Triage is the skill that separates good property management from liability
Answering the phone is not the hard part of property management. Deciding, in the first ninety seconds of a call, whether "there's water in my bathroom" means a spilled cup or a supply line failure — that is the hard part. Get it wrong in one direction and you dispatch an emergency plumber for a mopped floor. Get it wrong in the other and a slow leak becomes a multi-unit claim and a habitability complaint.
Triage is a system, not an instinct. This guide lays out the framework property managers use to classify every maintenance call correctly, escalate the genuine emergencies instantly, and queue everything else without waking anyone at 3 AM. It is written so you can hand the framework directly to an AI phone agent or a human team and get consistent results.
The habitability duty you are actually triaging against
Most US states do not require a property manager to answer the phone 24/7. Nearly all of them require a *timely response to habitability emergencies*. That legal duty is the reason triage exists, and it defines what "emergency" means in a way that is stricter than tenant urgency. The U.S. Department of Housing and Urban Development maintains state-by-state tenant-rights resources worth bookmarking, and the National Apartment Association publishes operational benchmarks on maintenance response.
| Condition | Typical legal expectation | Triage class |
|---|---|---|
| No heat in freezing conditions | Repair within ~24 hours | Emergency |
| Major water leak / flooding | Immediate mitigation | Emergency |
| Gas odor | Immediate response + safety instruction | Emergency |
| No running water | Repair within ~24–48 hours | Emergency/Urgent |
| Sewage backup | Immediate mitigation | Emergency |
| No electricity / electrical hazard | Immediate response | Emergency |
| Broken exterior lock/door (security) | Reasonable, prompt response | Urgent |
| No hot water | Prompt, not immediate | Urgent |
The exact standards vary by state and lease, so this table is a starting framework rather than legal advice — but it shows why triage is anchored to habitability, not to how loudly a caller insists.
The three-tier classification
Every maintenance call sorts into exactly one of three tiers. The tier determines the response protocol, and the protocol is the same every time regardless of who answers.
Tier 1 — Emergency (respond immediately)
Active water intrusion, gas odor, no heat in freezing weather, sewage backup, electrical hazard, fire damage, security breach. These threaten safety, health, or the building itself.
Protocol: Capture unit and nature of the hazard, give any immediate safety instruction, escalate to the on-call manager or dispatch a contractor within minutes.
Tier 2 — Urgent (respond same day)
No hot water, a broken lock, a failed appliance affecting daily life, a single-bathroom toilet out of service, pest infestation. Serious, but not a same-minute safety threat.
Protocol: Capture full details, create a work order, schedule a vendor same-day or next morning, notify the manager in the daily summary.
Tier 3 — Routine (respond within 24–48 hours)
Dripping faucet, running toilet, minor appliance quirk, paint touch-up, common-area bulb, cosmetic issues.
Protocol: Capture details, create a work order, schedule for the next available slot, no after-hours escalation.
Source: Estimated from multi-family property management maintenance-call data.
The critical insight: only about 12% of maintenance calls are true emergencies. A triage system that correctly identifies that 12% lets you escalate them hard while letting the other 88% wait for business hours — which is the entire mechanism behind protecting your evenings without exposing yourself to liability.
The emergency decision tree
An emergency is identified by asking, not by the caller's tone. This is the decision tree to configure into your agent or hand to your team. Each "yes" ends the tree with an escalation.
| Question | If yes |
|---|---|
| Is water actively leaking or flooding right now? | Emergency escalation + "shut off the supply valve if you can safely reach it" |
| Do you smell gas or see/smell smoke? | "Leave the unit now, call 911, then the gas company" + immediate manager alert |
| Is the heat completely out and is it near or below freezing? | Emergency escalation + HVAC dispatch |
| Is there no running water at all? | Emergency/urgent escalation depending on lease and duration |
| Is there sewage backing up into the unit? | Emergency escalation + mitigation dispatch |
| Is there an electrical burning smell, sparking, or exposed live wiring? | "Turn off the breaker if safe" + electrician dispatch |
| Are you locked out with a safety concern (child inside, medical, unsafe area)? | Urgent escalation + locksmith |
If none of the triggers fire, the call is Tier 2 or Tier 3, and it is captured and queued — never escalated.
Structured intake: capture it once, correctly
The second-biggest time sink in maintenance (after triage errors) is the back-and-forth to understand a vague report. "Something's wrong with my sink" becomes three callbacks. A structured intake captures everything on the first call.
| Field | Why it matters |
|---|---|
| Property + unit number | Routes the work order to the right building |
| Tenant name + callback number | Matches the lease, enables follow-up |
| Issue category | Routes to the correct vendor (plumbing, HVAC, electrical) |
| Specific description | Prevents the diagnostic callback |
| Location within unit | Guides the vendor on arrival |
| When it started | Assesses damage risk and urgency |
| Access permission + pets | Lets you schedule entry without a second call |
| Tenant-preferred window | Reduces failed visits |
A configured AI agent collects all of these in a natural conversation and produces a complete work order before the call ends, instead of a voicemail that says "call me back." The mechanics of that workflow are covered in managing tenant enquiries and maintenance calls.
Escalation configuration: who gets woken, and how
Escalation only works if the right person is notified through the right channel with enough context to act. Configure it explicitly.
| Setting | Recommended configuration |
|---|---|
| Emergency recipient | On-call manager mobile via SMS + call |
| Escalation content | Property, unit, tenant, hazard, time, safety instruction given |
| Consolidation rule | Group calls from the same building-wide incident into one alert |
| Fallback recipient | Secondary manager if primary doesn't acknowledge in X minutes |
| Vendor auto-dispatch | Optional direct dispatch for defined emergency categories |
| Routine handling | Queue silently for the morning summary — no after-hours alert |
The consolidation rule matters more than any other in property management: a burst riser that generates twelve resident calls should produce one escalation, not twelve. Systems that can answer all twelve simultaneously and roll them into a single incident alert are the ones that keep an on-call manager sane.
Measuring triage quality
Triage is tunable. Review these metrics monthly and adjust the decision tree and thresholds.
| Metric | What it tells you | Fix if poor |
|---|---|---|
| False-emergency rate | Non-emergencies escalated after hours | Tighten the decision tree wording |
| Missed-emergency rate | True emergencies queued instead of escalated | Broaden triggers; audit transcripts |
| First-call capture rate | Work orders complete without a callback | Add missing intake fields |
| Time-to-dispatch (emergencies) | Speed of the escalation-to-vendor path | Pre-authorize vendor auto-dispatch |
| Repeat-call rate per issue | Issues not resolved on first dispatch | Improve vendor briefing detail |
For portfolios where after-hours volume is the core pain, pair this framework with the coverage analysis in do property management companies need 24/7 call answering.
Keep Reading
- Best Property Management Answering Services — which services can actually triage
- Property Management Answering Service Cost — what 24/7 triage costs by door count
- Best Way to Manage Tenant Enquiries and Maintenance Calls — the end-to-end work-order workflow
- How Property Managers Handle Tenant Calls Without Burning Out — protecting your time
- Property management industry overview — configuring triage in OnCallClerk
Frequently Asked Questions
Q: How do I stop tenants from labeling everything an emergency?
You triage on facts, not framing. The decision tree asks objective questions — is water actively leaking, is there no heat near freezing — rather than accepting the caller's urgency. A caller who insists a running toilet is an emergency is handled politely, classified as routine, and given a realistic timeline. A consistent, objective tree removes the pressure to over-escalate.
Q: Can an AI phone agent be trusted to triage emergencies correctly?
When configured with a clear decision tree, yes — and it does so more consistently than a tired human at 3 AM, because it runs the exact same checklist every time. The key is auditing transcripts for the first few weeks and tuning the trigger wording. Track your missed-emergency and false-emergency rates and adjust.
Q: What counts as a legal emergency versus just urgent?
It varies by state and lease, but habitability conditions — no heat in freezing weather, major leaks, gas, sewage, no running water, electrical hazards — are the core emergency set nearly everywhere. Non-habitability issues like a dripping faucet or cosmetic damage are routine. Confirm your specific jurisdiction's standards; this framework is a starting point, not legal advice.
Q: How should building-wide incidents be handled?
Consolidate. When many residents in one building call about the same failure, the system should answer all of them, reassure each, log the affected units, and send the manager a single consolidated emergency alert rather than a flood of separate notifications. This is a defining requirement for multi-family portfolios.
Q: What information does my maintenance team actually need?
A complete work order: property, unit, tenant and callback number, issue category, specific description, location in the unit, when it started, access permission, pets, and preferred window. Capturing all of it on the first call is what eliminates the diagnostic callbacks that waste the most time.