A WEEKLY LOGISTICS-CONTROLLING SYSTEM FOR EXPORT SUPPLY CHAINS: CORRIDOR-LEVEL EVIDENCE FROM A TEXTILE EXPORTER

Abstract

Export-oriented manufacturers in landlocked, multi-border economies compete on the reliability of long international
corridors, yet such firms typically measure logistics performance retrospectively rather than as a live control loop.
This article develops a weekly logistics-controlling system whose novelty is procedural: it embeds the monitoring of order
timeliness, completeness, and quality into the firm’s routine weekly management cycle using three-zone control limits,
enabling the early detection of deviations and shortening management’s response time to export disruptions. The system
is demonstrated on a complete census of the 2025 export shipments of an Uzbek textile exporter (N = 50; seven corridors).
At the firm level, OTIF was 84% and POR was 68%.
The decisive empirical result is a corridor-level contrast: the Priority tier (Kazakhstan, Tajikistan, and Turkmenistan) substantially
outperformed the more distant Premium tier (Russia, Armenia, Türkiye, and Iran) in cycle time (4.3 vs. 11.3 days;
Mann–Whitney p < 0.001), OTIF (94% vs. 63%; Fisher p = 0.009), and POR (79% vs. 44%; p = 0.021). In contrast, defect
and documentation-error rates did not differ significantly, identifying timeliness on distant corridors, rather than quality
failures, as the primary operational constraint. A log-revenue regression with heteroscedasticity-robust errors confirmed
cycle time as a significant correlate of order value (β = −0.45, p = 0.004). When correctly specified, the indicator set exhibited
no multicollinearity (VIF ≤ 2.6).
The study contributes a low-cost, SME-feasible control routine that is distinct from capital-intensive digital control towers,
together with a corridor-prioritisation matrix and a four-phase implementation roadmap. As a single-firm baseline study, it
does not claim a proven before-and-after effect; instead, the proposed improvement targets are presented as hypotheses
for future controlled validation.