SLA Time-to-Done Breach Rate
The proportion of PSOM tickets that missed the “Time to done” SLA, based on the latest snapshot.
This metric reflects how well the team is resolving tickets within the agreed timeframe. It measures the whole ticket, including time spent escalated to another team. Only tickets where the SLA clock started are included; tickets with no clock are excluded.
Calculation
From jira.psom_sla, filter to the most recent date partition, to sla = 'time_to_done', and to rows where breached is not null (i.e., the clock started). Divide the count of rows where breached is true by the total count of those rows.
SELECT
COUNT(*) AS total_sla_tickets,
SUM(CASE WHEN breached THEN 1 ELSE 0 END) AS breached_count,
CAST(SUM(CASE WHEN breached THEN 1 ELSE 0 END) AS DOUBLE)
/ NULLIF(COUNT(*), 0) AS breach_rate
FROM jira.psom_sla
WHERE date = (SELECT MAX(date) FROM jira.psom_sla)
AND sla = 'time_to_done'
AND breached IS NOT NULL;psom_sla <- tbl(con, in_schema("jira", "psom_sla"))
latest_date <- psom_sla |> summarise(d = max(date, na.rm = TRUE)) |> pull(d)
psom_sla |>
filter(
as.Date(date) == as.Date(latest_date),
sla == "time_to_done",
!is.na(breached)
) |>
summarise(
total_sla_tickets = n(),
breached_count = sum(as.integer(breached), na.rm = TRUE),
breach_rate = sum(as.integer(breached), na.rm = TRUE) / as.double(n())
)Interpretation notes
The rate is a proportion between 0 and 1. Tickets whose clock never started (null breached) are excluded from both numerator and denominator. Check total_sla_tickets to understand the sample size underlying the rate.
The goal is set per ticket (16, 24, 40 and 72 working hours all occur), so a breach means the ticket missed its own target, not a shared one. Working hours run 09:00 to 17:00, so a 40-hour goal is one working week. To compare how long tickets took, use elapsed_minutes instead, which is populated from 2026-09-29.
A ticket that is still open and inside its goal is recorded as not breached (false), the same as a ticket that met the SLA. The rate on a live snapshot is therefore a lower bound: some of those open tickets may still breach before they are resolved. It is most meaningful once the snapshot’s tickets are largely resolved.
Alternative names
- time to done SLA breach rate
- resolution SLA compliance
- time-to-done SLA failure rate