report on | Purchase orders | Forecast |
| Accruals | Commercial | Stated basis: from progress, or independently estimated | Progress figure | Cost report |
| Invoices | Commercial | Dated to the work period, not the payment date | Delivery and measurement | Actual cost |
| Cost cut-off date | Commercial | Written down, and known to you before the month closes | Finance calendar | Reconciliation |
Two dates and one short list. Earned value calculated across two unreconciled month-ends is wrong even when both sides are individually correct.
Practical insight
Pick one shared material on your project — something that goes to several places, such as cable, fasteners or paint. Take last month's deliveries of it and ask two questions.
First: how many work items does the ledger think it went to? Usually one, sometimes two.
Second: how many does the site say it went to? Ask the engineer, not the system.
If those two answers are different, you have found something worth an email rather than an argument. And you have found it on the one material where it is easiest to demonstrate, which is why it is the one to start with.
Key takeaways
- A delivery note is correct for logistics and insufficient for cost allocation. It was built for a different question.
- Shared consumables expose the gap first, because a single code can't describe several destinations.
- Commitments, accruals and invoices arrive with different certainty and different lag.
- Commercial coding maps to the contract. Physical coding maps to where work happens. They are cousins, not twins.
- Finance and progress close on different dates, and work done in one month can land as cost in the next.
- Agree both cut-off dates, write them down, and reconcile the items that crossed the boundary before publishing.
- Accruals built from a progress percentage inherit whatever was wrong with that percentage, into a second document.
- The planner notices because the planner is the only one with all the records open. The allocation decision still belongs to commercial.
Records born here. Cut-off calendar · boundary item list · monthly reconciliation note.
What is coming next
Everything so far has come from inside the project. The last source is outside it, and it doesn't send data at all.
Next week: the client and the consultant — instructions, approvals, comments, and why a turnaround time is a programme input rather than an administrative detail.
Enjoyed this lesson?
Join with Google to get each new lesson the moment it's published — and help me see which topics matter most to you. No spam, one email a week, unsubscribe anytime.
Already following on LinkedIn works too — this is just for the weekly email.