SLA and escalation rules for operations teams

What an SLA and escalation rule actually are

A service level agreement (SLA), in an operations team, is a written target for how fast each kind of request is handled. It usually has two clocks: time to first response (someone owns it) and time to resolution (it is fixed). An escalation rule says what happens when a clock is about to run out or has run out: who is told, and what they are expected to do.

Without them, every request is urgent to the person asking and low priority to the person who is already busy. With them, lateness becomes a fact that the system can see, rather than an argument at the end of the shift.

This guide covers how to set the priorities, how to pick realistic targets, who to escalate to, and how to keep the whole thing honest. There is an example matrix in the middle; treat it as a starting point, not a standard.

Step 1: Define priorities in your own words

Priority should describe impact, not how loudly someone asked. Write one sentence per level and add real examples from your property.

• P1 Urgent. Risk to safety, property or a guest, resident or tenant who cannot use their space. Examples: water leak, no power to a unit, guest locked out, lift with people inside, gas smell, fire door wedged open.
• P2 High. Something an occupant is waiting on now, or that will become urgent if left. Examples: no hot water, air conditioning not cooling, toilet blocked, request for an item a guest needs before going out.
• P3 Normal. A real fault or request that does not stop anyone using the space today. Examples: a dripping tap, a noisy fan, a single light out in a room, an extra pillow.
• P4 Low. Cosmetic or planned work that can be batched. Examples: scuffed paint, a loose skirting board, a sign to replace.

Pin the definitions near the front desk, the reception dashboard or wherever requests are logged. The person logging the request sets the first priority; a supervisor can change it with a reason.

Step 2: Set first-response and resolution targets

Targets must be ones your team can actually meet on a normal busy day, with the staff you actually have on shift. Aspirational targets that are missed half the time teach everyone to ignore the clock.

A practical way to set them:

1. Pull two to four weeks of history, even if it is from a logbook or a chat group.
2. For each priority, find the time within which most jobs were already handled.
3. Set the target close to that, slightly tighter for P1 and P2.
4. Review after a month and tighten where the team is comfortably beating it.

Example SLA and escalation matrix (adapt before use)

The table below is an example only. It is written for a mid-sized property with a 24-hour duty engineer and a supervisor on every shift. Your numbers will differ with your size, layout, staffing and contracts. Use it as a template for the shape, not as a recommended set of values.

A few notes on the example:

• “Made safe” counts for P1. Shutting off a valve and putting down towels stops the damage. Fixing the pipe may need a part. Track both, but the first target is safety.
• Escalation names roles, not people. “Shift supervisor” resolves to whoever is on shift. Named people go on holiday.
• Two levels are enough. A third level for P1 safety issues (general manager or building manager) is reasonable. For anything else it trains senior people to ignore alerts.
• Housekeeping and guest requests often need their own rows. A towel request is not a repair. Many hotels run a separate, shorter target for guest amenity requests.

Step 3: Decide what escalation means

An escalation is only useful if the person receiving it knows what to do. Write it down.

• First escalation (supervisor): check whether the owner has accepted the task; if not, call them or reassign. If they have, find out what is blocking them: parts, access, another job. Decide whether to send help or reset expectations with the requester.
• Second escalation (department head): decide on extra resource: call in a second technician, call a contractor, move a room attendant from another floor, or offer the guest a room move.
• Safety escalation (manager on duty): the safety process takes over. The SLA becomes secondary.

Every escalation should leave a note on the task: who was told, what they decided. That note is what you learn from at the monthly review.

Step 4: Agree when the clock starts and stops

Most SLA arguments are about the clock, not the work. Settle these before go-live.

• Clock starts when the request is logged, not when someone reads it.
• First response stops when the owner accepts the task, not when they arrive.
• Resolution stops when the work is done, and a supervisor can reopen it if it was not.
• “Waiting for part” or “waiting for guest access” is a visible status with a reason, not a silent pause.
• Clock rules per priority (24/7 or business hours) are written down.
• Reprioritising a task is allowed, with a note, and the original priority is kept in the history.

Step 5: Review monthly, not daily

Daily, the escalation does the work. Monthly, look at patterns.

Adjust targets based on what you find. A target nobody can meet should be loosened and the underlying problem fixed. A target everyone beats by a wide margin can be tightened.

Telling requesters what to expect

An SLA is only half internal. The guest, resident or tenant who reported the problem has their own clock running, and it usually starts earlier and runs faster than yours. Closing that gap is mostly about communication.

• Acknowledge at first response. “Our technician is on the way” or “we have logged it for this afternoon” turns an anxious wait into a known one.
• Give a time, not a promise. “Within the hour” is better than “right away” that turns into ninety minutes.
• Call back when the target will be missed. A two-line update before the deadline passes prevents most complaints. Nobody minds a delay they were told about nearly as much as one they discovered.
• Confirm at close. A short check that the problem is really gone catches reopen cases early.

Write these touchpoints into the matrix as owner actions, usually for the front desk or the building’s community manager, so they happen on every shift and not only when someone remembers.

Common mistakes

• Everything is P1. Fix the definitions and let supervisors downgrade with a note.
• Escalating to a group chat. An escalation that goes to everyone goes to no one. Name a role.
• Hidden pauses. If waiting time disappears from the numbers, the numbers lie. Keep waiting visible.
• Using SLA breaches to punish. People will start closing tasks early or not logging them at all. Use breaches to find the system problem first.
• Copying a contract SLA into daily operations unchanged. Contract definitions are written for lawyers. Your staff need plain ones.

Where TaskLoop fits

TaskLoop supports the core of this: requests are logged with a priority and a deadline, response and resolution timers are configurable per category and priority, staff accept and complete them on the mobile app with timestamps, overdue tasks escalate to supervisors, and reports show response and completion times by department and by person. The matrix itself is a decision for your team; the tool makes it visible. See escalations and SLA for how escalation works.