CALL CENTERS (CENTRES) - Faculty of Industrial Engineering and ...
CALL CENTERS (CENTRES) - Faculty of Industrial Engineering and ...
CALL CENTERS (CENTRES) - Faculty of Industrial Engineering and ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
at .<br />
Abstract. This paper models the cross-selling problem <strong>of</strong> a call center as a dynamic service<br />
rate control problem. The key trade<strong>of</strong>f between revenue generation <strong>and</strong> congestion in a call<br />
center is addressed in a dynamic framework. The question <strong>of</strong> when <strong>and</strong> to whom to cross-sell<br />
is explored using this model. The analysis shows that unlike current marketing practice which<br />
targets cross-sell attempts to entire customer segments, optimal dynamic policies may target selected<br />
customers from different segments. Structural properties <strong>of</strong> optimal policies are explored.<br />
Sufficient conditions are established for the existence <strong>of</strong> preferred calls <strong>and</strong> classes; i.e., calls that<br />
will always generate a cross-sell attempt. Numerical examples, that are motivated by a real call<br />
center, identify call center characteristics that increase the significance <strong>of</strong> considering dynamic<br />
policies rather than simple static cross-selling rules as currently observed. The value <strong>of</strong> these dynamic<br />
policies <strong>and</strong> static rules are compared. The numerical analysis further establish the value<br />
<strong>of</strong> different types <strong>of</strong> automation available for cross-selling. Finally, the structural properties lead<br />
to a heuristic that generates sophisticated static rules leading to near optimal performance both<br />
for a loss system <strong>and</strong> a queueing system.<br />
Keywords: Call center, Cross-selling, Revenue management, Customer relationship management,<br />
Dynamic control, Loss system<br />
153. Sisselman, Michael E. <strong>and</strong> Ward Whitt. Value-based routing <strong>and</strong> preference-based routing in<br />
customer contact centers. Working paper, 2005. Also: A preference-based-routing example<br />
solved by linear programming with excel solver. Supporting material, 2005. Available at:<br />
.<br />
Abstract. Telephone call centers <strong>and</strong> their generalizations—customer contact centers—usually<br />
h<strong>and</strong>le several types <strong>of</strong> customer service requests (calls). Since customer service representatives<br />
(agents) have different call-h<strong>and</strong>ling abilities, contact centers exploit skill-based routing (SBR) to<br />
assign calls to appropriate agents, aiming to respond properly as well as promptly. Established<br />
agent-staffing <strong>and</strong> call-routing algorithms ensure that agents have the required call-h<strong>and</strong>ling<br />
skills <strong>and</strong> that constraints are met for st<strong>and</strong>ard congestion measures, such as the percentage <strong>of</strong><br />
calls <strong>of</strong> each type that ab<strong>and</strong>on before starting service <strong>and</strong> the percentage <strong>of</strong> answered calls <strong>of</strong><br />
each type that are delayed more than a specified number <strong>of</strong> seconds. We propose going beyond<br />
these traditional performance measures to focus on the expected value accrued from having<br />
the agent h<strong>and</strong>le the call. Expected value might represent expected revenue or the likelihood<br />
<strong>of</strong> first-call resolution. Value might also reflect agent call-h<strong>and</strong>ling preferences. We show how<br />
value-based routing (VBR) <strong>and</strong> preference-based routing (PBR) can be introduced in the context<br />
<strong>of</strong> an existing SBR framework, where the existing SBR is based on static-priority routing<br />
using a highly-structured priority matrix. Since VBR <strong>and</strong> PBR use the same SBR framework,<br />
they can be implemented with the existing SBR algorithm in the automatic call distributor<br />
(ACD); it is not necessary to replace the ACD. We use mathematical programming to find an<br />
effective priority matrix. We select the priority matrix to use during a specified time interval<br />
(e.g., 30-minute period) by maximizing the total expected value over that time interval, subject<br />
to constraints that ensure that st<strong>and</strong>ard performance constraints are met.<br />
Keywords: Customer contact centers, Telephone call centers, Skill-based routing, Value-based<br />
routing, Preference-based routing, Indirect value-based routing, Priorities, Mathematical pro-<br />
58