Twenty-two risks. Two empty branches.

You can't answer that from a list. A list can only tell you what's on it. To know what isn't, you need somewhere for the missing thing to be absent from.

What a WBS actually does

Track 1 spent a week on the work breakdown structure, and most people took the wrong lesson out of it. They think a WBS is for organising scope. It isn't, or not mainly.

A WBS is a completeness test. You break the work into branches, then look at each branch and ask whether what sits underneath it adds up to the whole of that branch. The value isn't in the boxes you filled. It's in the box you couldn't fill, which is now sitting there in front of you looking empty.

A risk breakdown structure does exactly the same job for everything that can go wrong.

Same discipline, same payoff, and — as we're about to find out — the same uncomfortable moment when a branch stares back at you with nothing in it.

Eight branches

Here's the structure for this job. Eight branches, chosen because they match how a construction project actually fails, not how a textbook categorises risk.

Ground and subsurface. Anything below formation level that behaves differently from the assumption.

Design and information. Drawings, details, coordination, and the timing of all three.

Procurement and supply chain. Price, availability, lead time, and the financial health of the people you buy from.

Subcontract and interface. What's inside a package, what's outside it, and what fell between two of them.

Site, access and environment. Getting to the site, working on it, and the consents that let you.

Execution and productivity. Whether the work goes as fast as the rate assumed it would.

Client and employer obligations. The things somebody else has to do before you can do yours.

Commercial and contractual. Payment, valuation, retention, variation, and the mechanisms that decide who carries what.

Sorting the pile

Now take the twenty-two and put each one where it belongs.

Ground and subsurface takes three: the rock in the pile bores, the groundwater reading taken in the wrong month, and the existing services diversion.

Design and information takes four: late detailing, the rebar clash between the core wall and the transfer slab, the embedded items nobody coordinated, and the drawings still marked for tender that we are already building against.

Procurement takes four: steel escalation, the long-lead switchgear, the pump, and one supplier whose accounts are two years overdue.

Subcontract and interface takes all four of the seams — pile trimming, waterproofing to blinding, builder's work to the mechanical package, and the temporary water for testing.

Site, access and environment takes three: the unconfirmed wayleave, the neighbour's working-hours restriction, and the discharge consent for dewatering.

Execution and productivity takes three: the winter shutdown, the single-crane dependency, and the gang productivity nobody has measured.

That's twenty-one.

Notice what happened to the query log on the way. Its four answers scattered into four different branches — the wayleave went to site, the drawing status went to design, the rock went to ground, the temporary power went somewhere else entirely. One document, four branches. Which is the first clue that documents and categories are not the same thing, and that you need both.

The two empty shelves

Client and employer obligations has one item on it — the temporary power that Query 31 said was by others.

Commercial and contractual has nothing.

Twenty-two risks on a million-dollar contract, and not one of them is about money changing hands. Nothing about the payment period. Nothing about retention release. Nothing about how a variation gets valued, or what happens when the client disputes one. Nothing about delay damages. Nothing about a bond being called.

Cost & Cash Week 17 spent a whole article on the fact that this job carries $130,404 of working capital because the employer pays twenty-eight days after certification — a period that isn't a rule of nature, but a negotiated line in the Contract Data. That is a risk. It has a probability, it has an impact, and it is nowhere on this register.

Why they're empty

This is the part worth sitting with, because the reason isn't carelessness.

Go back to the six documents from last week. The ground investigation. The query log. The drawing register. The subcontract enquiries. The programme. The site file.

Every single one of them is a document about the physical works. What's in the ground, what's on the drawings, what's in the packages, how long it takes.

We never opened the contract.

Not once. We read hundreds of pages about how to build the thing and none at all about the conditions under which we get paid for building it — and the register came out shaped exactly like the reading list that produced it.

That's the real lesson of a breakdown structure, and it has nothing to do with tidiness. A structure doesn't organise what you found. It exposes the shape of where you looked. The two empty branches aren't telling you this job has no commercial risk. They're telling you which cupboard nobody opened.

There's an organisational version of the same point, and it's sharper. Look at the six full branches and ask who, in your company, owns the document that produced them. Ground belongs to the engineer. Design belongs to the design manager. Procurement belongs to the buyer. Subcontract and interface belong to whoever wrote the packages. Execution belongs to the site team.

Now ask who owns the contract. On most projects it's the quantity surveyor — and on most projects the quantity surveyor wasn't in the room when the register was built. The planner built it, from the planner's documents, in the planner's language.

The empty branch isn't a gap in the analysis. It's a gap in the invitation list.

Practical insight

Take your own register — however many lines it has — and sort it into these eight branches. Twenty minutes, on paper, no software.

Then count each branch and be honest about the thin ones.

Two branches come up short on almost every construction project, and they're the same two every time. Commercial and contractual is empty because the planner built the register and the planner doesn't read the contract. Client and employer obligations is thin because nobody enjoys writing down, in a company document, that the person paying you is a risk.

Both of those tell you something about the organisation rather than about the job.

And if a branch has nothing in it, there are exactly two possibilities. Either it genuinely carries no risk on this project — which happens occasionally, and you should be able to say why in one sentence. Or nobody looked. It is almost always the second one.

Key takeaways

✔ A list tells you what is on it. Only a structure can tell you what is missing.
✔ A WBS is a completeness test, not a filing system. An RBS does the same job for risk.
✔ Eight branches cover how construction projects actually fail: ground, design, procurement, subcontract, site, execution, client, commercial.
✔ Sorting the twenty-two put twenty-one into six branches and left two branches carrying one line between them.
✔ Commercial and contractual had nothing in it — on a contract carrying $130,404 of working capital and a negotiated payment period.
✔ The register came out shaped like the reading list. Every document we read was about the physical works; none of them was the contract.
✔ An empty branch means one of two things: no risk, or nobody looked. Assume the second.

What's coming next

We now have twenty-two risks in eight branches, and we know which two branches are lying.

But there's a problem underneath all of it that we've walked past twice. Go back and read line one of the original register: Rock in the pile bores.

That isn't a risk. It's a noun. It doesn't say what causes it, it doesn't say what happens when it turns up, and it doesn't say who is supposed to do anything about it. You can't put a probability on a noun, you can't tell when it has happened, and you certainly can't hold anybody to account for it.

Next week we rewrite all twenty-two properly — cause, event, effect — and find that a good number of them turn out not to be risks at all.