Running scaffolding jobs across several sites

A firm running one or two sites operates on shared knowledge. Everybody roughly knows where the gangs are, what is standing, and what is due to come down. It works well and it does not scale, and the failure is not gradual — it arrives somewhere around the fourth or fifth concurrent site.

What stops working first

Allocation is usually the first casualty: not knowing without a phone call whether a gang can be moved tomorrow. Close behind is knowing what is standing where, which matters for inspection coverage as much as for commercial reasons. Both depend on information that previously lived in one person's head and now needs to exist somewhere else.

Consistency between sites drifts

The other thing that goes is consistency. Each site develops its own practice around adaptations, handovers and paperwork, generally because a capable supervisor built something sensible locally. Each is defensible on its own and together they mean the firm cannot answer a question about its own operation without asking three people.

Worth deciding early which things must be identical across sites — how structures are identified, how adaptations are logged, what a handover requires — and which can safely stay local.

The knowledge that is still only on site

Whatever the structure, some knowledge stays site-specific: access arrangements, which client contact actually decides things, why a particular structure was built the way it was. It is invisible until the supervisor is off for a fortnight, and writing it down as it is used is far cheaper than reconstructing it in a hurry.

Common mistakes

• –Allocation information that requires a phone call to establish
• –No central view of what is standing across all sites
• –Each site developing its own paperwork and identification conventions
• –Site-specific knowledge undocumented because it is obvious to whoever holds it
• –Assuming a system that worked at two sites will work at six
• –No agreed minimum that must be consistent across sites

In practice: the fourth site

A firm that had run two or three sites comfortably for years took on a fourth and found that everything got harder at once — not proportionally, but noticeably. Allocation required phone calls, nobody could say confidently what was standing where, and two structures went a fortnight past their inspection because each site assumed the other had been covered.

Nothing about the firm had got worse. It had simply passed the point where shared knowledge works, and the systems that had been unnecessary at three sites became necessary at four.

What to standardise first

If only one thing gets standardised, make it how structures are identified. Everything else — inspections, adaptations, handovers, hire periods, variations — hangs off being able to refer to a specific structure unambiguously across sites and documents.

Firms that standardise reporting formats first and identification later usually end up redoing the reporting.

Worth knowing: travel is a real constraint

Allocation across sites is frequently planned as though gangs teleport. Travel time between sites, and the practical limits on moving a gang mid-day, shape what is actually possible — and a schedule that ignores it produces plans that fail quietly, with the slack absorbed by people starting earlier and finishing later.

• –Build travel into allocation rather than treating it as slack
• –Group sites geographically where the work allows
• –Check whether gangs are absorbing unplanned travel in their own time

Where this connects: travel is a real constraint

Allocation across sites is frequently planned as though gangs teleport. Travel time between sites, and the practical limits on moving a gang mid-day, shape what is actually possible — and a schedule that ignores it produces plans that fail quietly, with the slack absorbed by people starting earlier and finishing later than the plan assumed.

Which is worth knowing, because it is a cost that never appears anywhere. A gang absorbing forty minutes of unplanned travel a day is a real expense being paid in goodwill, and goodwill runs out without giving notice.

• –Build travel into allocation rather than treating it as slack
• –Group sites geographically where the work allows
• –Check whether gangs are absorbing travel in their own time
• –Standardise how structures are identified before anything else
• –Hold one register across all sites, not one per site
• –Write down site-specific knowledge as it is used

One practical test: ask a simple portfolio question

Ask somebody who is not the owner how many structures are standing across all sites, and which of them are past their quoted hire period. Time how long it takes to get a confident answer.

Under a minute means the information exists somewhere real. A phone call means it lives in one person's head. An afternoon means the firm has grown past its systems and has been absorbing the cost in somebody's time without pricing it. The question is deliberately mundane, which is the point — it is the kind a client or an insurer asks without warning.