156
Requests handled
Submitted across the four weeks. Every one of them has an owner in the record — none drifted without one.
Sample evidence pack
What week 4 looks like, on paper.
This is the report a pilot team receives at the end of week 4: what was handled, what the numbers say, what improved — and what missed, with the miss included. A report that only shows wins is not evidence.
A sample, not a result.
Illustrative workflow using sample data. No customer data shown.
Week 4
The headline numbers
Four numbers, each with its definition agreed in week 0. Every one of them is fictional — this page shows the report's structure, not anyone's results.
91%
SLA attainment
First-response targets met across all urgencies. The target agreed in week 0 was 90% — met, but not everywhere (see below).
14
Safety flags
Safety-flagged requests; 14 of 14 acknowledged within the 15-minute target. Zero unacknowledged, all pilot long.
38 min
Median acknowledgement, week 4
Down from 2h 41m in week 1. The queue stopped being decided by whoever noticed a problem first.
Illustrative workflow using sample data. No customer data shown.
Volume & SLA
Requests handled, targets met — per urgency, not blended
The overall number is 91% against a 90% target. The per-urgency breakdown is where the honest reading lives: Medium missed.
156 total across the pilot. Steady volume, not a launch spike.
34Week 1
41Week 2
37Week 3
44Week 4
Illustrative workflow using sample data. No customer data shown.
Black marker: the target agreed in week 0. Amber: missed.
Critical100% · target 100%
High94% · target 90%
Medium89% · target 90% · missed
Low88% · target 85%
Illustrative workflow using sample data. No customer data shown.
Safety
The one measure that cannot miss
Safety escalations carry an explicit acknowledgement target agreed in week 0. This measure is reported separately from everything else, on purpose.
14 safety-flagged requests over four weeks.
Zero unacknowledged safety flags for the entire pilot. The median acknowledgement time for safety flags was 6 minutes.
Illustrative workflow using sample data. No customer data shown.
Why this gets its own section
A blended SLA number can hide a safety failure inside a good average. So safety is counted on its own, every week, with its own target — and a single unacknowledged safety flag would put it at the top of this report, ahead of every win on this page.
This pilot had none. The report says so plainly because the weeks where it does not will say that just as plainly.
Vendors
Commitments captured — and kept
A promise without a recorded date is not a commitment. The pilot tracked both: did we record the promised date, and was it met.
96 requests involved a vendor. Black marker: the 90% target.
Jobs with a recorded promised date78% · target 90% · missed
Of recorded commitments, met on time64% · target 90% · missed
75 of 96 vendor jobs had a promised date recorded; 48 of those 75 were completed by the promised date.
Illustrative workflow using sample data. No customer data shown.
Reading the miss
The misses are the most useful part of this section. Two of the three regular vendors would not give promised dates by email, so engineers logged “TBC” — which is how 21 jobs ended up with no commitment at all. That is a process gap, not a vendor conspiracy, and it has an agreed fix in the section below.
Improvement
Week 1 to week 4
A pilot should show the operation getting visibly better — or say honestly that it did not. These two measures moved.
Minutes from submission to someone taking responsibility.
161mWeek 1
104mWeek 2
72mWeek 3
38mWeek 4
Illustrative workflow using sample data. No customer data shown.
Share of the open queue with no owner assigned, end of each week.
23%Week 1
15%Week 2
9%Week 3
4%Week 4
Illustrative workflow using sample data. No customer data shown.
Honest accounting
What missed, and why
Three things did not hit their targets. Each one has a cause and an agreed fix — that is what makes the next four weeks different from the last four.
01
Vendor-commitment capture: 78% vs a 90% target
Why: Two of the three regular vendors would not give promised dates by email, so engineers logged “TBC” instead of a date. 21 of 96 vendor jobs had no recorded commitment — and a commitment you do not record is one you cannot hold.
Fix: Agreed fix: a vendor job can no longer move to “in progress” without a promised date. Phone confirmations count — logged with who confirmed and when.
02
Medium-urgency SLA: 89% vs a 90% target
Why: Three Medium requests missed the 24-hour first-response target. The furthest miss was a fictional request (JFS-2419), acknowledged 47 minutes late because the assigned owner was on leave and the request was never reassigned to the deputy.
Fix: Agreed fix: a deputy-owner rule for leave cover, so no request sits with someone who is not there.
03
One recurring issue still open
Why: The loading-bay shutter has been reported three times this month (fictional JFS-2422). Each fix held for a week. Root cause is not yet established, so it is flagged as a problem-record candidate instead of being closed again.
Fix: It is in this report because closing it a third time without a root cause would be the point of the exercise failing.
Illustrative workflow using sample data. No customer data shown.
Method
How this report is built
The structure matters more than any number on this page.
- · Every definition — what counts as “acknowledged”, what each urgency target is — is agreed in week 0, before any data exists to argue about.
- · Every number traces to the event record: each request's history of what happened, who did it, and when. Nothing here is written by hand or adjusted after the fact.
- · Every request counts, including the ones nobody looked at. A measure that excludes its misses is marketing, not evidence.
- · The report is generated from the desk, not assembled in a slide deck. What you see here is what the pilot team sees every week.
Your pilot
Agree the measures for your operation.
Tell us what “working” means for your team, and we will build the pilot's success criteria around it — then report against them, misses included.