Just Another Manic Monday in the Call Center: When Equal Demand Does Not Produce Equal Service

Abstract

Call centers ordinarily explain service-level variation through demand, service time, staffing, routing, and customer patience. Yet performance may also vary systematically by weekday even when workload is comparable. This study examines a Monday effect in operational data from a large U.S. insurance-service organization. The dataset contained 13,024 intraday records across 32 business days, 37 reporting queues, and 20 business groups. Because call counts were cumulative within day, the primary analysis used each queue’s final daily snapshot, producing 1,127 active queue-day observations. Monday service level averaged .658, compared with .715 on Tuesday, although log-transformed call volume did not differ significantly between the two days. Ordinary least squares models with queue-clustered standard errors showed that Monday remained associated with a 5.1-percentage-point reduction in service level after controlling for call volume, average handle time, and business-group fixed effects. The result persisted across all-weekday, logit-transformed, volume-threshold, nonboundary, weighted, paired-queue, and nonparametric sensitivity analyses. The findings distinguish workload effects from weekday effects and suggest that start-of-week operational frictions may impair service beyond what volume alone explains. Because staffing, schedule adherence, absence, backlog, and call-mix measures were unavailable, the analysis identifies a robust temporal pattern rather than a causal mechanism. The study contributes to service-operations research by treating weekday as an operational state variable and offers managers a focused diagnostic agenda for Monday readiness.