SLAs are supposed to set expectations between a service provider and a client. Most FM SLAs fail because they are designed like contracts instead of like operational targets, aggressive on paper, ignored in practice. Here is how to design an SLA matrix that actually drives behaviour.
The two SLA metrics that matter
Every SLA tracks two time intervals:
Response Time
From request logged until someone acknowledges or starts work.
Resolution Time
From work order started until job completed.
These are separate measurements because fast response doesn't guarantee fast resolution, and vice versa.
Priority levels (the SLA driver)
SLA targets are driven by priority, which is driven by impact:
| Priority | Description | Examples |
|---|---|---|
| P1: Emergency | Safety or major operations impact | Fire alarm, power outage, water leak |
| P2: Urgent | Major disruption to operations | HVAC failure in occupied area |
| P3: Routine | Minimal operational impact | Minor repairs, non-critical faults |
| P4: Planned | Scheduled activities | Preventive maintenance tasks |
Realistic SLA target matrix
These targets work in most commercial FM environments:
| Priority | Response | Resolution |
|---|---|---|
| P1 | 15 minutes | 2 hours |
| P2 | 1 hour | 6 hours |
| P3 | 4 hours | 24 hours |
| P4 | As scheduled | As scheduled |
Industrial, critical-infrastructure, or healthcare environments may need tighter SLAs. Small SME offices may need looser ones. These are the starting point.
Which work order types get SLAs
Only RM (Reactive Maintenance) should have SLAs
Reactive work has an external requester and a user waiting. That's the definition of an SLA-able ticket. PPM, CM, INS, and planned work don't need SLAs, they have schedules, KPIs, or compliance deadlines instead. Applying SLAs to PMs creates perverse incentives.
SLA pause conditions
SLA clocks need to pause for things outside the team's control:
- Waiting for material, spare not in stock
- Waiting for vendor/service, contractor hasn't arrived
- Waiting for customer access, user not in room, lab locked
- Waiting for PTW, permit approval pending
- Waiting for approval, scope change needs sign-off
All pauses must be logged and auditable. If pauses aren't tracked cleanly, vendors dispute every breach.
Status indicators
Dashboards should show SLA status in three colours:
- Green, within SLA (<80% of time consumed)
- Amber, approaching breach (80-100%)
- Red, breached (>100%)
Common SLA design pitfalls
- Aspirational targets. 99.9% resolution within 2 hours, sounds great, impossible to achieve, destroys team morale.
- No pause mechanism. SLAs breach because of stock-outs nobody controls.
- Applying SLAs to preventive work. PMs need compliance tracking, not SLAs.
- Too many priority levels. 7 priorities means nobody actually uses P4-P7 consistently.
- No exclusion windows. 24/7 SLA for a Monday-Friday operation creates phantom breaches every weekend.
Conclusion
Good SLA design is about honesty. Honest about what's achievable, honest about which work types need SLAs, honest about pause conditions. An aggressive SLA that nobody hits is worse than a realistic SLA that drives consistent performance. Start with 4 priorities and the target matrix above, adjust based on your operational reality, and track results honestly.
Written by Muhammad Abbas
Enterprise integration specialist. See also Work Order Types in CMMS.
