L.L.Bean staffed calls by total cost, not a service-level target

L.L.Bean was a large catalog telemarketer, so the number of phone lines, agents and queue positions shaped both customer experience and cost. The old approach used service-level criteria, which could overbuild capacity without proving the trade-off was worth it.

The team reframed the call center as an economic queue. It estimated the cost of capacity and the cost of abandoned or retried calls, then optimized the whole telecommunications deployment rather than tuning one lever at a time.

The recommended deployment differed substantially from service-level planning. The Interfaces article reports that L.L.Bean accepted the method and estimated annual savings of $9 million to $10 million.

Price the queue instead of chasing a target

The model converted soft customer-service trade-offs into a single cost objective.

It optimized lines, staff and waiting-room size together, avoiding local fixes.

Abandonment behavior was estimated empirically instead of assumed away.

A service level is a proxy. If you can price waiting, abandonment and capacity directly, optimize the economics rather than the proxy.

The paper became a reference point for later call-center staffing research, cited in surveys and models of nonstationary service centers and abandonment-sensitive capacity sizing.