| Research |
25.10.2007
Use of ‘Simulation’ as a Contact Centre ‘HR Management Tool’?
By Vijay Pereira MCMI, UK
Researcher and Lecturer, University of Portsmouth, UK
Contact centres are highly-complex environments. As a result of this complexity, the need to change and evolve is extremely important. ‘Simulation’, it is claimed allows contact centres to understand this complete customer handling process by duplicating all processes that affect the centre. Further, it enables contact centres predict results before making changes that can impact its bottom line. ‘Simulation’ is defined as the construction of a mathematical model to reproduce the characteristics of a phenomenon, system, or process, often using a computer, in order to infer information or solve problems.
Personally I see ‘Simulation’ as a ‘HR Management Tool’ in a contact centre, as a significant portion of the cost of operating a contact centre is staffing cost. As contact centres become larger and more complex, simulation modelling is becoming a required tool enabling users to manage growth and expenses. While the cost to procure contact centre simulation is initially high (as is evident from Jens Moeller’s article below dated 2 October 2007 ‘Call Center Simulation Tools – demystified’), the return on investment (ROI) that results from operational efficiencies would be fast if the tool is used effectively. Simulation tools would thus make contact centre operations far more effective. In today’s fast-paced contact centre industry, it is a tool that is becoming essential to management.
Contact centre HR departments often have a problem of minimising staffing costs in an inbound call centre, while maintaining an acceptable level of service in multiple time periods. The problem is complicated by the fact that staffing level in one time period can affect the service levels in subsequent periods. Moreover, staff schedules typically take the form of shifts covering several periods. Interactions between staffing levels in different time periods, as well as the impact of shift requirements on the staffing levels and cost should be considered in the planning. Traditional staffing methods based on ‘stationary queueing formulas’ do not take this into account.
This problem has received a great deal of attention in contact centre HR circles and so one
can reasonably ask why there is a need for a computational tool of this sort (Simulation). At this stage it would be convenient to determine the staffing process in a contact centre. The essential structure is sequential in nature and is as follows:
1. (Forecasting) Obtaining forecasts of customer load over the planning horizon, which is typically one or two weeks long. The horizon is usually broken into short periods that are typically between 15 minutes and 1 hour long.
2. (Work requirements) Determining the minimum number of agents needed during each period to ensure satisfactory customer service. Service is typically measured in terms of customer waiting times and/or abandonment rates in the queue.
3. (Shift construction) Selecting staff shifts that cover the requirements.
4. (Rostering) Allocating employees to the shifts.
Again as is evident from Jens article below, the market place is abuzz with various simulation solution providers providing software solutions in the name of ‘Contact/Call Centre Simulations’. From a HR point of view, these are aimed at solving some of the staffing or workforce problems faced by contact centres, especially when trying to discern how many operators are needed at any given time.
Such software’s it seems simulates staffing problems and enables the contact centre to try out solutions to their staffing problems. Their main purpose is to enable contact centres to find out exactly when a problem occurs, why it occurred and the best solution to use in the future. The HR departments of contact centres would obviously like answers to such situations, in order to staff correctly.
However it would be no good knowing how many staff/operators are needed if there has been no allowance made for operators having breaks. ‘Simulation’ providers claim to find answers to tricky situations like these as is shown in the following example:
Suppose the call rates increase from one level to another and a decision has to be made when to start another operator and for how long. Similarly there could be situations when operators take breaks, e.g. how long can they be away from their station before calls are delayed for too long, and how long the interval must be between one operator returning and the next one going off for their break. Providers claim their programs allows contact centres to run the simulation of increasing call volumes, calculate the call delay and show how many calls are unanswered. When another operator(s) is brought in, they can be seen to reduce the call delay. When the call delay reaches zero, all the calls are answered.
This model can be a useful tool and can be of a lot of help to the HR Manager and Operations Managers of contact centres when trying to determine how many operators have to be on duty because unlike equations to calculate average waiting times of customers the program will calculate individual waiting times quickly.
It seems simulators or formulas can calculate "Average Call Delay" and ‘Each Callers Delay’. This tool could be of a lot more use because in practise to calculate "Average Call Delay" is illogical. If the "Average Call Delay" is above zero then the operators can not handle the rate at which calls come in. Therefore you have two results, either the callers hang up and you loose their business, or the call delay tends to increase to infinity. However using available simulators contact centres can test different solutions to find the exact point when and for how long an extra operator is needed.
It seems certain providers claim their program is a stochastic (statistics involving or showing random/problematic/guesswork behaviour) model wherein the program can be run many times with the same parameters and it will produce different results each time to give you a range of results just as each day is different in ‘real life’. Use of this tool would thus achieve a better idea of the situation. Providers also claim that programs can also be used to test what would happen in
different situations, for instance, if an operator went on a break when everything was going smoothly.
Would everything be going smoothly ten minutes later? Hence as mentioned above by using the data available from the contact centres, a stochastic simulation model of the operations of a contact centre is created.
As was evident from the above example ‘Simulation’ can be used as an excellent ‘HR Management Tool’ in a contact centre to focus on their complex staffing needs and problems. Contact centres could thus be able to solve, or at least approximately solve, a number of moderately sized call centre staffing problems.
Author- Vijay Pereira is an experienced consultant and researcher. He is currently lecturing and researching at the University of Portsmouth, UK. He has over 15 years experience in consulting for the IT and BPO industry. He specialises in Human Resource Management. His current research interest lies in looking into the strategic role of HR post HR outsourcing. His overall experience has been gained in Industry, Consultancy and Academia in countries like India, UAE, Israel and the UK. Contact him at vijay.pereira@port.ac.uk .
|
16.10.2007
SYNERGISTIC MODELING OF CALL CENTER OPERATIONS (EXTRACT)
By DENNIS C. DIETZ AND JON G. VAVER
We synergistically apply queueing theory, integer programming, and stochastic simulation to determine an optimal staffing policy for a repair call handling center. A stationary Markovian queueing model is employed to determine minimal staffing levels for a sequence of time intervals with varying call volumes and mean handling times.
These staffing requirements populate an integer program model for determining the mix of call agent shifts that will achieve service quality standards at minimum cost. Since the analytical modeling requires simplifying assumptions, expected performance of the optimal staffing policy is evaluated using stochastic simulation.
Computational efficiency of the simulation is improved dramatically by employing the queueing model to generate an analytic control variate.
1. Introduction
Many commercial enterprises and public agencies operate centralized call centers to provide effective and responsive service for patrons. For example, communication service providers operate call centers to ensure timely restoration of service following equipment failures within the communications network or at the customer premises. The call center is staffed by repair service agents who are trained to effectively interact with the customer, diagnose the problem, and dispatch appropriate repair resources. Typically, the goal is to completely restore service within a few hours.
The operating cost of a repair call center is dominated by personnel expense, so the economic efficiency of the systemis determined almost entirely by the quality of the agent scheduling process. The scheduling problem is characterized by a highly variable demand pattern and a requirement to schedule agents in shifts that are constrained by labor rules.
Fortunately, the weekly demand profile is quite predictable and seasonally consistent. The fundamental challenge is to schedule agent shifts such that resulting agent availability will enable consistent achievement of a specified service level at minimum cost.
|
02.10.2007
Call Center Simulation Tools – demystified
By Jens Moeller
The USP of Call Center simulation software is that they remove the risks that go along with an ErlangC based planning. They are doing this while recording every contact, the number of call reasons and more of such sophisticated real time variables. They tend to pay off quicker in larger call centers with skill based routing challenges in environments with -at least partly- highly skilled call center agents.
With some 7000 visitors, Call Centre Expo in Birmingham/UK was probable more successful than last year. At least, I had the impression. This leading event in the UK was truly buzzing and I needed to give my walk along the stands some focus.
I always wondered why so many Call Centres simply follow the trial and error principles, when serious mistakes in planning and realising operations can be so costly. So I had a look at simulation tools.
What is it, anyway? Basically, you can check what kind of business impact you can expect when changing some variables with influence on the
workforce productivity. For instance, you may want to shift some call centre agents to another team for a while, which had been reduced lately due to intense and long training measures. You need to put them on a multi-skill team instead of the single-skill one. And you are wondering what the impact on your call centre operations will be before you run into any pitfalls and bad results.
First, the question is whether it pays off. A common business model is to just buy it rather than renting or leading models, because it will only be used on a couple of seats – those of forecasting and planning managers. Vendors seem to target the larger call centers assuming that only the big ones have the need to run simulations. In the UK, ballpark figures for purchasing a call centre simulation software vary from one vendor to another, ranging from 27 K £ to some 80-100 K £ at the moment, no matter how many seats the call centre organisation has got. Costs largely depend on the needed degree of integration and customisation.
These two components play a key role. One vendor likes to keep his solution as open as possible to any environment - which does not make the product look user friendly without customisation, but will be in the end after adaptions are made. Others prefer to make call center simulation as user friendly as possible out of the box, which may need less customisation effort if the customer goes with the standard - but which may cause even more professional services costs if the customer still needs changes.
Second, there are a lot of Workforce Management software tools - and usually they have a forecasting component in order to allow detailed staff planning and scheduling. So what would we need dedicated Call Center simulation and Call Center forecasting tools for, anyway? The answer is: because of their algorithm. I was told dedicated simulation tools contain more variables, so the prognosis would be more accurate. The technology behind this is called "Discrete Event Simulation", nothing new, but new to many call centre organisations. Forecasting traditionally relies on the ErlangC formula and many call centers have made an attempt to improve the forecasting quality using HillsB as well or even instead ErlangC.
Both make a number of assumptions to calculate the future inbound workload, which represent a risk of miscalculation. Some simulation tools go beyond this, eliminating ErlangC and the respective risks. Instead, they record every contact, the number of call reasons and more of such sophisticated real time variables, enabling the call center to estimate which skill groups are confronted with a too high or too low inbound volume. This concept, then, is ideal for skill based routing concepts, which are widely used in larger contact centers.
|
|
|
44
Users online 
