SLA health

Three scorecards, deliberately not one blended number. A strong internal result would otherwise hide a weak partner one - and as the business sells to other businesses, partner expectations are the ones that tighten.

Service review pack
Internal
Meeting goal
Staff can route around a slow fix, so targets are the loosest.

SLA attainment · last 30 days

96%

50 of 52 resolved tickets inside target · goal 85%

Resolved
52

in the window

Avg time to resolve
1d 12h
Open now
14
Open past target
6

still breaching - not in attainment

SeverityTargetWithin
Critical4h1/2
High1d8/9
Medium3d14/14
Low5d27/27
Consumer
Below goal
Customers can't route around anything and will leave - tighter targets.

SLA attainment · last 30 days

75%

30 of 40 resolved tickets inside target · goal 90%

Resolved
40

in the window

Avg time to resolve
1d 20h
Open now
2
Open past target
0

still breaching - not in attainment

SeverityTargetWithin
Critical2h1/1
High8h6/9
Medium2d8/11
Low4d15/19
Partner
Below goal
Partners have contracts and their own customers waiting downstream - the tightest targets, and the highest goal.

SLA attainment · last 30 days

60%

18 of 30 resolved tickets inside target · goal 95%

Resolved
30

in the window

Avg time to resolve
1d 13h
Open now
6
Open past target
0

still breaching - not in attainment

SeverityTargetWithin
Critical1h1/2
High4h5/6
Medium1d4/9
Low3d8/13
Resolution-time targets
How long a ticket may stay open before it counts as a miss. This is the table the scorecards are computed against - read from configuration, not hardcoded into the scoring, and tighter for every step outward. Edit a cell to see the scorecards above recompute instantly; edits are saved in this browser only.
AudienceCriticalHighMediumLowAttainment goal
Internal
h

4h

h

1d

h

3d

h

5d

%
Consumer
h

2h

h

8h

h

2d

h

4d

%
Partner
h

1h

h

4h

h

1d

h

3d

%

About these numbers: the 120 resolved tickets in the history behind these scorecards are generated by a seeded generator in this repository, not measured from any real system. Every percentage above is computed from that data plus the current queue, which includes a few real captured records - so it is real arithmetic, mostly on invented tickets but not entirely. Attainment counts only resolved tickets; the “open past target” figure is shown beside it because attainment alone can't see tickets that are still breaching.