The short answer

Site teams do not run out of hours. They run out of uninterrupted ones. A community manager's day is structurally fragmented by walk-ins, calls, vendor arrivals and escalations, which means capacity is lost to recovery time rather than to task volume. The way out is not working faster. It is sorting every task by two questions, whether it must happen at the property and whether it must happen now, then removing everything that fails both tests.

Most work that lands on a community manager does not need to be done on site, and most of it does not need to be done immediately. Almost none of it is sorted that way.

Why doesn't adding software make site teams less busy?

Because software was added to the tasks, and the tasks were never the constraint.

Gloria Mark, a professor of informatics at the University of California, Irvine, has spent two decades observing how people actually work, initially by following them with clipboards and later with tracking software. Her research on interrupted work, discussed in this Fast Company interview, produced one of the most cited findings in the field: after an interruption, workers took an average of more than twenty minutes to return to the original task at the same depth of focus. Her observational work also found that people rarely go straight back. They typically complete two other tasks first.

Sophie Leroy at the University of Washington named the related mechanism attention residue: after switching, part of your cognitive capacity remains allocated to the previous task, degrading performance on the new one.

Now apply that to a leasing office. A community manager is working through renewal offers. A resident walks in about a parking space. The phone rings. A vendor arrives needing the gate code. A regional manager wants last month's variance explained. Each is small. Each is legitimate. And each resets a recovery period that was already running.

The shape of the day matters more than its length. When interruptions arrive faster than the work can be reoriented, a substantial share of the day goes to returning to things rather than finishing them. That is not a claim about any individual's capacity. It is a claim about structure, and it explains why the work feels enormous while the output looks modest, and why the honest answer to "what did you get done today" is so often difficult to give.

It also explains something operators find genuinely puzzling. You buy software, staff report it is good, and nothing gets easier. Of course it does not. A tool that improves how a task is completed does nothing about how often that task is started, stopped and restarted. Worse, most tools add a notification stream, which is a new interruption source pointed at the same person.

The lever is not the task. It is whether the task reaches that person, at that moment, at all.

The Presence Matrix

Every piece of work landing on a site team can be sorted by two questions. Does it require someone physically at the property? Does it require action right now?

Only Q1 genuinely requires this person, in this place, at this moment. It is also the smallest quadrant by volume and the one most often crowded out.

Q2 requires the person and the place but not the moment, so it can be clustered into protected blocks.
Q3 requires immediacy but not presence, which makes it the natural candidate for a centralised queue.
Q4 requires neither, which means every minute of it spent on site is a choice rather than a necessity.

Most "doing more with less" advice tries to make the community manager faster across all four quadrants. The better move is to make three of them somebody else's problem, or nobody's.

Quadrant 1. Here and now: what actually requires you on site?

Less than the day suggests, and it is worth being strict about the definition.

Genuine Q1 work has a consequence that worsens materially with delay and cannot be discharged remotely. A flood. A lockout at 9pm. A fire panel alarm. A resident in distress. A police or code enforcement attendance. Anything where a statutory clock has started, which we covered in detail in where maintenance automation should stop.

What is not Q1, despite feeling like it: a resident who is upset about something that happened last week, a vendor who arrived without an appointment, a corporate request for a number, or a prospect who walked in during a scheduled inspection. These are urgent to the person raising them, which is a different thing from urgent.

The practical discipline is to name Q1 explicitly and in writing, then treat everything outside that list as interruptible by default rather than the reverse. Site teams generally operate on the opposite assumption, treating everything as immediate unless proven otherwise, which is how the smallest quadrant gets buried under the other three.

Quadrant 2. Here but not now: what can you batch?

The work that has to happen at the property but does not have to happen this minute is the largest recoverable block in the week.

Unit inspections, vendor escorts, common area walks, turn sign-offs, punch lists, amenity checks. Each requires presence. None requires a specific moment. Yet in most communities they are performed reactively, as gaps appear, which means they are constantly interleaved with everything else.

Clustering them changes the economics. Two protected three-hour blocks a week, with the office covered and phones routed, converts a stream of interruptions into two scheduled activities. It also improves the work itself, because inspections done between two phone calls are not inspections, they are glances.

The prerequisite is scheduling infrastructure rather than willpower. Vendor arrival windows, inspection calendars and turn sequencing need to live somewhere the whole team can see, which is what maintenance planning and scheduling is for, and what makes a community manager portal more than a dashboard.

Quadrant 3. Now but not here: what belongs in a centralised queue?

Anything where speed matters and location does not.

Lead response is the clearest case. Response speed is the single most controllable conversion variable in leasing, as we set out in nine ways to improve occupancy rates, and it is also the one a site team is structurally worst placed to deliver, because the inbound lead arrives precisely when the community manager is doing something else. A centralised leasing queue or an AI first-touch does not make the site team lazier. It removes an interruption the site team could never have serviced well anyway.

The same logic covers after-hours calls, emergency triage and escalation intake. All demand immediacy. None requires the person standing in that building.

This is where the industry has been moving. Propmodo has documented large owners including AvalonBay and Greystar experimenting with consolidated leasing structures. Executives at some of those operators report lower turnover in leasing roles under centralised models, which they attribute to steadier income than commission-dependent site positions offer. That is practitioner experience rather than measured evidence, but it points at something structural: a role that is hard to staff on site may be easier to staff off it.

Quadrant 4. Neither: what should leave the property entirely?

Administration, and this is where the biggest gains sit.

Invoice coding, application processing, renewal document preparation, collections follow-up, compliance filing, reporting. None of it requires presence. None of it requires immediacy. All of it lands on site teams because that is where it has always landed.

The National Apartment Association's reporting makes the point that this, rather than leasing, is the real centralisation story. Covering the Cardinal Group's approach, the NAA quotes Jacob Kosior on the principle that separates programmes that work from those that do not: "It is the tasks and the processes that we centralize, not the people."

That distinction matters more than it sounds. Centralising people removes a body from the property and hopes service holds. Centralising tasks removes work from the property and leaves the relationships in place. The first is a headcount decision. The second is an operating model decision, and only one of them survives contact with a resident who wants to talk to someone who knows the building.

Where full centralisation is not available, automation covers much of the same ground, which is what workflow automation is genuinely good at. Q4 work is repetitive, rule-governed and low-consequence per instance, which is exactly the profile automation handles well.

Is centralisation actually the answer?

Partly, and it fails in specific ways worth knowing before anyone proposes it.

• It can degrade service visibly.
  Propmodo's reporting notes that removing on-site staff can reduce service quality in ways that surface as poor reviews, weaker retention and eventual pricing pressure, alongside reduced oversight of the physical asset. Review content now carries disproportionate weight with prospects, as we covered in how to measure resident experience, so service degradation reaches the leasing funnel faster than it reaches a satisfaction survey.

• It requires conditions most operators do not have.
  NAA's analysis identifies two determinants of whether centralisation is viable: control of the operating environment and submarket density. Public REITs typically have both, with standardised platforms and properties close enough to share staff. Most operators have neither, which is why models copied from large REITs frequently fail in mid-sized portfolios.

• Done badly, it just relocates the interruption.
  Moving a queue to a call centre without changing the handoffs produces the same fragmented work in a cheaper location, plus a new coordination cost between the centre and the site. If the centralised team has to call the community manager to answer half its queries, nothing was centralised. A second interruption source was created.

The honest position is that centralisation is a portfolio-level decision with real trade-offs, and a community manager reading this may not be in a position to make it. Which raises the more useful question.

What can a community manager do without waiting for corporate?

Four things, none of which require a budget or a restructure.

• Sort your recurring work into the four quadrants once.
  Take an hour, list everything you did last week, and place each item. Most people find Q4 is larger than they expected and Q1 smaller. That list is the entire argument, and it is also the document to take to a regional manager.

• Protect two blocks a week for Q2.
  Phones routed, door signed, work clustered. Two blocks is defensible in a way that "I need focus time" is not, because it maps to specific inspections and walks with dates attached.

• Change your defaults, not your speed.
  Route inbound requests through the tenant portal and service request management so they queue rather than interrupt. A request that arrives as a ticket can be batched. The same request arriving as a phone call cannot.

• Measure interruptions for one week.
  A tally mark for every unplanned interruption, with a one-word category. It takes seconds per entry and produces something no time-tracking exercise will: evidence about the shape of the day rather than the length of it. That tally is a far stronger case for centralised support than any general statement about being busy.

Is the 1:100 staffing ratio still valid?

It was never an operating model. It was a cost formula.

One employee per hundred units has served as the industry's staffing heuristic for decades, and it has come under sustained challenge as amenity programmes expanded, resident expectations rose and compliance obligations multiplied. The ratio says nothing about what those hundred units require: whether the property is one building or nine, whether the resident base is students or seniors, whether there is a commercial component, or how much Q4 work corporate pushes down.

The more useful measure is not units per employee. It is uninterrupted hours per week available for Q1 and Q2 work. Two communities at identical ratios can differ enormously on that number, and it is the one that predicts whether inspections happen, whether renewals get called, and whether the manager is still there in a year. The hiring and retention guide covers the burnout side; fragmentation is the mechanism underneath it.

Which quadrant is eating your week?

Frequently asked questions

Q1. Why do property management teams feel busier after adding software?
Because tools improve how tasks are completed without changing how often they are interrupted, and most add a notification stream of their own. Capacity is lost to recovery time between switches, not to task volume.

Q2. What work should a community manager stop doing on site?
Anything that requires neither presence nor immediacy: invoice coding, application processing, renewal paperwork, reporting and collections follow-up. It lands on site teams by convention rather than necessity.

Q3. Does centralisation improve site team capacity?
It can, when tasks and processes are centralised rather than people. It fails when a queue is relocated without redesigning handoffs, which creates a second interruption source instead of removing the first.

Q4. What are the risks of removing on-site staff?
Trade reporting notes reduced service quality showing up as weaker reviews and retention, pricing pressure, and less oversight of the physical asset. Review content reaches prospects faster than service problems reach a survey.

Q5. Is one employee per 100 units still a valid staffing ratio?
It is a cost formula, not an operating model. It ignores building count, resident type, commercial components and how much administrative work corporate pushes down to the property.

Q6. How long does it take to recover from an interruption?
Research led by Gloria Mark at UC Irvine found workers often take more than twenty minutes to fully return to an interrupted task, though recovery varies with the task and the nature of the interruption.

Q7. What can a community manager change without corporate approval?
Sort recurring work into the four quadrants, protect two scheduled blocks a week, route inbound requests into a queue rather than a phone call, and tally interruptions for one week as evidence.

Q8. Which tasks should never leave the property?
Emergencies, lockouts, incident response, anything with a safety or habitability clock, and situations where a resident needs a person who knows the building. That list is shorter than most days imply.

The real job of doing more with less

The phrase usually means working faster with the same tools and fewer people, which is why it lands badly with the people expected to do it. Nobody at a property is running at half speed.

The constraint is structural. Site work is interrupt-driven, interruptions carry a recovery cost far larger than their duration, and most of what interrupts a community manager did not need to happen there or then. Capacity is not created by pressing harder on a fragmented day. It is created by changing what reaches the property in the first place.

The operators who get this right are not the ones whose teams work hardest. They are the ones who worked out which quadrant each task belonged in, and stopped sending three of them to the site.

RIOO gives site teams a community manager portal that queues requests rather than interrupting, on the same NetSuite data layer as leasing, maintenance and accounting, so administrative work can be routed away from the property without anyone losing visibility of it. See how it works.