1,434 cases · 8,577 events · 27 activities · 10 work groups
A process-mining review of a real case-handling workflow — variants, rework, handoffs, queue time, service levels and automation screening. The headline is that almost none of it matters as much as it looks like it does.
Everything a conventional process review produces — variant maps, handoff diagrams, queue times, rework counts — describes the middle band. Remove every queue, handoff delay and rework loop in the logged workflow and mean cycle time falls from 59.1 to 53.6 days. The other 91% of the clock is not instrumented at all.
| Scope | Cases | Cycle (d) | Intake | Worked | Closure | % worked |
|---|---|---|---|---|---|---|
| all channels | 1,327 | 59.1 | 22.6 | 5.5 | 30.9 | 9.4% |
| Internet | 1,146 | 57.9 | 22 | 5.3 | 30.5 | 9.2% |
| Desk | 107 | 68.8 | 25.7 | 5.9 | 37.1 | 8.6% |
| Post | 52 | 59.4 | 26.3 | 7.7 | 25.4 | 12.9% |
| 21 | 76.2 | 32.2 | 10.8 | 33.2 | 14.1% | |
| Intern | 1 | 22.6 | 13.4 | 1 | 8.3 | 4.3% |
The pattern holds in every channel. It is structural, not a quirk of one route.
The bottleneck analysis came out clean: one transition accounts for 2,422 days of waiting across 791 occurrences — about 31% of all queue time. The obvious thing to fix. Then the arithmetic refused to close: the gaps between logged steps total roughly 5.4 days per case against a 59-day cycle. The bottleneck is real, and it is 31% of a window worth 9% of the total — about 1.8 days of the 59. A programme that opened with it would have reported a real improvement and moved the number almost not at all.
| From | To | Count | Mean (d) | Total days |
|---|---|---|---|---|
| T05 Print and send confirmation of receipt | T06 Determine necessity of stop advice | 791 | 3.1 | 2,422 |
| Confirmation of receipt | T02 Check confirmation of receipt | 1,079 | 0.8 | 901 |
| T04 Determine confirmation of receipt | T05 Print and send confirmation of receipt | 1,177 | 0.6 | 758 |
| T10 Determine necessity to stop indication | T02 Check confirmation of receipt | 155 | 3.1 | 475 |
| T06 Determine necessity of stop advice | T10 Determine necessity to stop indication | 1,165 | 0.4 | 440 |
| T02 Check confirmation of receipt | T04 Determine confirmation of receipt | 1,119 | 0.3 | 339 |
| Confirmation of receipt | T06 Determine necessity of stop advice | 239 | 1.1 | 258 |
| T10 Determine necessity to stop indication | T05 Print and send confirmation of receipt | 102 | 2.5 | 255 |
| Channel | Cases | Mean cycle (d) | Breach rate | Mean days late |
|---|---|---|---|---|
| Internet | 1,146 | 57.9 | 28.2% | 30.3 |
| Desk | 107 | 68.8 | 38.3% | 35.5 |
| Post | 52 | 59.4 | 32.7% | 15.6 |
| 21 | 76.2 | 38.1% | 70.5 | |
| Intern | 1 | 22.6 | 0.0% | 0 |
The highest-volume channel performs best. e-mail is worst on both counts — small volume, but when it breaches it breaches by more than two months.
| Steps in case | Cases | Mean cycle (d) | Breach rate |
|---|---|---|---|
| a. 1-4 steps | 119 | 36.2 | 15.1% |
| b. 5-6 steps | 1,068 | 60 | 29.4% |
| c. 7-9 steps | 48 | 92.1 | 56.2% |
| d. 10+ steps | 92 | 60.5 | 32.6% |
Cases with 7–9 steps breach at 56.3% against 15.1% for 1–4. Usable, because step count is visible on day three while cycle time is only knowable at the end. Not monotone — 10+ step cases breach less than 7–9 — so it is a flag worth testing, not a scoring model.
| Activity | Cases | With repeat | % repeating | Worst | Wasted |
|---|---|---|---|---|---|
| T06 Determine necessity of stop advice | 1,309 | 59 | 4.5% | 10 | 107 |
| T02 Check confirmation of receipt | 1,316 | 35 | 2.7% | 6 | 52 |
| T03 Adjust confirmation of receipt | 37 | 11 | 29.7% | 5 | 18 |
| T07-2 Draft intern advice aspect 2 | 15 | 8 | 53.3% | 6 | 17 |
| T07-1 Draft intern advice aspect 1 | 29 | 8 | 27.6% | 9 | 16 |
| T07-5 Draft intern advice aspect 5 | 16 | 5 | 31.2% | 6 | 11 |
| T17 Check report Y to stop indication | 20 | 2 | 10.0% | 5 | 6 |
| T18 Adjust report Y to stop indicition | 2 | 2 | 100.0% | 4 | 4 |
Highlighted rows are the process's own explicit correction steps. One of them repeats within a case up to five times — a correction that itself needs correcting.
| From | To | Handoffs | Mean wait (h) | Max wait (h) |
|---|---|---|---|---|
| Group 3 | Group 2 | 992 | 11.6 | 314.5 |
| Group 4 | Group 3 | 857 | 1 | 138.9 |
| Group 2 | Group 1 | 800 | 73.7 | 6,455.3 |
| EMPTY | Group 4 | 712 | 0.7 | 166.6 |
| EMPTY | Group 1 | 429 | 19.1 | 2,397.4 |
| Group 1 | Group 4 | 336 | 11.9 | 917.7 |
| Group 1 | EMPTY | 321 | 112.6 | 2,663.4 |
| EMPTY | Group 3 | 171 | 19.1 | 172.5 |
The slowest handoff averages 73.7 hours, with a worst case of 6,455 hours — 269 days on a single handoff. "EMPTY" is the log's own value where no group was recorded; kept rather than dropped.
| Activity | Executions | People | Mean hours | CoV | Score |
|---|---|---|---|---|---|
| T05 Print and send confirmation of receipt | 1,300 | 33 | 19.1 | 2.7 | 353 |
| T04 Determine confirmation of receipt | 1,307 | 37 | 10.2 | 4.9 | 222 |
| T02 Check confirmation of receipt | 1,368 | 40 | 27.9 | 5.3 | 217 |
| T06 Determine necessity of stop advice | 1,416 | 34 | 51.8 | 5.5 | 217 |
| T10 Determine necessity to stop indication | 1,283 | 34 | 11.3 | 6.2 | 179 |
| T03 Adjust confirmation of receipt | 55 | 17 | 129.7 | 2.5 | 16 |
Frequent, consistent and quick steps score highest — a ranking, not a decision. The top candidate is executed 1,300 times by 33 people, and is also the source of the biggest queue in the process.
| Check | Cases | Treatment |
|---|---|---|
| Still open at extract | 105 | no end date, so no cycle time |
| End date before start date | 2 | impossible; excluded, not clamped to zero |
| Committed date differs from deadline | 37 | two fields disagree; SLA uses the committed date |
| Usable for cycle time | 1,327 | the denominator behind every timing figure |
Real-life event log, publicly released for research. A public-sector case-handling process, not a banking one — the domain differs, the method does not. No cost or headcount data, so benefits are stated in days rather than money. · SQL, workbook and memo on GitHub