Activities

Why progress claims stop being true

The ninety-per-cent syndrome is one of the best-documented patterns in project management, and it is not a story about laziness. It is a story about measurement.

When percent-complete is subjective, it behaves like an opinion — and opinions are optimistic, in a specific and predictable way. The first eighty per cent of an activity is visible and easy to claim. The last twenty is fiddly, unglamorous and hard to see, so it gets rounded up. Every week. And the activity sits at ninety, then ninety-two, then ninety-two again, while the remaining work quietly turns out to be a third of the job.

The fix is not to demand more honesty. It is to remove the opportunity for optimism: rules-based percent-complete, decided BEFORE the work starts, measured against physical milestones or installed quantities against the bill of quantities. Fifty cubic metres poured out of two hundred is forty per cent, and it is forty per cent whether or not anyone is pleased about it.

And note what the rule does to the conversation. Decided in advance, it is a method. Decided afterwards, it will be decided by whoever needs the number — which is the definition of a broken control.

What activity data is actually worth

• Earned against burned is the productivity lens that tells you first. Earned hours come from installed quantities multiplied by your norms. Burned hours come from the timecards. The ratio between them is the AACE-recognised productivity measure, and its virtue is timing: it moves BEFORE the cost report, because the cost report is waiting on invoices and the invoices are waiting on a month-end. A crew burning faster than it earns is losing money today. You can either know that today, or you can find out in six weeks when the option to do anything about it has expired.

• Contemporaneous progress records are claim evidence — and they cannot be made later. Delay and disruption analysis needs period-by-period production data: what was achieved, where, with how many people, week by week. The measured-mile method compares an unimpacted period against an impacted one, and it is the most persuasive disruption analysis there is precisely because it uses the project against itself. But it requires the data to have existed at the time. Reconstructed progress fails in exactly the way a reconstructed site diary fails — a tribunal can tell, and it will say so.

• Progress is what payment is certified against. The application says the work is done; the certificate agrees or does not. Measured progress is the evidence in between. And inflated progress is not a victory — it is a loan, taken at an unfavourable rate. The clawback arrives later, usually at the worst moment, and usually in front of the people whose trust you needed.

• Track by location, or the disruption stays invisible. An aggregate percentage hides everything that matters. Track by zone, level or grid and the picture changes: three trades stacked in one area fighting for the same front, while an adjacent zone sits empty because its predecessor never finished. Trade stacking and front availability are the mechanics of disruption, and they are simply not visible in a single project-level number — which is why that number is so often the last thing to move.

What it costs to guess

Progress is claimed subjectively, so the forecast is built on it, so the forecast is wrong in the same direction every month — and everyone is surprised at the same moment, together, late.

Productivity is never measured against earned hours, so a losing crew looks like a busy one until the cost report catches up. And there is no contemporaneous production record, so when the disruption claim finally has to be made, the strongest analysis available — the project’s own unimpacted period, measured against its impacted one — cannot be run at all.

Meanwhile the payment application is certified against a number that came from a conversation. Which is fine, until it is not.

How Zepth tracks activities

Activities carry their quantities, their locations, their photos and their manpower — so progress is a measurement rather than an assertion, and it is attached to the evidence that proves it.

Earned hours compute from installed quantities; burned hours arrive from the timecards. The ratio is a dashboard rather than a quarterly exercise. And the same record feeds the daily report, the payment application and the forecast — which is the point: one measurement, not three arguments.