Field study 01/Tests Thesis 003

Decision rights, redrawn

A national parcel operator had automated dispatch and still missed cut-offs. The constraint was not throughput. Six roles could halt a route and none could restart one.

Strengthened
Organization
Logistics · 4,000 people · 14 hubs
Duration
2 quarters · 2026
Sponsor
Chief operating officer
Institute
Gabriel Rivera, Anneli Lindqvist
Field study 012 quarters · 2026Tests Thesis 003Strengthened

Thesis 003 · Management Is Becoming an Operating System

The management layer has no design principle.

Spans and layers were set by the cost of coordination. That cost has collapsed and the structure has not moved. Nobody can say what a manager is for in an automated firm, how many are needed, or what they should be measured on. The system is a management operating model built for the firm as it now runs.

Quarter by quarter

  1. Q0

    Baseline written with the sponsor: cycle time, rework and cut-off misses by hub, for the two quarters before the study. Handoffs mapped for the twelve most common exception types.

  2. Q1

    Decision rights rewritten for exceptions. Six roles that could halt a route were reduced to two, and a restart right that no role held was created and assigned. No tooling changes. No headcount changes.

  3. Q2

    Measurement against baseline at all fourteen hubs. Verification of self-reported rework by record review at nine hubs. Peak-season adjustment applied using the prior year's curve.

Against the written baseline

18%shorter cycle time in two quarters
11%less rework, self-reported and verified at 9 of 14 hubs
6 to 2roles able to halt a route

Cycle time fell 18 percent against the written baseline, with the largest gains in the hubs that had the most halt-capable roles before the change.

Rework fell 11 percent by self-report; verified figures at nine hubs were within 2 points of self-report.

Cut-off misses fell, but the effect could not be separated from the seasonal adjustment and is not claimed.

What changed in the thesis

The study strengthened the decision-rights finding in Thesis 003 and provided its first evidence from physical operations, where the survey effect had been weakest. It did not test the compounding claim; two quarters is too short.

What this study cannot establish

  • Observational, single organization, no control group.
  • Peak season fell in Q2 and was adjusted using the prior year's curve, which assumes a stable seasonal pattern.
  • The sponsor was the chief operating officer, whose attention is itself an intervention.