Professional Documents
Culture Documents
ASSIGNMENT 1
BUSINESS ANALYSIS AND DECISION MAKING
MGMT 90141
MUKUL SETHI -632244
DIVYA GUPTA – 724894
ALEXANDRA PITSAS – 748975
HE ZHU- 648868
PIRADA P. - 684284
Table of Contents
1. Introduction .......................................................................................................................... 3
6. Conclusion ........................................................................................................................... 10
Appendix ................................................................................................................................... 12
1. Introduction
wide range of services to over 10 million customers each day (Singtel 2015, p.22).
more than 20 countries globally (‘yes’ Optus 2013, p.4). Like many
from its competitors (‘yes’ Optus 2010, p.2). However, this can be a challenging
estimated that $300 billion is spent on call centres globally and approximately
80% of a company’s call centre costs are personnel costs (Bhulai, Koole & Pot
2008; Mandelbaum & Zeltyn 2009). Thus, by reducing this labor cost by just a
modeling to determine the optimal scheduling of their staff to reduce wait times
required during different periods of the day to meet customer demand, thus
reducing average wait times, enhancing customer experience and minimizing
costs. The decision variables that need to be considered in this problem include the
minutes), hourly wages for different skilled staff, the maximum office capacity,
and the minimum amount of staff needed during peak and off peak times.
1.1 Methodology
and collect 10 international journal articles, we used the online sources Scopus,
IEEE Explore Digital Library, Informs Pubs Online, and Science Direct by typing
The evaluative criteria specifically used to analyze the strengths and weaknesses
of the mathematical models used in the 10 journal articles were based around the
constraints faced by call centers like Optus. These constraints were examined in
each article and included, the average customer wait time, multi-skilled staff, the
number of staff required for specific shift times to meet customer demand, and the
2. Literature Review
Recently call centers have grown rapidly and have become an integral part of
quick support services to customers (Seada & Eltawil 2015). Therefore, call
centers need to ensure they have an accurate match of customer demand and staff
supply during different periods of the day (Madelbaum & Zeltyn 2009).
Understaffing call centers typically results in longer wait times for customers,
that 63% of Australian customers have ended business with a company after a bad
call center experience (Bhulai et al. 2008, p.411; ‘yes’ Optus 2013, p.18).
Therefore, it would be beneficial for call centers to address such staff scheduling
The literature analyzed in this report offers a variety of mathematical models and
Gaussian linear mixed model calculations. These models and algorithms were used
majority of the research considers constraints such as customer waiting time (not
(2008) and Cheng and Huo (2011) considered the potential possibility of
employing permanent and casual staff, in which the permanent staff are always
available to work, while the casual staff only work when the call center is busy,
thus minimizing the costs associated with increased staff scheduling. Overall, the
research analyzed has demonstrated that LP and IP have successfully been applied
during different periods of the day in an effort to minimize personnel costs, whilst
article.
Although call centers seem peripheral in most organizational operations, they are
dissatisfied customers and loss of future business or revenue. With the aim to
reduce costs while maintaining a good level of customer service, numerous studies
Bhulai, Koole and Pot (2008) approach the staffing optimization problem with a
could be useful in a fast paced environment. However, the key strength of this
model is that it considers the multiple skill sets call center personnel can have,
which is similar to that of the work of Cezik & L’Ecuyer (2008), Schrieck,
Chevalier & Aksin (2011), and Ertogral & Bamuqabel (2008). The mathematical
models that support multiskilling staff are very practical and could be adapted to
solve Optus staffing problem as Optus call center supports multiple areas of
business such as sales, billing, technical support, and so on, which will require
Accordingly, the aforementioned papers take the costs of individuals with varying
skillsets into consideration. That is, the cost of staff is not in one single rate but
differs based on the amount of skills an individual has. This differing staff cost
Wolf & Balter (2005) and Cheng & Huo (2011) do not build their models based on
multi-skill call center, they consider the different wages in different working shift,
One strong aspect of the models presented by Cheng & Huo (2011) and Noiman,
Feigin & Mandelbaum (2009) is that they use inhomogeneous Poisson process
arrival rate to realistically forecast call center staff’s workload that is not constant
throughout the day. The latter model assumes that the arrival rate remains constant
in every 30-minute interval to simulate real life uncertain rate of calls received and
The detailed real world constraints are also presented in multiple models and add
value to solving staffing optimization problems. The model by Ma & Liu (2012)
applies rigorous business rules such as the number of working hours and resting
hours per week, which can be adapted to workplace rules and regulations at Optus
when scheduling staff. Additionally, Mandelbaum & Zeltyn (2006), Cezik &
L’Ecuyer (2008), Cheng & Huo (2011), and Baron & Milner (2009) consider the
fact that customers may abandon the call after a certain period of waiting time.
Lastly, in addition to staff wage, Bhandari, Wolf & Balter (2005) and Baron &
Milner (2009) both include penalty cost as a service level constraint on service
delays when forming staffing optimization models to reflect real world cost to the
service quality.
According to the evaluation criteria and the relevance to Optus call center case, the
An essential constraint in Optus call center case is the varying wages for staff with
varying skills. In their articles, Mandelbaum and Zeltyn (2006), Ma and Liu
(2012), Baron and Milner (2009) and Noiman, Feigin and Mandelbaum (2009)
have not taken the multi-skills staff into consideration when they build the model
to solve the problem. Both varying wage and multi-skill of staff are not discussed
in their articles, which limits the relevance to Optus call center case.
conducted in most articles. Although arrival rates are considered in most articles as
a key constraint to conduct staff scheduling, Bhulai, Koole and Pot (2008),
Mandelbaum and Zeltyn (2006) and Baron and Milner (2009) have not considered
uncertain arrival rates and simplify the arrival rates as constant during each one
hour period. Furthermore, in the articles by Bhulai, Koole and Pot (2008),
Bhandari, Wolf and Balter (2005), and Ma and Liu (2012), the fact of customer
high customer satisfaction in Optus call center case. Then, another fact of potential
call abandonment resulted by lengthy waiting time is also not considered in these
articles.
discussed, although the main objective of optimizing the number of staff could
essentially lead to that. Similarly, Baron and Milner (2009) focus more on
maximizing profit rather than minimizing cost, as the call center in Optus is
Finally, three more common issues are identified. None of the ten articles have
considered the length of each service session or average length, which is also an
discussed in each article, and the scope is normally limited and simplistic for a
Having examined a number of articles for the Optus call centre problem, one of
the recurring limitations was that some of the main constraints were not
were abandonment rate, uncertain arrival rate, queuing, staff cost, customer-
to Optus.
expected profit for the Service Level Agreements (Baron & Milner 2009).
Additionally, one of the main aspects that was apparent in many articles (Bhulai,
Koole and Pot 2008; Ertogral & Bamuqabel 2008) is regarding the multi-skill
group to be involved in Optus call centre. Thus, including multi-skill groups will
help customers to choose the right service according to the needs which will
6. Conclusion
As discussed above the paper presented has tried to cover the strengths and
weaknesses of all the articles presented and tried to relate it to the problems that
Optus faces. On the basis of the presented papers and their relation with Optus’s
problem, a following report will suggest the solution for the same.
REFERENCES
Baron, O & Milner, J 2009, ‘Staffing to maximize profit for call centers with alternate
service-level agreements’, Operations Research, vol. 57, no.3, pp. 685-700.
Bhandari, A, Scheller-Wolf, A & Harchol-Balter, M 2005, ‘An exact and efficient algorithm
for the constrained dynamic operator staffing problem for call centers’, viewed 10 August
2015,
<http://www.cs.cmu.edu/~harchol/Papers/dynamic-staffing.pdf>.
Bhulai, S, Koole, G & Pot, A 2008, ‘Simple methods for shift scheduling in multiskill call
centers’, Manufacturing & Service Operations Management, vol. 10, no.3, pp. 411-420.
Brigandi, AJ, Dargon, DR, Sheehan, MJ & Spencer III, T 1994, ‘AT&T’s call processing
simulator (CAPS) operational design for inbound call centers’, Interfaces vol. 24, no.1, pp.
6–28.
Cezik, MT & L’Ecuyer, P 2008, ‘Staffing multiskill call centers via linear programming and
simulation’, Management Science, vol. 54, no.2, pp. 310-323.
Cheng, F & Huo, J 2011, ‘Dynamic operator staffing problem for call centers with
uncertain arrival rates’, 2011 International Conference on Business Computing and Global
Information (BCGIN), Shanghai, China, 29-31 July 2011, viewed 13 August 2015,
<http://ieeexplore.ieee.org.ezp.lib.unimelb.edu.au/stamp/stamp.jsp?tp=&arnumber=60039
75>.
Ma, Y & Liu, L 2012, ‘Solving call center agent scheduling problem through improved
adaptive genetic algorithm’, 2012 Fifth International Symposium on Computational
Intelligence and Design (ISCID), Beijing, China, 28-29 October 2012, viewed 12 August
2015,
<http://ieeexplore.ieee.org.ezp.lib.unimelb.edu.au/stamp/stamp.jsp?tp=&arnumber=64055
57>.
Mandelbaum, A & Zeltyn, S 2009, ‘Staffing many-server queues with impatient customers:
constraint satisfaction in call centers’, Operations Research, vol. 57, no.5, pp. 1189-1205.
‘yes’ Optus 2010, Optus submission to Australian Communications and Media Authority
public enquiry: Reconnecting the customer, viewed 20 August 2015,
<http://www.acma.gov.au/webwr/_assets/main/lib311946/public_inquiry_submission_121-
optus_21-09-10.pdf>
‘yes’ Optus 2013, Customer Service: Perceptions and Expectations: A report by Optus,
viewed 21 August 2015,
<https://www.optus.com.au/dafiles/OCA/AboutOptus/MediaCentre/SharedStaticFiles/Shar
edDocuments/CustomerExpectationsReport.pdf>.
Appendix
Simple Methods This paper The model is simple and The model
for Shift introduces an easy to implement, determines staffing
Scheduling in integer- which makes calculation levels and
Multiskill Call programming possible in seconds. generates shifts
Centers model for shift separately in two
scheduling in multi steps, which could
skill call centers. result in sub
The objective of Consider differences in optimality/inefficie
the model is to constraints such as call ncy
minimize personnel types, skill groups, time
costs, while period, shift types (diff Low complexity:
meeting the pay in diff shift) doesn’t consider
staffing abandon calls,
requirements for return calls,
each group in each queuing, arrival
The model allows agents
period. rates, average wait
to work in different
groups during the same
shift time, etc.)
Suggests offering
arriving customers with
an estimated waiting
time
Staffing Multiskill This paper is based The solution provided is Although the
Call Centers via around optimizing optimal in case of large solution is optimal
Linear the staffing problem instances. as discussed in the
Programming and problem at a paper, but is not the
Simulation multiskill call best solution
center while always. But still is
minimizing the Improves the satisfaction a better result than
operating costs. as the call goes to the diverting the call
The paper uses concerned authority. randomly to
cutting plane anyone.
methodology of It also minimizes
Atlason et al. operating costs. This may increase
(2004) to solve this the wait time.
problem
Dynamic Operator This paper talks Uses the concept of Do not consider the
Staffing Problem about how to Flexible staffing i.e. the adjustment cost of
for Call Centers maximize profit ability to dynamically call center
with Uncertain and meet call- adjust staffing levels managers.
Arrival Rate center level goals with traffic load.
by considering Does not consider
single-class call Considers uncertain analysis of call-
center as a queuing arrival rates. centers
model having performance.
uncertain arrival Determine the optimal
rate. This is done number of permanent Does not consider
by flexible staffing. and temporary operators. service-level
Thus, by constraints.
deterministic fluid Considers customer-
approximations we abandonment and Doesn’t consider
get the number of customer-waiting cost multi-class call
permanent and center.
temporary Reduces unnecessary
operators to hiring cost
maximize the
expected net
revenue
Peakedness-Based This paper talks The paper Helps in Still in early stage
Staffing for Call about the understanding the in the area of
Center Outsourcing variability in the variability in call peakedness so its
types of calls a call variance credibility cannot
Centre gets and be established
how it can be This is the first paper to
understood by propose the use of
using the peakedness to measure
peakedness call arrival variability in The research is not
principle. call centers work with a 100%
accuracy
The variability in It also does help by
the type of the calls giving a simple approach
will help the to measuring variance in
company staff the calls thus helping in
people accordingly staffing accordingly
giving more cost
effectiveness.
It demonstrates a
high degree of
accuracy for the
approximations and
the resulting
staffing levels.
It considers
uncertain arrival
rates and allows
abandonment from
the queue.
It determines the
staffing level that
maximizes the
expected profit of
an outsourced call
center subject to the
alternate SLAs.
MINUTES OF MEETING - 1
PIRADA P.
ALEX
HE ZHU
DIVYA
MINUTES OF MEETING - 2
PIRADA P.
ALEX
HE ZHU
DIVYA