Everyone Remembers the Big Blow. Nobody Counts the Jabs.

Ask anyone who worked on a difficult project what went wrong and you'll usually get an answer within about ten seconds. The design change in month four, the subbie who walked or the on and off again rain for six weeks. There is nearly always one single event that the whole team agrees on and it gets named in the project post-mortem, written into the lessons learned document and carried onto the next job by everyone who was there.

But then you go and look at where the margin actually went and a surprising amount of the time, it didn't go there at all.

The big blow tends to be the most visible thing that happened on a project and because it's so visible, it's usually the thing that ends up being managed best. It gets escalated, it gets a meeting, somebody writes it up and somebody puts a claim together whilst a recovery programme gets built around it. It's painful and it's expensive but it gets fought over and some proportion of it usually comes back. The events big enough to be remembered are the events big enough to be defended.

What quietly drains a project is everything sitting underneath that threshold.

The Jabs

A truck sits at the gate for two hours because the delivery landed outside its slot and there's nowhere to put it, a trade turns up and the preceding works aren't quite finished so half of them stand around until after lunch, an order arrives split across two deliveries rather than one, so the same crane, the same traffic management and the same supervision all get paid for twice, an item gets substituted for an equivalent that isn't quite equivalent and somebody loses a day and a half sorting it out or a delivery slot moves by a week, which gets absorbed by resequencing two trades, which then pushes a third trade into a fortnight where they're also committed on another job.

None of these are worth raising and that is the entire problem. Each one is individually too small to escalate, too ambiguous to pin on anyone cleanly and too easily solved by a competent site manager doing exactly what competent site managers do, which is deal with it and get on with the next thing. The person who absorbed it doesn't think of it as a loss at all. They think of it as Tuesday.

Multiply that across the length of a project though and then again across every project in a portfolio and it becomes a really surprising number. It's just a number that nobody has ever added up because at no point was anyone actually asked to.

Why There Are More of Them Right Now

This has always been true to some extent but the current market has made it considerably worse and when time and cost are of the essence we should be asking why.

Australian construction is running an enormous pipeline that's heavily weighted toward a small number of very large projects. Data centres in particular are absorbing mechanical and electrical packages at a scale the market simply hasn't had to accommodate before and that's alongside major infrastructure and defence work, all against a forecast pipeline heading toward a quarter of a trillion dollars by the end of the decade. Long lead items on imported and energy intensive lines are frequently sitting past twelve months. The specialist installation workforce is short by something in the order of 141,000 people with credible projections that the gap gets wider before it gets narrower.

When supply gets that tight, commercial priority follows long-term pipeline. That isn't a criticism and it isn't anyone behaving badly either. Having spent years on the manufacturing and supply side of building services, feeding equipment into commercial construction sites, it's a decision that makes complete sense from where it's being made. A customer representing consistent volume over several years is a genuinely different proposition to a single project order and when a factory can't serve everyone at once, the person defending the slot for the bigger relationship simply has the easier argument to make. Anyone who has sat in on those conversations understands the logic behind them, even when the outcome is frustrating.

The consequence though is that the volatility doesn't disappear anywhere, it just moves. It gets pushed down the chain to whoever has the least leverage and it arrives at those projects not as one dramatic failure but as a steady drip of small slippages, split deliveries, moved slots and substituted items, every one of which lands on a site team who absorb it and keep going.

That's the mechanism. Not one big blow, a hundred jabs.

What the Ones Who Cope Are Doing

The difference between the customers who came through periods like this reasonably well and the ones who really didn't was rarely about size and it was almost never about how forcefully they complained.

It came down to how far out they could commit.

The ones who fared best were simply a lot less reactive. They gave their dates further ahead, they confirmed properly rather than provisionally and they came back with the information that had been asked of them without needing to be chased for it three times. That made them easy to plan around and being easy to plan around is a genuine commercial asset in a constrained market, because a supplier allocating scarce capacity will always find it easier to protect a slot that's firm than one that's still soft. Predictability buys you something that urgency doesn't.

That's an uncomfortable finding for construction because reactivity is fairly close to the industry's default operating mode and a lot of the reason projects order late is that the information they need in order to order earlier is itself late. But it's the lever that actually exists and it's the one genuinely sitting within a project team's control.

The Problem With Acting On Any Of This

This is usually where it stalls.

Say a project manager reads all of the above, accepts it and decides to do something about it. The first question is a fairly obvious one. Which suppliers are actually causing this, how often and on what?

That question is close to impossible to answer on most projects and it isn't because the information never existed. It existed on the day. Somebody stood at that gate and knew the truck was two hours late, somebody knew the order came in two halves, somebody knew the trade stood around until after lunch and exactly why they did. Every single one of those facts was known by a person at the moment it was happening, then it got absorbed into the far more pressing job of getting the day finished and by the following week it had gone.

What survives instead is impression. A general view that a particular supplier has been a bit ordinary lately, which isn't the same thing as evidence at all and which tends to be shaped disproportionately by whatever went wrong most recently rather than by what has gone wrong most often. Walk into a supplier review with an impression and the conversation quickly becomes a discussion about one bad day, which the supplier can explain and almost certainly will.

So the hundred jabs never get addressed because they were never counted and things that don't get counted don't get raised. Meanwhile the one big blow, which was already the best managed event on the entire project, gets yet another meeting.

What Changes When You Can Count Them

The shift isn't a more forceful complaint, it's data.

When every delivery is recorded as it actually arrives, with the supplier, the promised time, the real arrival, the turnaround and whether it held up the work that was waiting on it, the individual jab stops being a fleeting annoyance and becomes a countable event. Once they're countable, they can be sorted.

Filter by supplier and the pattern everybody half sensed turns into a figure. Not "they've been a bit ordinary lately" but the number of occurrences, over what period, on which product lines and what it did to the programme each time. Sort it the other way and you find your genuinely strong performers, which matters every bit as much, because those are the ones worth giving longer notice to and worth protecting the relationship with the next time capacity gets tight.

That's a completely different supplier review. It moves the conversation off the most recent failure and onto the pattern and a frequency is very difficult to argue with. It's also close to the only real leverage available to a project that isn't the biggest customer in the room, because leverage in a constrained market comes from being able to demonstrate the relationship rather than just assert it.

None of this asks anyone on site to do more administration either, it asks that the information they already have, at the exact moment they already have it, gets captured somewhere it can be added up and used in a beneficial way later, when it’s needed.

The big blow will always get remembered. It's the jabs that need counting and SiteVector exists because they never were.