Case 1 · Service operations & SLA¶
Data used on this page
Service desk tickets - synthetic. Details on the data sources page. With synthetic data the findings illustrate the method; they are not claims about a real organisation.
Context¶
A service desk with four queues handled 10,637 tickets in 2024. Each priority carries a resolution promise (SLA): P1 8 h, P2 24 h, P3 72 h, P4 120 h. Agents pick the highest priority first, then the oldest ticket. On 01 September 2024 a product rollout started generating more field visits.
Question¶
Are we keeping the resolution promise - and if not, where, since when and why?
KPIs · 2024Q4¶
SLA compliance is the outcome the customer feels; it is lagging (measured when a ticket closes). Aged backlog is its leading indicator: today's old tickets are next month's breaches. Reopen rate is the counterweight that stops speed being bought with quality.
Finding¶
Show the data behind this chart
| month | value | status |
|---|---|---|
| 2024-01-01 | 98% | green |
| 2024-02-01 | 96% | green |
| 2024-03-01 | 97% | green |
| 2024-04-01 | 97% | green |
| 2024-05-01 | 98% | green |
| 2024-06-01 | 96% | green |
| 2024-07-01 | 96% | green |
| 2024-08-01 | 96% | green |
| 2024-09-01 | 94% | green |
| 2024-10-01 | 91% | green |
| 2024-11-01 | 88% | amber |
| 2024-12-01 | 88% | amber |
Show the data behind this chart
| bucket | queue | 0-2 days | 3-7 days | 8-14 days | 15-30 days | > 30 days |
|---|---|---|---|---|---|---|
| Field Support | 19 | 17 | 22 | 39 | 113 | |
| Technical | 7 | 2 | 0 | 0 | 0 | |
| Accounts | 5 | 0 | 0 | 0 | 0 | |
| Billing | 3 | 0 | 0 | 0 | 0 |
Reading it together. Since the rollout, Field Support receives about 64 tickets a week instead of 44 (+45%) with the same team. Priority rules protect P1/P2, so the wait piles up in P3/P4 and in the backlog. The fleet median resolution time (11.9 h) barely moves - the problem lives in the tail and in one queue. Reopenings in the other queues stayed at 3.2%: the quality drop is local too.
Decision¶
- Close the capacity gap in Field Support - the gap is about 20 tickets a week. Options: reassign trained agents from the queues that are well above target, or overtime for a burn-down period. Size it with the gap, not with the SLA percentage.
- Burn down the aged backlog first - 174 tickets are already older than 7 days. A daily stand-up on the oldest tickets is cheaper than the breaches they will cause.
- Do not reward closure speed alone - pair every resolution-time target with the reopen rate, so agents are not pushed to close tickets that come back.
How we keep measuring¶
| KPI | Owner | Cadence | Target | Trigger for action |
|---|---|---|---|---|
| Aged backlog (> 7 days) | Queue Team Lead | Weekly, per queue (leading) | ≤ 10.0% | > 20% in any queue for 2 weeks |
| SLA compliance | Service Desk Manager | Monthly, per queue and priority | ≥ 90.0% | Below 85% in any queue |
| Resolution time (median) | Service Desk Manager | Weekly, median and P90 | ≤ 24.0 h | P90 above the SLA of the priority |
| Reopen rate | Quality Lead (Service) | Weekly, per queue | ≤ 5.0% | Above 8% - review closed tickets |
| First contact resolution (FCR) | Service Desk Manager | Monthly, per channel | ≥ 70.0% | Below 60% in a channel |
The aged-backlog rule fires weeks before SLA compliance drops, which is why it is reviewed weekly.