Versailles was the most expensive reactive loop in history

Versailles was a small city wearing a house costume, and it employed enough people to prove it.

Yes, comparing your maintenance backlog to the fall of the French monarchy is a stretch. Stay with us anyway. Versailles is the cleanest example we have of a problem that still runs most FM operations today, and it comes with far better wigs.

At its height, the palace ran on thousands of servants, uncountable candles, and gardens full of fountains that drank more water than the estate could supply. The fix was theatre: courtiers walking ahead of Louis XIV would whistle to the fountain operators, who switched on each fountain just in time for the Sun King to admire it, then shut it off once he moved on. A palace so expensive to run that it could only afford to work when the right person was watching.

That is a symptom more than it is a lifestyle. And if you've ever seen a site suddenly get very clean the day before an audit, you already know the disease.

A palace that couldn't see its own costs

Nobody at Versailles knew what anything cost. The accounts were scattered across departments, keepers, and favours, and the numbers moved depending on who was asking. Historians still argue about how much of the French state's money the court actually consumed, and they argue precisely because the books were a mess. One reading of the 1789 figures puts royal household spending at roughly a fifteenth of state revenue. Others push it much higher. The useful part isn't the number itself but the very fact that no one at the time could give you one.

You can't manage a cost you can't see, and Versailles couldn't see its own. Most operations can't either. The real spend is hidden: emergency call-outs, after-hours labour, the contractor invoice nobody questioned, the asset quietly bleeding money because no one's tracking its lifecycle. The reported cost is the fountain the king walks past. The actual cost keeps running in the dark.

Every department has its own fiefdom

The kitchens, the stables, the gardens, or the wardrobe: each ran as a separate operation with its own staff, its own budget, and no particular reason to talk to the one next door. Coordination happened by rank and rumour. Work got duplicated in one wing and dropped entirely in another, and when something fell through the gap, it was always somebody else's gap.

Swap the wigs for spreadsheets and this is a Tuesday. The in-house team logs jobs one way, the contractor another, procurement lives somewhere else, and the facilities manager finds out a job was missed when the complaint arrives. Nobody built a shared place for the work to live, and siloed teams fail by nature.

Etiquette instead of workflow

Versailles ran on protocol so elaborate it needed to be memorised: who could sit in whose presence, who was permitted to hand the king his shirt at the morning dressing, and which door you were allowed to use. It looked like a process. It behaved like a process. It was, in fact, a theatre-worn process as a disguise.

When something actually broke, there was no workflow to follow. People improvised, deferred, or waited for someone senior to notice and issue a command. That works until it doesn't, and by then, you're managing a queue of things going wrong faster than anyone can respond.

Firefighting, occasionally literal

Fire was a genuine risk in a building lit by open flame and packed with fabric, and the court fought plenty of them. Most of the firefighting, though, was budgetary and political: react to this creditor, appease that faction, patch the crisis in front of you and hope the next one waits its turn. It rarely did. Every hour spent reacting to the last emergency was an hour not spent preventing the next, so the emergencies kept coming.

That's the whole machine. Four forces feeding each other: no shared data, siloed teams, no shared process, and constant firefighting. Each one makes the others worse. This is the part worth slowing down for, because it's the part everyone gets wrong.

Firefighting isn't a people problem

The instinct is to blame the firefighters for the firefighting. If only the staff were more disciplined. If only the technicians planned better. If only someone cared more.

They cared plenty at Versailles. The court was full of capable people working hard inside a system designed, without anyone deciding it, to keep them reacting. That's what the reactive loop is: not a staffing problem and not an attitude problem, but a system that produces urgent work faster than anyone can get ahead of it. It survives because teams, tools, and data keep working apart. Bad data leads to wrong decisions. Silos duplicate and drop work. No shared process means everyone improvises. And fragmentation lets a small failure cascade into an expensive one.

Your operation is a small Versailles. Fewer courtiers (we hope) but the very same loop. The urgent work eats the time you'd need to prevent urgent work, so more of the work turns urgent, and round it goes.

The good news the palace never got: the loop is a system, which means it can be redesigned. Connect people, processes, and data in one place and the forces stop reinforcing each other. Shared data means you can finally see the cost. Shared process means the response doesn't depend on who's on shift. Connected teams mean the gap between the kitchen and the stables closes. This is the work Infraspeak was built for: getting an entire operation onto the same page so it can plan rather than react.

Versailles never broke its loop. It kept reacting, department by department, crisis by crisis, until there was nothing left to react to. Break yours first.

Ask Louis XVI how the other option ends (not literally, it would be awkward to be seen speaking to his grave at Saint-Denis about strategic FM, but you do you).