There’s a decision every growing commercial operation eventually faces, usually after a bad quarter of missed PMs and escalating vendor chaos: the coordination is breaking, so do we hire another coordinator — or change how coordination works?

It looks like a budget question. It’s actually an architecture question, and getting it wrong is expensive in a way that compounds: the wrong answer doesn’t just cost a salary, it locks the operation into a model with a structural ceiling.

Most operations that feel understaffed are actually misallocated: skilled people spending the majority of their week on coordination that a system should run, leaving no capacity for the judgment work that justifies their salary.

Model A: Scale the people

The headcount model is the default for a reason. It’s fast, it’s familiar, and for a specific range of operations, it genuinely works. One skilled coordinator can hold roughly a handful of sites in their head — the vendors, the schedules, the quirks of each building. When the operation sits comfortably inside that range, hiring is the right call: the new person absorbs the overflow, relationships deepen, and judgment improves with tenure.

Where it stops working is precisely where most operations apply it: past the range. Coordination load doesn’t grow linearly with sites — it grows with the interactions between sites, vendors, assets, and deadlines. Double the sites and you roughly quadruple the threads someone has to hold. The new hire doesn’t halve anyone’s load; they add a communication layer, because now two people must stay synchronized on which vendor was called and which PM was rescheduled. And the labor market makes the model fragile on top of slow: IFMA’s workforce research has tracked persistent vacancy and an aging pipeline — with roughly 40% of FM managers over 55 per FacilitiesNet — meaning the people you’d hire increasingly don’t exist to be hired.

Model B: Scale the infrastructure

The automation model changes what coordination is. Work orders create themselves from PM calendars, sensor signals, and verified requests. Dispatch routes to pre-qualified vendors by specialty, score, and availability. Status, escalation, and follow-up run on rules instead of memory. Payment releases on verified completion. In multi-site facility management, this is the difference between a person juggling twelve sites and a system coordinating twelve sites while a person supervises the exceptions.

Where this model fails is equally specific: when it’s installed on top of chaos. Automation executes the process it’s given. If asset data is wrong, PM schedules are fiction, and vendor records live in someone’s phone, automation accelerates the mess — a pattern consistent with McKinsey’s finding that only 5% of AI programs in commercial real estate achieved their stated objectives. The failure point is almost never the technology. It’s the data underneath it.

MODEL A Scale the People

Fast and familiar, but adds communication layers and hits a structural ceiling rapidly as site interactions quadruple.

MODEL B Scale the Infrastructure

Systematizes coordination so rules drive dispatch and follow-up, but requires clean asset and vendor data to succeed.

THE REALITY Misallocation vs. Understaffing

Adding a person buys time but not scale if the work breaking the operation is coordination rather than judgment.

The variable that decides

So the honest framework isn’t “headcount versus automation.” It’s a question about the nature of the work breaking your operation:

  • If the breaking work is judgment — repair-versus-replace calls, vendor negotiations, capital planning — you need a person, and no system replaces that.
  • If the breaking work is coordination — creating, dispatching, chasing, reconciling, documenting — adding a person buys time but not scale, because you’re paying a salary to be a router. That work has a structural ceiling in human form and effectively none in automated form.

Most operations that feel understaffed are actually misallocated: skilled people spending the majority of their week on coordination that a system should run, leaving no capacity for the judgment work that justifies their salary. The interviews behind our research said it plainly — operators described their ratio of coordinating versus actually managing, and the ratio was the complaint.

What it looks like in a real operation

A regional operations manager with nine sites stops being the person who finds the plumber. The system finds the plumber. She becomes the person who notices, from the dashboard, that site 4’s plumbing spend has doubled year-over-year and decides it’s time for a repipe conversation — the decision a single point of failure operation never gets to, because its people are too busy routing. That reallocation of human attention is the product Sweven FM actually delivers; the software is just how.

The Architecture Question

The question to ask before the next requisition: is the work that’s drowning your team work that requires their judgment — or work that requires their memory and their phone?

Sources: