Flow Dataset Diagnosis

Move from a set of degrading flow metrics to a non-blaming system hypothesis and a reversible experiment.

Scenario

Foundation case

The support team's last four weeks show: WIP 18 → 23 → 29 → 34; median cycle time 4.2 → 5.1 → 6.4 → 7.8 days; throughput 42 → 41 → 43 → 42 tickets; critical reopen rate within 14 days 5% → 7% → 11% → 14%. Incoming demand stayed between 46 and 48 tickets per week, and team size did not change. The data shows relationships, not causes.

Your role

You are the service manager investigating degradation in support flow.

Your task

Separate data from cause, then propose at least three system hypotheses, discriminating evidence, and a reversible experiment.

Four weeks of flow data

WIP rose 18 → 23 → 29 → 34; median cycle time 4.2 → 7.8 days.
Throughput stayed roughly flat: 42 → 41 → 43 → 42.
Critical reopen rate rose 5% → 14%.
Demand stayed 46-48 per week and team size was unchanged.

Your response

There is no single correct answer; we will evaluate the reasoning chain you build.

Separate data from interpretation0/4000 · min 20

At least three system hypotheses0/4000 · min 25

What additional data would you check0/4000 · min 15

Two-week, reversible experiment0/4000 · min 15

Success and balancing signals0/4000 · min 15