The introduction of the services sector
The services sector is gaining more and more importance in the international economy today. Irene et al., (1998) says that the services account for a growing percentage of the gross national output of most countries. As such, service industries are maturing and have become more competitive, and there is a growing need to increase efficiency, productivity and competitiveness (Wirtz et al., 1998). Adenso-Diaz et al. (2002) says that the growth of this sector makes the need for adequate management of service operations more and more apparent. Capacity planning is amongst nine fundamental research areas Johnston (1999) identifies in this field over the next few years. Slack et al., (2009) says, "Insufficient capacity leaves customers not served and excess capacity incurs increased costs". As such, capacity management is very important in services industries and it gains more and more importance in today's business environment.
Different authors have already tackled this from different viewpoints of staffing and scheduling (Goodale and Tunc, 1998; Thomson, 1996). However, according to Irene et al. (1998) there seems to be a divergence between what companies should do, according to academic literature, and what they are actually doing. In other words, the theory is not put into practise many times. They think this divergence is because the literature often ignores the role of strategy when dealing with capacity issues. Armisted and Clark, (1993) says that there is an interaction between capacity management, quality management and resource productivity or efficiency management. Rhyane (1987) says that capacity management has a considerable impact on the quality of service perceived by customers. Van Looy et al., (1998) and Adenso-Diaz (2002) it is very difficult and complex to achieve better capacity management without quality level being affected. It is apparent, that service capacity management is far more complex than manufacturing or logistics capacity management (Irene et al., 1998)).
In this paper, I will be evaluating different models or methods of capacity management and compare their practical usage. First, the minimum staff model proposed by Adenso-Diaz et al. (2002) that tries to find the minimum capacity level below which quality may be affected. Short-term resource allocation could be done on the basis of this model. The second, unused capacity model explains how to better utilise the unused capacity by taking the practises of different service organisations as basis and comparing it with literature.
Capacity management -a literature review of the existing strategies
Lovelock (1992) defines the capacity of a service as the highest possible amount of output that may be obtained in a specific period with a predefined level of staff, installations and equipment. "Capacity management is the ability to balance demand from customers and the capability of the service delivery system to satisfy the demand", (Armisted and Clark, 1993). This requires first to understand the nature of demand by forecasting and second to manage capacity to meet that demand. In simple terms, the aim is to minimise the customer waiting time and to avoid unused capacity but at the same time without affecting the quality of service provided. As said above, the number of service organisations is growing in many countries. This leads to increased competition and the firms are forced to increase their efficiency and productivity that requires adequate management of available capacity.
There are many writers and literature on how to cope with the demand and supply imbalances (Lee, 1989; Lovelock, 1988; Sasser, 1976). In general, there are three ways to deal with such variances as identified by Sasser (1976), the level capacity, chase demand and manage demands. Slack et al., (2009) calls them as 'pure' plans but suggests that in practise most organisations will use a mixture of them rather than sticking with a single plan.
Level capacity plan
The first method is to have a fixed capacity irrespective of the demand. This is a very simple strategy and if the demand is lower than the capacity, the extra idle capacity is wasted. When the demand is more than the capacity, it cannot meet the higher demand. "Level strategies are applicable when demand is more visible before the time of use and the service organizations can effectively tell customers to wait when demand cannot be satisfied, i.e. the service is valued by the customers and they are willing to wait" (Armisted and Clark, 1993).
The second strategy is to manage supply to demand. Irene et al. (1998) says that
In goods manufacturing or in logistics when the demand is low, the firms could continue its production to make and keep an inventory level for future demand. This will help to meet the high customer demand when capacity is lower than demand. However, in service delivery there is not the possibility of producing the complete service in advance of demand and holding it as an inventory since the services are perishable by nature (Armistead and Clark, 1993). In addition, services tend to keep additional capacity in anticipation of additional business. Most service sectors have strict capacity constraints. "Once the capacities are established, the cost of making any adjustment-renting new gates and airplanes or building new hotels -are quite high" (Kim et al., 2004). Moreover, if the customer demand is not met, then there is a high risk of losing customer base. So to serve as many customers as possible and to have competitive advantage companies prefer to keep the capacity at the maximum anticipated level (Irene et al., 1998). This also helps them to avoid implementation of the complicated capacity management techniques.
Problems in Capacity management of Service Sector
Adenso-Diaz et al. (2002) says that capacity management in service sector presents additional problems to those of manufacturing industries. He says that on the one hand the service firms are faced with a strong seasonality in demand. Kim et al., (2004) also notes that many service industries face considerable demand uncertainty and seasonal variations. "For instance, market demands typically are much higher during summer holidays and Christmas than during the rest of year" (Kim et al, 2004). On the other, the need for the customer to be actually present when the service is given is fundamental to many sectors. This personalised demand directly affects the quality of service offered.
In addition, services are perishable by nature and hence for each day those services are not put to profitable use, they cannot be saved (Bateson, 1977; Thomas, 1978). For example if an airline passenger seat is vacant on a particular journey, the airline company cannot use that on the next journey as an additional seat. This is applicable to most services. In addition, most services have strict capacity constraints. For example, an airline company or a hotel when reached its capacity it is very costly to make adjustments. To add another hotel or a flight requires huge investment.
Capacity management - minimum staff model
Duder and Rosenwein (2001) show that by using simple formulas to rearrange the number of staff in a 'call centre', it is possible to reduce the percentage of abandoned calls and increase the number of calls answered without waiting. This methodology also has been applied for postal services (Assad et al., 1998) and nursing services (Van Looy et al., 1998) where optimal staffing decisions are formulated. Adenso-Diaz et al. (2002) propose a modified model to determine the minimum staff numbers to provide coverage to various departments. In this model, they use historical data to determine the minimum staff required in the departments of a service.
In services, the capacity management is a very complex and difficult task. The failure to synchronize supply and demand, leads to a loss in opportunity to attend certain customers when demand is higher and to high costs due to the loss in income when demand is insufficient (Sasser, 1976). Another of the barriers in services is the problem of seasonal demand. Adenso-Diaz et al. (2002) says that one of the most important aspect in capacity planning is the human resource planning. This deals with assigning the right number of people at the right place and time. The Adenso model could be used to get an idea of the lower limits when a limited number of resources should be distributed in a number of departments. There is of course a trade-off between the productivity and quality. So, "the goal of our model is to establish minimum capacity levels below which quality may be affected, so as to be able to assign short-term resources on the basis of the demands that arise" (Adenso-Diaz et al. 2002).
According to the model, the following are the major steps in the process of determining minimum staff.
a) Determining the time to execute the tasks
The first step in the model is finding the real execution time, of each of the different tasks which are carried out on day-to-day basis in a particular service, as exactly as possible. The calculation should be done on a standard performance conditions and for a standard class of customer.
b) Calculation of average number of activities
Other basic data required is finding the average number of times all the different activities are carried out as a function of the type of customer. When the standard time varies highly according to the type of customer, the frequency calculated in this step is used for correction.
c) Calculation of theoretical staffing
For all the historical data that is available, a calculation is done to find what would have been the staffing level theoretically using a system for evaluating loads. This allows those responsible for the unit to determine the capacity needs.
d) Determination of the ratio
Comparing the relative staffing data with the actual staffing levels, calculate the ratio. This is by dividing actual data by theoretical data.
e) Consolidation of quality data and calculation of minimum staff
The model now compares the available quality to specific quality indices, recorded by the company from its experience. This information when combined with the ratio, we will get the following table.
When we group the ratio into different range of values, we will get the quality values associated with each range.
We could show this in a graph as shown below representing different quality indices for each range of ratios.
From this, we shall obtain a global quality function
f) Calculation of minimum staff for a mix of customers
So any particular day, the theoretical number of staff on any department may be determined as a function of the type of customer received as well as minimum admissible number of employees on the staff (multiplying this by acceptable ratio).
Application of the model into a hospital nursing service
Adenso-Diaz et al. (2002) applied the capacity management model in a hospital to find the outcomes. The hospital chosen is in Spain with 202 beds, 554 staffs and offers medico-surgical hospitalisation.
Step 1) The Delphi methodology was applied to determine the standard execution time of different hospital nursing tasks. A questionnaire was used to get the data from staff. The statistical values are calculated from these received data. The specific list of the estimated execution time of various tasks is created from this calculations.
Step 2) Taking the sample of patients as the basis, the number of activities normally dealt with were obtained as a function of the dependency of the patient (Type1: Autonomous, Type2:semi-independent, Type3: very dependent on nurse). A random sample of 236 patients were taken for this purpose.
Step 3 and 4) Taking the real historical data of patients and staff, ratios are calculated for every month. The following table shows this data
Step 5) Nine quality indices are already identified and implemented internally in the hospital, therapy sheets, nurses' notes, admission evaluation, discharge reports, phlebitis rate, fall rate, scab rate, care planning sheets and rate of infections. Out of this nine, the last two were not used for the purpose of the study. The others were of quantitative nature and not based on subjective perception. Staff planning is calculated as a function of the seven factors.
Call centre capacity management
"Call centres often experience large fluctuations in demand over relatively short periods of time. However, most centres also need to maintain short response times to the demand", (Betts et al., 2000). This places great emphasis on capacity management in call centre operations. The difficulty in dealing with capacity management of call centres is characterised by the demand seasonality that we saw earlier as common to any service industry. It is more evident here with call centres experiencing short-run spikes in demand and fluctuations in daily, monthly and yearly time horizons. Also as seen before, the capacity management have a strong influence in efficiency and quality of the service provided.
The capacity management in call centres is also very typical of the queuing environment, which is modelled in the management literature. The pattern with which calls arrives can be a complex variation of the Erlang delay problem (Brockmeyer et al., 1948). Erlang says that at low levels of resource utilisation, the randomness in demand creates backlogs or queues. The computation however is time consuming and complex. Kolesar and Green (1998) apply the Erlang model to call centres and use their approximations of the formulae to offer general guidance in capacity decisions. They made two basic propositions. First, aggregation of demand that says a small number of big call centres is better than large number of small centres. Second, to achieve customer satisfaction requires more staff and this might leads to increasing the staff levels even uneconomically. The figure below illustrates this point.
Heskett (1986) also said this point that service quality affects as the utilisation of resources is more. He said some estimates show that the quality of a service drops rapidly when demand exceeds even as low as 75 per cent of the service firm's capacity. But the problem with Kolesar and Green model (1998) is that they made some assumptions about the demand pattern to simplify the Erlang model. The assumptions are one, demand remains constant and two the system does not become too heavily congested. But in case of call centres both of these assumptions are questionable (Betts et al., 2000).
Since the primary task is to cope with the varying demand, the range and response flexibility in organisations is useful to study (Slack, 1995). In capacity management, range flexibility is defined as the capability of operations to move to much lower or higher output rate over a long period. Response flexibility is the ability move from one output rate to another without significant lag. Betts et al., (2000) proposes a new model for capacity management in call centres. The y use the range and response flexibilities to determine this model. The model is illustrated in the figure below.
Capacity management model for call centres
Source: Betts et al., 2000
The table below summarises the call centres which were included in the study. It is evident from the table that the managers already put some capacity management by adding flexibility to number of operators.
Source: Betts et al., 2000
For the purpose of the study, they used the quantitative data from each call centre that is recorded by the call centre operations. "Four month's data were used (86 separate data sets) to identify statistical links between forecasting accuracy (the independent variable) and four separate dependent variables, calls answered, percentage service levels, calls answered within target and calls abandoned", (Betts et al., 2000).
The table below summarises the results of the research. The data taken includes daily forecast demand, actual demand, staffing levels and service levels. The theoretical discussion identifies that high service levels imply low staff utilisation levels, thereby strengthening the views by Kolesar and Green (1998), and Heskett (1986).
Also it shows that the UK based call centres show very poor resource utilisations, less than 50% in all the cases and even up to 11% in some cases. But this is justified with the fact that high level of response targets, as much as 95%, were set for many of the call centres. Low levels of service might have helped in better staff utilisation.
Main sources of variation
The range and response in capacity management that we saw early is used now to determine the variation. The figure below shows how the different call centre sites were categorised for their range and response requirements. The results show that only two sites are low in range and response flexibility. These two sites (E and G) need both range and response flexibility. Also it shows that there is cluster of sites where response requirement is high.
Effect of forecasting
The analysis of the forecasting shown above suggests that many call centres have effective method of forecasting. "The analysis indicated a linear relationship between each of the sets of dependent and independent variables". The analysis also showed a high dependence between the performance measures and forecast approach. This is illustrated in the table below
Also, it is important to note that the linear fit does not support the idea that Erlang patterns (Brockmeyer et al., 1948) might create exponential pattern. Another important finding of the study is that some call centres had to limit the chase strategy (Sasser, 1976). This was because of various practical reasons. The centres not yet done any plans to reduce or shift the staff during lower demand periods.
The study also shows that
Goodale, J.C. and Tunc, E. (1998), "Tour scheduling with dynamic service rates", International journal of service industry management, Vol.9 No.3, pp. 226-47
Johnston, R. (1999), "Service operations management: return to roots" International journal of operations and production management, Vol.19 No.2, pp-104-24
Adenso-Diaz, B., Gonzalez-Torre, Pilar and Garcia, Virginia (2002), "A capacity management model in service industries", International Journal of Service Industry Management, Vol.13 No.3, pp.286-302
Lovelock, H.C. (1992), "Seeking synergy in service operations: seven things marketers need to know about operations", European Management Journal, Vol.10 No.1, March, pp. 22-9
If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please click on the link below to request removal: