Int. J. Services and Operations Management, Vol. X, No. Y, xxxx 1
Copyright 200x Inderscience Enterprises Ltd.
Integrating customer preferences and organisation strategy for resource allocation
Amit Sachan* Management Development Institute Mehaurali Road, Gurgaon 122001, India E-mail: [email protected] E-mail: [email protected] *Corresponding author
Subhash Datta Jaipuria Institute of Management 1, Bambala Institutional Area Sanganer, Jaipur 302033, Rajasthan, India Fax: 1412771334 E-mail: [email protected]
A.P. Arora Management Development Institute Mehaurali Road, Gurgaon 122001, India Fax: +911242341189 E-mail: [email protected]
Abstract: The objective of the current research is to develop a resource allocation plan based on customer preferences and organisational strategy. The Automatic Teller Machine (ATM) service of the bank is chosen as the service in the current research. Conjoint experiments were used to identify choice patterns of customers. Customer share maximising service designs were obtained under three generic strategies, i.e., cost leadership, differentiation and focus strategy. Simulation is used to determine the effect of changing the level of policy attributes of optimal service designs on customer share. The effect of concentration of the banks ATM on the banks X customer share is determined by simulation for the three strategies. The results show how users trade off between various ATM attributes when choosing an ATM service. Various service designs were obtained under each strategy and then various scenarios were obtained by simulation. The paper provides the approach which will help managers in effectively managing resources.
Keywords: resource allocation; service design; optimisation; simulation; conjoint.
Reference to this paper should be made as follows: Sachan, A., Datta, S. and Arora, A.P. (xxxx) Integrating customer preferences and organisation strategy for resource allocation, Int. J. Services and Operations Management, Vol. X, No. Y, pp.000000.
2 A. Sachan, S. Datta and A.P. Arora
Biographical notes: Amit Sachan is a Fellow of the Management Development Institute (MDI), Gurgaon, India. He graduated in Industrial Engineering from the Indian Institute of Technology (IIT), Roorkee, India. His research interests are the strategic uses of IT in services and manufacturing, service design and supply chain management. To date, he has published research papers and cases in refereed international journals and conference proceedings.
Dr. Subhash Datta is currently the Director of the Jaipuria Institute of Management, Jaipur. He has over 32 years of experience in teaching, training, research and consulting. Dr. Datta is a Life Member of the Operational Research Society of India and an International Fellow of Operational Research. In 1990, he received the Sir Charles Goodeve Gold Medal for the most outstanding paper of the year in the Journal of the Operational Research Society, UK. His research interest are in the following areas: production and operations management, DSS and decentralised planning and decision making.
Professor Ashok Pratap Arora, is a Fellow of the IIM Ahmedabad with over 25 years of teaching experience in institutes such as the IIM Calcutta, University of Bocconi, Milan, University of Torino, National Institute of Design, Ahmedabad, University of Allahabad, etc. He had been in the board of several organisations such as the Calcutta Stock Exchange, Indian Institute of Management Calcutta, etc. Professor Arora has also been consulted by over 30 organisations in India and abroad. His main interests are in the areas of cyber marketing, marketing research, consumer behaviour, marketing strategy and brand management.
1 Introduction
Resource allocation lies at the heart of the operation of systems. It is the careful choice of resources. This function is gaining even more importance as systems become more flexible in their arrangement and use of their resources. Resource allocation decisions are important because of their impact on a firms value (Merchant, 1997). Resource allocation embodies the choice of resources, their assignment and timing to perform operations while satisfying client, processing and capacity constraints. It is broader in its context, fulfilling both the planning and scheduling aspects. The planning aspect is in the choice of resources and their assignment, while the latter is the time ordering of operations and resources. Four types of resources are financial resources, physical resources, human resources and technological resources.
In past research on resource allocation in services (or in general), customer views are incorporated in the form of customer perception on service quality (Soteriou and Hadjinicola, 1999). Resource allocation in the past has not been done on the basis of customer preference. Panagiotis et al. (2002) have taken customer utility function in their theoretical model for resource allocation in integrated services connection-oriented networks. Chao et al. (2003) also developed a theoretical model in multisite service system with intersite customer flows. Until now no research has considered customer preferences and empirically tested the model. Further, no research in the past has considered strategic framework for resource allocation. The current research fills this gap by integrating customer preferences and strategy for resource allocation. Chandler (1962)
Integrating customer preferences and organisation strategy 3
mentioned that strategy is the determination of the basic long-term goals of an enterprise and the adoption of courses of action and the allocation of the resources necessary to carry out these goals. Swink and Way (1995) have mentioned that most of the strategy definitions discuss about the enhancement of the firms competitive position in the marketplace through resources building or positioning.
The objective of the current research is to develop resource allocation plan based on customer preferences and organisational strategy. This paper is divided into six sections. Section 2 presents the literature review on past research in resource allocation. Section 3 presents the resource allocation model that integrates customer preferences and organisation strategy. After that, research methodology is presented in Section 4. Section 5 discusses the results and the paper ends with conclusions in Section 6.
2 Literature review
Past researchers have presented many techniques for effective resource allocation in organisations. Xie et al. (2000) presented the fault tree analysis technique to improve the reliability of complex systems and to help in prioritising the improvement efforts and optimising resource allocation. Cheng and Li (2001) focused on the use of the Analytic Hierarchy Process (AHP) to prioritise different forms of information. Ranking of the information may provide additional insight in allocating scarce resources. The ranking of industrial projects using the AHP has been raised to help management in the efficient allocation of companies resources (Alidi, 1996). Greasley and Barlow (1998) presented an assessment of the use of simulation modelling to assist in business process reengineering projects. Juha and Mikko (2000) mentioned that formula-based allocation schemes are often proposed as a means of making the allocations of public funding more equitable, and more efficient. The Granot and Zuckerman (1991) model is developed, which would facilitate the analysis of activity sequencing and resource allocation in R&D projects. Lee et al. (2003) proposed a heuristic genetic algorithm for solving resource allocation problems. Leu et al. (1992) integrated three Artificial Intelligence (AI) techniques to model the resource allocation decision process.
Chao et al. (2003) studied the resource allocation problem in multisite service systems with intersite customer flows. They applied a cost-effective measure to construct a mathematical model for the resource allocation problem, i.e., how to allocate a limited amount of service capacity to different service sites so that a system-wide service quality measure is optimised. Chao et al. (2003) results show that the optimal resource allocation rule exhibits the structure of one large and many small, i.e., a healthcare system should have one large centre and many small facilities to cover the service regions. Lawler and Moore (1969) formulated a functional equation, similar to that for the knapsack problem, applying it to the solution of a problem of resource allocation in critical path scheduling, and to a variety of single-machine sequencing problems with deadlines and loss functions. Among these are: minimisation of the weighted number of tardy jobs, maximisation of weighted earliness, minimisation of tardiness with respect to common relative and absolute deadlines, and minimisation of weighted tardiness with respect to a common deadline. Jegadheesan et al. (2007), examined the Failure Mode and Effects
4 A. Sachan, S. Datta and A.P. Arora
Analysis (FMEA) implementation in the service industry and designed an improved model, named Modified service FMEA. The models implementation was examined in a transport company and pinpointed the seriousness of failures through the portrayal of Service Lost (SL) and Cost Lost (CL).
Past research in resource allocation took customers into account. Venkatesan and Kumar (2004) evaluated the usefulness of Customer Lifetime Value (CLV) as a metric for customer selection and marketing resource allocation by developing a dynamic framework that enables managers to maintain or improve customer relationships proactively through marketing contacts across various channels and to maximise CLV simultaneously. Panagiotis et al. (2002) presented an approach to the resource allocation problem in connection-oriented networks that offer multiple services to users. The objective of the optimisation problem is to determine the amount of, and required resources for, each type of service to maximise the sum of the users utilities. Users preferences are summarised by means of their utility functions, and each user is allowed to request more than one type of service. Soteriou and Hadjinicola (1999) presented a modelling framework that provides marketing and operations viewpoint for resource allocation. The framework was used at different stages of multistage service delivery system, where the managers goal is to improve service quality perception given some budget. The three factors which their resource allocation scheme is captured are:
1 current level of customer perception of service quality at each stage
2 the cost of implementing service quality improvement at each stage
3 the importance placed by customers on each stage.
Until now no research has considered customer preferences and empirically tested the model. Further, no research in the past has considered a strategic framework for resource allocation. The current research fills these gaps by integrating customer preferences and strategy for resource allocation. The next section presents the resource allocation model, which integrates customer preferences and organisation strategy.
3 Resource allocation model
The resource allocation model which integrates customer preferences and organisation strategy, with the objective to develop customer share maximising resource allocation is presented in Figure 1. The formulation builds on the choice patterns of customer. The customer share of the service offered is the function of customer choice patterns, service attributes offered by company and the service attributes offered by competitors. Figure 1 also shows that the cost is a function of the service attributes. The optimal service design1 takes into consideration customer preferences and develops customer share/profit maximising service design. After getting optimal service profiles, the research moves to the resource allocation2 in alignment with organisation strategy.3 The criteria for resource allocation is that the optimal level of resources that should be directed by the bank is a response function describing the relationship between the customer share of that service design and the investment made in resources for that design under particular strategy.
Integrating customer preferences and organisation strategy 5
Figure 1 Resource allocation by integrating customer preferences and organisation strategy
3.1 Assumption of the model
Customers ascribe measurable amounts of importance to different features of the product or service.
When called upon to choose between the alternatives, customers use an additive model for overall evaluation.
Customers are seeking to maximise their total utility.
4 Research methodology
For integrating customer preferences and organisation strategy for resource allocation, the model is tested in the banking industry and ATM service is chosen for research. Koutouvalas et al. (2005), revealed several factors that influence customer perceptions of service quality in the banking industry. Past researches on ATMs had given some attributes which customer look for while deciding about ATM services. Location and accessibility of ATMs, new functions and services provided through ATMs by Moutinho and Brownlie (1989), breakdown of ATMs by Howcroft (1991), security by El-Haddad and Almahmeed (1992), time and cost savings, greater control over the service delivery, reduced waiting time, a perceived higher level of customisation by Meuter and Bitner (1998), location by Kauffman and Lally (1994), and fun or enjoyment from using the technology by Dabholkar (1994; 1996). The relevant attributes for conjoint experiments
Company AService Attribute
CompetitorsService Attribute
Competitors UtilityCompanyA Utility
Customer ChoicePattern
Customer Share
Optimal Service Design
Resource Allocation
COST OrganisationStrategy
6 A. Sachan, S. Datta and A.P. Arora
were selected by reviewing the past researches on ATM service and by customer and manager interviews. The final list of attributes and their levels is given in Table 1. The total number of possible experiments that can be developed by these six attributes and their levels are (4*2*2*2*2*2 = 128). Fractional factorial design is used to reduce this number of design to 16.
Table 1 ATM attributes and their levels
S no. ATM attributes Levels
ATMs at markets only
ATMs at railway stations and bus stands only
ATMs at branches only
1 ATM location
ATMs in residential areas only
Only banking services 2 ATM functionality
Banking services + utility bill
ATM not working once in ten visits 3 ATM breakdown
ATM not working once in 50 visits
No fee for four transactions in year (provided you maintain 5,000 minimum balance in saving account)
4 Fees on cash withdrawal from other banks ATMs
Fee of 50 rupees per transaction
Security guard always available and entry by card 5 Security and privacy inside the ATMs Entry by card
Limit on single as well as monthly withdrawal 6 Limit on cash withdrawal
Can withdraw up to the balance in your account
Data is collected from the customer by asking them to rank these 16 ATM designs. Conjoint Analyser is used to analyse the ranking data (ranks of 16 ATM designs). The output of the conjoint analyser is the part worth utilities of levels and relative importance of each attribute. Once a conjoint output is obtained, a nonlinear programming model is developed, which maximises the customer share considering the cost and organisational constraint (see Appendix 1). For evaluation of alternate strategy, the three generic strategy by Porter (1980) are used. Porter posits three generic competitive strategies that confer a sustainable competitive advantage for the firm. The three generic strategies are cost leadership, differentiation and focus. For resource allocation various service designs were obtained from nonlinear programming under three generic strategies, which were (1) cost leadership, differentiation and focus strategy, (2) considering customer preferences, and (3) cost and organisational constraints. In identifying the market segment for focus strategy, cluster analysis was used for market segmentation. These strategies are operationalised in the current research as:
1 In cost leadership strategy, the variable cost per ATM per month is fixed to Rs 32,000.
2 In differentiation strategy, the service designs are derived which would give the maximum customer share, keeping variable cost per ATM as a nonconstraint by nonlinear programming.
Integrating customer preferences and organisation strategy 7
3 In focus strategy, the service designs are derived which would give the maximum marker share in two clusters. Here the variable cost is taken into consideration.
In the above three strategies, customer share-maximising optimal designs were obtained. Simulation is also used to see the effect on customer share of these service designs by changing the levels of four attributes (ATM functionality, breakdown, fees on cash withdrawal and limit on withdrawal), which are mainly policy decisions. After that simulation is used to see the effect of these service designs on customer share by changing the concentration of Bank X ATM in the market. Virupaxi and Biswajit (2007), modelled and simulated current practices at the service centres to analyse and understand the drawbacks. They identified various configurations by simulation and discussed them highlighting the necessity of capacity addition, customer convenience and service quality.
The data for this work was collected from a public sector bank (Bank X). Customer data was collected in two phases. First was the interview phase where some qualitative data was collected for selecting the relevant attributes for measuring customer preferences. The second phase involved data collection on the questionnaire based on conjoint experiments. In both the phases, customer data were collected from the Delhi region. Since Delhi is the capital of India, almost all the banks operating in India have branches and ATMs in that region. The customers in this region are aware of their bank ATM service as well as other bank ATMs. And since the objective of the research was to do resource allocation for the future, the Delhi region came out to be a good option because the customers here are mature enough and have seen the advantages and disadvantages of various service configurations. Bank X has 144 branches in Delhi, divided into two regions: North Delhi and South Delhi. Fifteen branches were selected randomly in both the regions. Data was collected by customers in the bank branches. First, they were asked if they were using ATM or not, if the response was yes, then they were requested to fill up the questionnaire. The next section presents the results and discussions of the research.
5 Results and discussion
The conjoint analysis output of customers gave the relative importance of attributes and part worth utility of levels of ATM service. The conjoint analysis revealed that users of ATM were most influenced by ATM breakdown then by ATM locations, then by ATM security and lastly by ATM functionality (Figure 2). The lowest relative importance is on the attribute Limit on cash withdrawal. The relative importance of the attribute shows the weight a customer gives to these attributes. The levels having positive utility were ATM at markets, ATM at residential areas, Banking services + Utility bill, ATM not working once in fifty visits, No fee for four transactions in a year, Security guard always available and entry by card and Limit on cash withdrawal.
For resource allocation, the service designs are obtained in all three strategies. Then the effect of changing the level of attributes on customer share in all three strategies is seen and lastly, the effect of changing the concentration of Bank Xs ATM in market on customer share in all three strategies is observed.
8 A. Sachan, S. Datta and A.P. Arora
Figure 2 Relative importance and part worth utilities of customers
5.1 Cost leadership
In the cost leadership strategy, cost reduction becomes the major theme. In this strategy, the upper limit of cost is fixed to Rs 32,000 per month per ATM, and the optimal service design was obtained by putting various organisation constraints (refer to Table 1 and Appendix 2). The first design which one gets without organisation constraint is SD1. Bank Xs customer share (59.7%) for this design is maximum. But this design is not feasible because putting all the ATMs in the residential area is not a feasible proposition. So, various organisation constraints are put to get the various designs. Organisation constraints are put by the Equations (18)(22) in nonlinear programming formulation. Figure 3 shows Bank Xs customer share for the various service designs. Service designs SD1, SD2, SD3, SD5, SD6, SD11 and SD12 are the designs which provide Bank X more than 30% customer share under the given cost and organisation constraints.
Figure 3 Bank Xs customer share for service designs under the cost leadership strategy
24.79
16.75
28.82
9.37
19.22
1.0505
101520253035
Location Functionality Operational Fees on cashwithdraw
Security Limit on cashwithdrawal
0
1
2
3
M S B RA
00.5
11.5
2
OB BU0
2
4
10 500
0.5
1
No fee Fee of 500123
SG Card.0
0.050.1
0.15
Limit No
0
10
20
30
40
50
60
70
SD1 SD2 SD3 SD4 SD5 SD6 SD12 SD7 SD8 SD9 SD10 SD11
Cus
tom
er sh
are
Integrating customer preferences and organisation strategy 9
5.2 Differentiation
In the differentiation strategy the service designs are derived that would give the maximum customer share, keeping variable cost per ATM as a nonconstraint by nonlinear programming. The optimal service design was obtained by putting various organisational constraint (see Table 2 and Appendix 2). It is important to differentiate the service based on service attributes because as Moutinho and Meidan (1989) mentioned, it will become more difficult for customers to differentiate the services offered by one bank from those offered by another, as any technological advantage gained will be short-lived. The first design is one which has maximum customer share for Bank X, the cost of this design is Rs 44,000 per ATM per month. The customer share of Bank X for this service design was 69%. Other service designs were obtained by putting various organisation constraints. Organisation constraints are put by the Equations (18)(22) in nonlinear programming formulation. Figure 4 shows Bank Xs customer share for various service designs. Except SD2, SD6, SD8 and SD9, all the designs provide Bank X more than 55% customer share under the given organisation constraints. Only SD6 provide Bank X a customer share of 40%; this shows the effect of increasing the ATM in branches.
Figure 4 Bank Xs customer share for service designs under the differentiation strategy
5.3 Focus strategy
In the focus strategy, the company targets the narrow segment of the market. A market segmentation analysis was conducted, using all attribute ratings, to ascertain if clearly defined clusters of consumers exist or not. Cluster Analysis is used to segment the market based on part worth utility of each customer. Table 2 presents the average utilities of attributes levels and number of customers in the four clusters. Cluster 2 has maximum number of customers, so the service designs are obtained, which would give the maximum customer share in Cluster 2.
0
10
20
30
40
50
60
70
SD1 SD2 SD3 SD4 SD5 SD6 SD12 SD7 SD8 SD9 SD10 SD11
Cus
tom
er sh
are
10 A. Sachan, S. Datta and A.P. Arora
Table 2 Average utility of levels in four clusters
Attribute levels Cluster 1 (n = 28)
Cluster 2 (n = 150)
Cluster 3 (n = 57)
Cluster 4 (n = 48)
ATM at markets 3.80 0.02 0.19 4.06
ATM at railways 1.98 1.13 1.52 3.21
ATM at branches 1.49 0.44 4.01 2.28
ATM at residential areas 4.29 0.66 5.33 1.42
Only banking services 0.34 0.97 0.22 1.48
Banking + utility services 0.34 0.97 0.22 1.48
Breakdown once in ten visits 0.46 2.23 0.83 0.32
Breakdown once in 50 visits 0.46 2.23 0.83 0.32
No fees 0.32 0.49 0.20 0.82
Fee of Rs 50/ 0.32 0.49 0.20 0.82
Card only 0.82 1.41 0.56 0.16
Card + security guard 0.82 1.41 0.56 0.16
Withdrawal limit of Rs 15,000 0.03 0.03 0.17 0.03
Withdrawal up to balance 0.03 0.03 0.17 0.03
Here, the variable cost is kept fixed to Rs 35,000 per month per ATM. Then customer share maximising service designs are derived by nonlinear programming. Various organisational constraints were put to get the service designs by nonlinear programming. Organisation constraints are put by the Equations (18)(22) in nonlinear programming formulation. Figure 5 shows Bank Xs customer share for various service designs. Service designs SD1, SD3, SD4, SD5, SD6, SD7, SD8, SD9, SD10 and SD12 are the designs which provide Bank X more than 40% customer share under the given cost and organisation constraints. SD1, SD3 and SD6 are the designs for which the customer share is more than 50%.
Figure 5 Bank Xs customer share for service designs under the focus strategy
0
10
20
30
40
50
60
SD1 SD2 SD3 SD4 SD5 SD6 SD7 SD8 SD9 SD10 SD11
Cus
tom
er sh
are
Integrating customer preferences and organisation strategy 11
5.4 Effect of changing the level of attributes in all three strategies
The optimisation model automatically takes the attribute level, which provides maximum utility given the cost and organisational constraints. For attributes that are not in the cost function but are more of policy decisions, the model always considers the level with maximum utility. So it is important to understand the effect of changing levels of those attributes. The attributes are ATM functionality, breakdown, fees on cash withdrawal and limit on withdrawal. This effect is found by simulation, and the results of service design obtained under each strategy as shown in Figure 6 presents the result for the cost leadership strategy. A similar pattern was obtained in the differentiation strategy and focus strategy. Maximum reduction in customer share is observed when the level of attribute ATM breakdown is changed (Figure 6). ATM breakdown is the attribute to which customers have given maximum relative importance. After ATM breakdown, attribute ATM functionality is second in terms of relative importance of these four attributes. The reduction in customer share is also at second place after ATM breakdown. A similar trend is observed in the attributes fees on cash withdrawal and limit on cash withdrawal. Similar trend was observed in the other two strategies.
Figure 6 Effect of changing the level of attributes on Bank Xs customer share for service designs under the cost leadership strategy
5.5 Effect of changing the concentration of Bank Xs ATMs
The effect of the concentration of Bank Xs ATMs in the market on Bank Xs customer share is shown in Figure 7 for the differentiation strategy. Past research documents the existence of an S-shaped growth model relating outlet expansion and customer share gains (Naert and Bultez, 1975; Lilien and Rao, 1976). Implicit in the S-shaped growth model is the concept of saturation, where an increase in outlet produces decreasing returns to customer share. For the lower percentage of Bank Xs ATM, the Bank Xs customer share is close to zero (or it increases at a diminishing rate) in all the three
0
10
20
30
40
50
60
70
SD1 SD2 SD3 SD4 SD5 SD6 SD12 SD7 SD8 SD9 SD10 SD11
Cus
tom
er sh
are
CS Normal of ATM functionality of ATM breakdown of fees of limit on withdrawal
12 A. Sachan, S. Datta and A.P. Arora
strategies and then as the percentage of Bank Xs ATM increases from 30% to 40%, the customer share increases and then it becomes constant again or increases at a diminishing rate. The graphs are asymptotic at the ends for the two other strategies. This graph is in agreement with the distribution network literature. A similar trend was observed in the two other strategies.
Figure 7 Effect of changing concentration of Bank Xs ATMs on Bank Xs customer share under the differentiation strategy
6 Conclusion
The objective of the current research was to develop a resource allocation plan based on customer preferences and organisational strategy. Customer preferences in terms of part worth utility were obtained from conjoint analysis. The three strategies: cost leadership; differentiation strategy and focus strategy as given by Porter (1980) were considered for resource allocation. In the cost leadership strategy, cost constraint was kept fixed at Rs 32,000. In the differentiation strategy, cost is not kept as a constraint for obtaining service design while organisational constraints are used. In the focus strategy, first cluster analysis is used for identifying the segment where the bank should focus. And then under cost and organisational constraint, service designs were obtained under nonlinear programming. The results show that on average, maximum customer share is in the differentiation strategy. The customer share of Bank X in the differentiation strategy for most of the designs is above 50%. On an average, the lowest customer share was in the cost leadership strategy. And customer share in the focus strategy was in the middle of those in cost leadership and differentiation strategies. These results provide an important insight to managers about which strategy to take depending on their budget and organisation policies.
0102030405060708090
25.9% 28.6% 31.0% 33.3% 35.5% 37.5% 39.4%
Bank X ATM concentration
Ban
k X
cus
tom
er sh
are
SD1 SD2 SD3 SD4 SD5 SD6 SD7 SD8 SD9
SD10 SD11 SD12
Integrating customer preferences and organisation strategy 13
Further, Bank Xs customer share is determined by changing the attributes, which are policy decisions for all the service designs obtained under the three strategies. This analysis was carried out by running a simulation. This analysis showed that the ATM breakdown attribute has maximum impact on customer share followed by functionality, followed by fees on cash withdrawal from other bank ATM and lastly, limit on cash withdrawal. The impact of changing the attributes, which are policy decisions, is very helpful to managers because through this, they would know the impact of each attribute on customer share under each strategy. They can now deploy resources accordingly. The impact of concentration of the banks ATM on its customer share is also determined for three strategies. There it is observed that for lower percentage of Bank Xs ATM the Bank Xs customer share is close to zero (or it increases at a diminishing rate) in all the three strategies and then as the percentage of Bank Xs ATM increases from 30% to 40%, the customer share increases at an increasing rate and then it becomes constant again or increases at a diminishing rate. This shape resembles the S-shaped growth model of relating outlet expansion and customer share gains suggested in past research. This analysis will help managers in taking decisions about the number of ATMs bank should go for in future. This study contributes to resource allocation literature by considering customer preferences and organisation strategy for resource allocation. The managerial implication of this research is that it provides several scenarios for resource allocation in organisations.
The limitations of the current research are first, nonlinear programming developed in the current research is not tested for complexity theories. Future research should test the nonlinear programming formulation for complexity theories. Second, in the current research, the nonlinear programming model is run on the data of 200 customers for optimisation because of the limitation of the available Lindos Whats Best software. The commercial version of Lindos Whats Best software can take only 200 nonlinear variables. Future research should run the formulation on a higher version of software, which can take more number of customers. Third, simulation is used for resource allocation. Future research should consider more deterministic approaches for resource allocation.
References
Alidi, A.S. (1996) Use of the analytic hierarchy process to measure the initial viability of industrial projects, International Journal of Project Management, Vol. 14, No. 4, pp.205208.
Chandler, A.D. (1962) Strategy and Structure: Chapters in the History of American Industrial Enterprise, Cambridge, MA: MIT Press.
Chao, X., Liu, L. and Zheng, S. (2003) Resource allocation in multisite service systems with intersite customer flows, Management Science, Vol. 49, No. 12, pp.17391752.
Cheng, E.W.L. and Li, H. (2001) Information priority-setting for better resource allocation using Analytic Hierarchy Process (AHP), Information Management & Computer Security, Vol. 9, No. 2, pp.6170.
Dabholkar, P.A. (1994) Incorporating choice into an attitudinal framework: analyzing models of mental comparison processes, Journal of Consumer Research, Vol. 21, No. 1, pp.100118.
Dabholkar, P.A. (1996) Consumer evaluations of new technology-based self-service options: an investigation of alternative models of service quality, International Journal of Research in Marketing, Vol. 13, No. 1, pp.2951.
14 A. Sachan, S. Datta and A.P. Arora
El-Haddad, A. and Almahmeed, A.M. (1992) ATM banking behaviour in Kuwait: a consumer survey, International Journal of Bank Marketing, Vol. 10, No. 3, pp.2532.
Granot, D. and Zuckerman, D. (1991) Optimal sequencing and resource allocation in research and development projects, Management Science, Vol. 37, No. 2, pp.140156.
Greasley, A. and Barlow, S. (1998) Using simulation modelling for BPR: resource allocation in a police custody process, International Journal of Operations & Production Management, Vol. 18, Nos. 910, pp.978988.
Howcroft, B. (1991) Customer service in selected branches of a UK clearing bank: a pilot study, in the service-manufacturing divide: synergies and dilemmas, Proceedings of the Service Industries Management Research Unit Conference, Cardiff Business School, University of Wales College of Cardiff, 2526 September.
Jegadheesan, C., Arunachalam, V.P., Devadasan, S.R. and Srinivasan, P.S.S. (2007) Design and development of modified service failure mode and effects analysis model, International Journal of Services and Operations Management, Vol. 3, No. 1, pp.111126.
Juha, A.M. and Mikko, S.A. (2000) Merit rating and formula-based resource allocation, The International Journal of Educational Management, Vol. 14, No. 3, pp.95100.
Kauffman, R.J. and Lally, L. (1994) A value platform analysis perspective on customer access information technology, Decision Sciences, Vol. 25, Nos. 56, pp.767794.
Koutouvalas, D.K., Siomkos, G.J. and Mylonakis, J. (2005) Perceived service quality management and loyalty in public versus private banks operations: some empirical evidence, International Journal of Services and Operations Management, Vol. 1, No. 2, pp.101122.
Lawler, E.L. and Moore, J.M. (1969) A functional equation and its application to resource allocation and sequencing problems, Management Science, Vol. 16, No. 1, pp.7784.
Lee, Z-J., Su, S-F., Lee, C-Y. and Hung, Y-S. (2003) A heuristic genetic algorithm for solving resource allocation problems, Knowledge and Information Systems, Vol. 5, No. 4, pp.503511.
Leu, Y-Y., Rakes, T.R., Rees, L.P. and Ceccucci, W.A. (1992) Modelling resource allocation in a decentralized organization with an AI-based, goal directive model, Decision Sciences, Vol. 23, No. 5, pp.10271049.
Lilien, G.L. and Rao, A. (1976) A model for allocating retail outlet building resources across market areas, Operations Research, Vol. 24, No. 1, pp.114.
Merchant, K. (1997) Accounting performance measures and the Myopia problem (Chapter 12), Modern Management Control Systems Text and Cases, Upper Saddle River, NJ: Prentice Hall.
Meuter, M.L. and Bitner, M.J. (1998) Self-service technologies: extending service frameworks and identifying issues for research, in D. Grewal and C. Pechman (Eds.) Marketing Theory and Applications, Chicago, IL: American Marketing Association, Vol. 9, pp.1219.
Moutinho, L. and Brownlie, D.T. (1989) Customer satisfaction with bank services: a multidimensional space analysis, International Journal of Bank Marketing, Vol. 7, No. 5, pp.2327.
Moutinho, L. and Meidan, A. (1989) Bank customers perceptions innovations and new technology, International Journal of Bank Marketing, Vol. 7, No. 2, pp.2227.
Naert, P. and Bultez, A. (1975) A model of a distribution network aggregate performance, Management Science, Vol. 21, No. 6, pp.11021112.
Panagiotis, T., Teneketzis, D. and Mackie-Mason, J.K. (2002) Market-based approach to optimal resource allocation in integrated-services connection-oriented networks, Operations Research, Vol. 50, No. 4, pp.603616.
Porter, M.E. (1980) Competitive Strategy, New York, NY: Free Press.
Soteriou, C.A. and Hadjinicola, C.G. (1999) Resource allocation to improve service quality perceptions in multistage service systems, Production and Operations Management, Vol. 8, No. 3, pp.221238.
Integrating customer preferences and organisation strategy 15
Swink, M. and Way, M.H. (1995) Manufacturing strategy: propositions, current research, renewed directions, International Journal of Operations & Production Management, Vol. 15, No. 7, pp.426.
Venkatesan, R.K. and Kumar, V. (2004) A customer lifetime value framework for customer selection and resource allocation strategy, Journal of Marketing, Vol. 68, No. 4, pp.106125.
Virupaxi, B. and Biswajit, M. (2007) Exploring the operational strategies for two-wheeler service centres using discrete-event simulation, International Journal of Services and Operations Management, Vol. 3, No. 1, pp.7494.
Xie, M., Tan, K.C., Goh, K.H. and Huang, X.R. (2000) Optimum prioritisation and resource allocation based on fault tree analysis, International Journal of Quality & Reliability Management, Vol. 17, No. 2, pp.189199.
Notes
1 Service design encompasses the roles of the people (service providers), technology, physical facilities, equipment, and the specific processes by which the service is created and delivered.
2 Resource allocation embodies the choice of resources and their assignment while satisfying organisation constraints.
3 Strategy is the search for a favourable competitive position in an industry, the fundamental arena, in which competition occurs. Strategy aims to establish the profitable and sustainable position against the forces that determine the industry competition.
16 A. Sachan, S. Datta and A.P. Arora
Appendix 1
Nonlinear programming model
The ATM service of the bank is chosen in the current research, whose objective is to attract new customer or retain existing customer. So the objective function in the current research is to maximise the customer share of the organisation. The overview of optimal service design approach developed in this research was presented earlier in Figure 1. Customer share is the function of the number of customers using the product/service of that organisation and competitors (Equation (3.7)). Suppose there are three organisations x, y and z and Tx is the number of customers using the services of organisation x; similarly Ty and Tz can be calculated. For computing the value of Tx, Ty and Tz (see Equation (1) for Tx) it is assumed that customer i will use x company product when the Total Utility of x company product (Uxi) is more than the Total Utility of y company product (Uy) and Total Utility of z company product (Uz). Once the number of customer in each organisation is known, customer share can be calculated as follows:
x
x y z
TMarket Share
T T T= + + (1)
{ xixi
1 if U ( , )0 if U ( , )
1
xi yi zi
xi yi zi
i NMax U U andU
x i i Max U U andUi
T y where y= =
=
= = (2) where:
N = total number of customers
Uxi = utility of x company for customer i
Uyi = utility of y company for customer i
Uzi = utility of z company for customer i.
The utility of the product given by the company is the function of customer choice pattern and the product attributes offered by the organisation. In current research, the ATM service attributes and levels are given in Table 1. Since there are 14 levels, the output of the optimisation programme should be the 14 decision variables x1 to x14 . The value of these variables will give rise to the utility of each customer, which in turn will decide the value of total number of customers in Bank X and by that customer share can be determined. These decision variables x1 to x14 also affect the cost. In the current research for optimisation, the variable costs are considered. Variable costs are rent cost of the particular location, IT service provider cost, security guard cost and depreciation. Fixed costs such as cost of machines, installation cost and physical infrastructure are not considered in the current research because these costs are almost the same for all the competitors. Further, because the number of ATMs does not significantly vary across competition in the region selected for the study, fixed cost owing to number of ATMs is not taken as a constraint in nonlinear programming for getting optimal service designs.
Variable Cost = Location Rent Cost + IT Service Provider Cost + Security Guard Cost + Depreciation. (3)
Integrating customer preferences and organisation strategy 17
The rent of ATM depends on the location of the ATM. For ATM in the market the rent is Rs 25,000 on an average, for residential area the rent is Rs 5,000 and for railway station the rent is Rs 30,000. The depreciation is charged as 9.75% per annum on machine and 3.00% per annum on the room. The cost of ATM is Rs 4, 30, 000 plus 12.00% taxes. Room construction takes about Rs 3, 00, 000. The depreciation per ATM comes to about Rs 4,400. The IT service provider cost and security guard cost is the same for every location. The cost of IT service provider is 12,000 per month for any location. Security guard cost around Rs 10,600 per month for any location. The minimum cost per ATM is Rs 16,400. The ATM design corresponding to this cost is ATM at Branches, without the security guard and one level each from the rest of the four attributes. The cost in this design is of service provider cost and depreciation.
Table 1 Cost of attribute levels (per ATM per month)
S no. Level Cost (INR)
1 ATM in market 25,000
2 ATM in residential area 5,000
3 ATM in railway station 30,000
4 Depreciation 4,400
5 IT service provider 12,000
6 Security guard 10,600
Table 2 ATM profile of banks for determining customer share
Bank X Bank 2 Bank 3
x1% ATMs at markets only
x2% ATMs at railway stations and bus stands only
x3% ATMs at branches only
x4% ATMs in residential areas only
x5% Only banking services
x6% Banking services + utility bill
x7% ATM not working once in ten visits
x8% ATM not working once in 50 visits
x9% No fee for four transactions in year (provided you maintain 5,000 minimum balance in saving account)
x10% Fee of 50 rupees per transaction
x11% Security guard always available and entry by card
x12% Entry by card
x13% Limit on single as well as monthly withdrawal
x14% Can withdraw up to the balance in your account
40% ATMs at markets only
2% ATMs at railway stations and bus stands only
40% ATMs at branches only
18% ATMs in residential areas only
100% Banking services + utility bill
100% ATM not working once in 50 visits
100% No fee for four transactions in year
100% Security guard always available and entry by card
100% Limit on single as well as monthly withdrawal
35% ATMs at markets only
10% ATMs at railway stations and bus stands only
20% ATMs at branches only
35% ATMs in residential areas only
100% Banking services + utility bill
100% ATM not working once in 50 visits
100% No fee for four transactions in year
100% Security guard always available and entry by card
100% Can withdraw up to the balance in your account
18 A. Sachan, S. Datta and A.P. Arora
The nonlinear programming model was used to model customer preferences cost and organisation constraint. The nonlinear optimisation formulation (Equations 410) provides a measure of the customer share, Equation (11) is of cost. The objective of the formulation is to maximise the customer share of Bank X (Equation 4). Equations (12)(17) are for controlling that the summation of ATM in each level of attribute should be equal to the total number of ATM. The output of the model is in percentages so in Equations (12)(17) the value on the right-hand side is 1.
Equations (18)(22) are various organisation constraints. Each constraint has at least or at most options which are later explored while running the model for getting service design under various constraints. The optimisation model in the current research assumed that all six attributes range across their levels i.e., location attribute ranges. For customer share calculations, the ATM profile for the two other banks (Bank 1 and Bank 2) are fixed as shown in Table 2 in the Appendix.
BankX
Bank X B C
TMaximise
T T T+ + (4)
{ Bank XiBank Xi283 1 if U ( , )0 if U ( , )1
Bank Xi bi ci
Bank Xi bi ci
iMax U U andU
Bank X i i Max U U andUi
T y where y= =
=
= = (5) { BiBi283 1 if U ( , )0 if U ( , )
1
Bank Xi bi ci
Bank Xi bi ci
iMax U U andU
B i i Max U U andUi
T y where y= =
=
= = (6) { CiCi283 1 if U ( , )0 if U ( , )
1
Bank Xi bi ci
Bank Xi bi ci
iMax U U andU
C i i Max U U andUi
T y where y= =
=
= = (7) UXi = x1ULMi + x2ULRi + x3ULBi + x4ULREi + x5USBi + x6USBUi + x7UBTi + x8UBFi + x9UFNi + x10UFYi + x11USCi + x12USGi + x13ULYi + x14ULNi (8)
UBi = 0.40 * ULM + 0.02 * ULR + 0.40 * ULB + 0.18 * ULRE + USBU + UBF + UFN + USCG + ULY (9)
UCi = 0.35 * ULM + 0.10 * ULR + 0.20 * ULB + 0.35 * ULRE + USBU + UBF + UFN + USCG + ULN (10)
(RATMM +SATMM) x1 + (RATMS + SATMS) x2 + SATMB x3 + (RATMR + SATMR) x4 + CSGx12 + D C (11)
x1+x2+x3+x4 = 1 Locations (12)
x5+x6 = 1 Services (13)
x7+x8 = 1 Breakdown (14)
x9+x10 = 1 Fees (15)
x11+x12 = 1 Security (16)
x13+x14 = 1 Withdrawal limit (17)
x1 > or < C1 at least or at most C1% ATM at Market (18)
Please verify the numbering of the equations.
TBankX or TBank X?
Integrating customer preferences and organisation strategy 19
x2 > or < C2 at least or at most C2% ATM at railway stations (19)
x3 > or < C3 at least or at most C3% ATM at Branches (20)
x4 > or < C4 at least or at most C4% ATM at residential area (21)
x12 > or < C5 at least or at most C5% ATM will have security guard. (22)
All xi > = 0
where:
ULMi = Utility of ATM at markets for customer i
ULRi = Utility of ATM at railway stations for customer i
ULBi = Utility of ATM at branches for customer i
ULREi = Utility of ATM at residential areas for customer i
USBi = Utility of ATM providing only banking services for customer i
USBUi = Utility of ATM providing only banking services + utility bills for customer i
UBTi = Utility of ATM with breakdown once in ten visits for customer i
UBFi = Utility of ATM with breakdown once in 50 visits for customer i
UFNi = Utility of ATM which has No Fees on Cash withdraw for customer i
UFYi = Utility of ATM which has Fees on Cash withdraw for customer i
USCi = Utility of ATM which allows entry by card for customer i
USGi = Utility of ATM which has entry by card plus security guard for customer i
ULYi = Utility of ATM where there is limit on cash withdrawal for customer i
ULNi = Utility of ATM where one can withdraw up to balance for customer i
RATMM = Rent of ATM at markets
RATMR = Rent of ATM at residential areas
RATMS = Rent of ATM at railway stations
CSG = Cost of Security Guard
SATMM = Service provider cost of ATM at markets
SATMR = Service provider cost of ATM at residential areas
SATMS = Service provider cost of ATM at railway stations
SATMB = Service provider cost of ATM at branches
D = Depreciation
C = Cost
Tx = Number of customer using Bank Xs ATM
20 A. Sachan, S. Datta and A.P. Arora
Tb = Number of customer using Bank 2 ATM
Tc = Number of customer using Bank 3 ATM
UXi = Utility of Bank X for customer i
UBi = Utility of Bank 2 for customer i
UCi = Utility of Bank 3 for customer i
x1 = Percentages of ATM at markets
x2 = Percentages of ATM at railway stations
x3 = Percentages of ATM at branches
x4 = Percentages of ATM at residential areas
x5 = Percentages of ATM providing only banking services
x6 = Percentages of ATM providing only banking services + utility bills
x7 = Percentages of ATM in which breakdown is once in ten visits
x8 = Percentages of ATM in which breakdown is once in 50 visits
x9 = Percentages of ATM which have No Fees on Cash withdraw
x10 = Percentages of ATM which have Fees on Cash withdraw
x11 = Percentages of ATM which allows entry by card
x12 = Percentages of ATM which has entry by card plus security guard
x13 = Percentages of ATM where there is limit on cash withdrawal
x14 = Percentages of ATM where one can withdraw up to balance.
The nonlinear optimisation problem was solved using the Whats Best!, an add-in to Microsoft Excel. Whats Best! allows one to build large-scale optimisation models in a free-form layout within a spreadsheet. Whats Best! combines the proven power of linear, nonlinear and integer optimisation with Microsoft Excel. Whats Best! supports many of the mathematical and logical functions of Excel and also adds some functions of its own.
Integrating customer preferences and organisation strategy 21
Appendix 2
Table 1 Cost leadership strategy
SD12
35
32
30
2
20
48
100
100
100
49
51
100
32(m
)
.3(l
)
.02(
l)
.2(l
)
NA
NA
SD11
47
32
30
2
0
68
100
100
100
39
61
100
32(m
)
.3(l
)
.02(
l)
NA
.3(l
)
NA
SD10
21
32
12
2
56
30
100
100
100
100
100
32(m
)
.3(m
)
.02(
l)
.2(l
)
.3(m
)
NA
SD9
35
32
30
2
20
48
100
100
100
48
52
100
32(m
)
.2(l
)
.02(
l)
.2(l
)
.5(m
)
NA
SD8
27
32
48
2
20
30
100
100
100
15
85
100
32(m
)
NA
.02(
l)
.2(l
)
.5(m
)
NA
SD7
27
32
40
5
20
35
100
100
100
22
78
100
32(m
)
.4(l
)
.05(
l)
.2(l
)
NA
NA
SD6
35
32
40
5
10
45
100
100
100
17
83
100
32(m
)
.4(l
)
.05(
l)
.1(l
)
NA
NA
SD5
48
32
40
60
100
100
100
24
76
100
32(m
)
.4(l
)
NA
NA
NA
NA
SD4
20
32
10
5
40
45
100
100
100
88
12
100
32(m
)
.1(l
)
.05(
l)
.4(l
)
NA
NA
SD3
33
31.2
5
40
55
1
00
1
00
1
00
1
00
1
00
32(m
)
NA
.05(
l)
.4(l
)
NA
NA
SD2
37
30
40
60
100
100
100
100
100
32(m
)
NA
NA
.4(l
)
NA
NA
SD1
57
32
100
100
100
100
100
100
32(m
)
NA
NA
NA
NA
NA
Ser
vice
Des
ign
(SD
)
Cus
tom
er s
hare
Var
iabl
e co
st/A
TM
(R
s. 0
00)
AT
M lo
cati
on
AT
Ms
at m
arke
ts o
nly
AT
Ms
at r
ailw
ay s
tatio
ns
AT
Ms
at b
ranc
hes
only
AT
Ms
in r
esid
entia
l are
as
AT
M f
unct
iona
lity
Onl
y ba
nkin
g se
rvic
es
Ban
king
+ U
tili
ty b
ill
AT
M B
reak
dow
m
Onc
e in
ten
visi
ts
Onc
e in
50
visi
ts
Fee
s
No
fee
Fee
of
Rs.
50
Sec
urit
y
Sec
urity
gua
rd +
Car
d
Ent
ry b
y ca
rd
Lim
it o
n w
ithd
raw
al
Lim
it on
with
draw
al
Wit
hdra
w u
p to
bal
ance
Con
stra
ints
Var
iabl
e co
st/A
TM
(R
s. 0
00)
AT
Ms
at m
arke
ts o
nly
AT
Ms
at r
ailw
ay s
tatio
ns
AT
Ms
at b
ranc
hes
only
AT
Ms
in r
esid
entia
l are
as
Sec
urity
gua
rd +
Car
d
Not
es:
(l)
is a
t lea
st th
is m
uch;
(m
) is
at m
ost t
his
muc
h.
22 A. Sachan, S. Datta and A.P. Arora
Table 2 Differentiation strategy
SD12
10
0
5
9
39.8
35
5
10
50
1
00
1
00
1
00
1
00
1
00
NA
.35(
l)
.05(
l)
0.1(
l)
0.5(
l)
NA
SD11
10
0
5
5
40.3
35
5
20
40
1
00
1
00
1
00
1
00
1
00
NA
0.35
(l)
.05(
l)
0.2(
l)
0.35
(m)
NA
SD10
100
64
42
45
5
10
40
100
100
100
100
100
NA
0.4(
l)
.05(
l)
0.1(
l)
0.4(
m)
NA
SD9
10
0
5
4
40.3
40
2
20
35
1
00
1
00
1
00
1
00
1
00
NA
0.1(
l)
.05(
l)
0.2(
l)
0.35
(m)
NA
SD8
10
0
5
3
39.3
35
5
20
40
1
00
1
00
1
00
1
00
1
00
NA
0.1(
l)
.05(
l)
0.2(
l)
0.4(
m)
NA
SD7
10
0
6
1
40.5
35
10
20
35
1
00
1
00
1
00
1
00
1
00
NA
0.1(
l)
0.1(
l)
0.2(
l)
0.35
(m)
NA
SD6
10
0
4
0
39.5
30
10
20
40
1
00
1
00
1
00
1
00
1
00
NA
0.1(
l)
0.1(
l)
0.2(
l)
0.4(
m)
NA
SD5
100
58
44
50
10
10
30
100
100
100
100
100
NA
0.5(
m)
0.1(
l)
0.1(
l)
0.1(
l)
NA
SD4
100
60
42
40
10
10
40
100
100
100
100
100
NA
0.4(
m)
0.1(
l)
0.1(
l)
0.1(
l)
NA
SD3
1
00
5
6.5
40
30
10
10
50
1
00
1
00
1
00
1
00
1
00
NA
0.3(
m)
0.1(
l)
0.1(
l)
0.1(
l)
NA
SD2
100
51
38
20
10
10
60
100
100
100
100
100
NA
0.2(
m)
0.1(
l)
0.1(
l)
0.1(
l)
NA
SD1
100
69
44
60
40
100
100
100
100
100 44
NA
NA
NA
NA
NA
Ser
vice
Des
ign
(SD
)
Num
ber
of A
TM
s
Cus
tom
er s
hare
Var
iabl
e co
st/A
TM
(R
s. 0
00)
AT
M lo
cati
on
AT
Ms
at m
arke
ts o
nly
AT
Ms
at r
ailw
ay s
tati
ons
AT
Ms
at b
ranc
hes
only
AT
Ms
in r
esid
entia
l are
as
AT
M f
unct
iona
lity
Onl
y ba
nkin
g se
rvic
es
Ban
king
+ U
tili
ty b
ill
AT
M B
reak
dow
m
Onc
e in
ten
visi
ts
Onc
e in
50
visi
ts
Fee
s
No
fee
Fee
of
Rs.
50
Sec
urit
y
Sec
urit
y gu
ard
+ C
ard
Ent
ry b
y ca
rd
Lim
it o
n w
ithd
raw
al
Lim
it on
with
draw
al
With
draw
up
to b
alan
ce
Con
stra
ints
Var
iabl
e co
st/A
TM
(R
s. 0
00)
AT
Ms
at m
arke
ts o
nly
AT
Ms
at r
ailw
ay s
tati
ons
AT
Ms
at b
ranc
hes
only
AT
Ms
in r
esid
entia
l are
as
Sec
urity
gua
rd +
Car
d
Not
es:
CS
is C
usto
mer
Sha
re; (
l) is
at l
east
this
muc
h, (
m)
is a
t mos
t thi
s m
uch.
Integrating customer preferences and organisation strategy 23
Table 3 Focus strategy
SD11
1
00
44
29.
5
50
50
1
00
1
00
1
00
1
00
1
00
35(m
)
NA
NA
NA
0.5(
m)
NA
32
SD10
1
00
44
34.
7
15
5
30
50
1
00
1
00
1
00
1
00
1
00
35(m
)
0.15
(l)
.05(
l)
0.2(
l)
0.5(
m)
NA
40
SD9
1
00
44
33
10
5
35
50
1
00
1
00
1
00
1
00
1
00
35(m
)
0.1(
l)
.05(
l)
0.2(
l)
0.5(
m)
NA
40
SD8
100
47.
3
34.3
10
5
20
65
100
100
100
100
100
35(m
)
0.1(
l)
.05(
l)
0.2(
l)
NA
NA
48
.4
SD7
1
00
47
32.
2
5
20
75
1
00
1
00
1
00
1
00
1
00
35(m
)
NA
.05(
l)
0.2(
l)
NA
NA
48
SD6
1
00
50
32.
7
5
10
85
1
00
1
00
1
00
1
00
1
00
35(m
)
NA
.05(
l)
0.1(
l)
NA
NA
50
SD5
1
00
44
33.
25
15
0
35
50
1
00
1
00
1
00
1
00
1
00
35(m
)
0.15
(l)
NA
NA
0.5(
m)
NA
40
SD4
1
00
47.
3
35
15
5
25
55
1
00
1
00
1
00
1
00
1
00
35(m
)
0.15
(l)
.05(
l)
0.2(
l)
NA
NA
47
SD3
1
00
48.
6
35
20
0
20
60
1
00
1
00
1
00
1
00
1
00
35(m
)
0.2(
l)
NA
0.2(
l)
NA
NA
53
SD2
100
38
35
20
5
45
30
100
100
100
100
100
35(m
)
0.2(
l)
.05(
l)
NA
NA
NA
22
SD1
100
51
35
10
5
5
80
100
100
100
100
100
35(m
)
0.1(
l)
.05(
l)
NA
NA
NA
53
Ser
vice
Des
ign
(SD
)
Num
ber
of A
TM
s
M S
of
Clu
ster
2
Var
iabl
e co
st/A
TM
(R
s. 0
00)
AT
M lo
cati
on
AT
Ms
at m
arke
ts
AT
Ms
at r
ailw
a y s
tati
ons
AT
Ms
at b
ranc
hes
AT
Ms
in r
esid
entia
l are
as
AT
M f
unct
iona
lity
Onl
y ba
nkin
g se
rvic
es
Ban
king
+ U
tili
ty b
ill
AT
M B
reak
dow
m
Onc
e in
ten
visi
ts
Onc
e in
50
visi
ts
Fee
s
No
fee
Fee
of
Rs.
50
Sec
urit
y
Sec
urit
y gu
ard
+ C
ard
Ent
ry b
y ca
rd
Lim
it o
n w
ithd
raw
al
Lim
it on
with
draw
al
With
draw
up
to b
alan
ce
Con
stra
ints
Var
iabl
e co
st/A
TM
(R
s. 0
00)
AT
Ms
at m
arke
ts o
nly
AT
Ms
at r
ailw
a y s
tati
ons
AT
Ms
at b
ranc
hes
only
AT
Ms
in r
esid
entia
l are
as
Sec
urity
gua
rd +
Car
d
C
usto
mer
sha
re (
popu
lati
on)
Not
es:
(
l) is
at l
east
this
muc
h; (
m)
is a
t mos
t thi
s m
uch.