Master thesis
Multi-site ERP implementation versus single-site
ERP implementation – An investigation via
inventory analysis in a case study
Student: Marit Pool
Student number: S3213021
MSc Technology and Operations Management 1st supervisor: prof. dr. ir. J.C. Wortmann
2nd supervisor: dr. ir. S. Fazi
2
Abstract
An ERP system is a company wide information system that integrates all aspects of an organization. As companies are expanding and are becoming more complex, it can be interesting to implement multi-site ERP systems, however, implementing these information systems can be costly and complex. According to literature the way different sites are controlled and coordinated play an important role whether to implement multiple single-site ERP systems or a multi-site ERP system.
In this research the main question is ‘What are the criteria to choose for either a multi-site ERP implementation or multiple single ERP implementation?’ and the problem of multi-site ERP implementation versus single-site ERP implementation is research at the company Company A as they have both a multi-site and a single-site ERP implementation. This company wanted more insight in their inventory levels and this is studied in this thesis. Two cases were analyzed, Site X – Site Z, the multi-site ERP implementation, and Site X – Site Y, the single site ERP implementation. Inventory levels are analyzed and interviews were used.
The hypothesis which is tested is; the route Site X – Site Y is performing worse than the route Site X – Site Z by internal deliveries. This is outcome is expected as Site Y is not included in the multi-site ERP system, and less information is available. For internal deliveries is focused on the inventory levels, and it is expected that the inventories for the sites Site X and Site Z are better managed and are performing better.
For analyzing the inventory levels the following steps were taken; segmenting the products according the Kraljics Purchasing Portfolio Model, inventory levels were measured, product turnover is calculated, inventory was divided in different groups and the EOQ was calculated.
3
Table of content
Abstract ... 2 List of tables ... 6 List of abbreviations ... 6 List of figures... 7 1. Introduction ... 8 2. Theoretical Background ... 10 2.1 Characteristics of ERP ... 102.2 Multi-site ERP implementation ... 10
2.3 Criteria multi-site ERP implementation ... 11
3. Methodology ... 14
3.1 Setting of the research ... 14
3.2 Unit of Analysis ... 14
3.3 Case selection ... 14
3.4 Hypothesis ... 15
3.5 Data collection ... 15
3.6 Guarantee quality of the research ... 16
4. Theory inventory management ... 17
4.1 Kraljics Purchasing Portfolio Model ... 17
4.2 Product turnover ... 17
4.3 Components of inventory ... 18
4.4 Economic Order Quantity ... 19
4.5 Time-phased order point ... 19
5. Set-up ERP systems ... 21
4
7.1.2 Product turnover ... 28
7.1.3 Components of inventory ... 29
7.1.4 EOQ ... 30
7.1.5 Explanations inventory levels ... 30
7.2 Site X – Site Z ... 32
7.2.1 Inventory levels ... 32
7.2.2 Product turnover ... 33
7.2.3 Components of inventory ... 34
7.2.4 EOQ ... 35
7.2.5 Explanation inventory levels ... 35
7.3 Cross – case analysis ... 36
8. Discussion ... 37
8.1 Reflection on results ... 37
8.2 Managerial implications ... 38
8.3 Implications for future research ... 39
9. Conclusions ... 40
10. References ... 41
Appendix I. Interview questions ... 44
Appendix II. Summary interviews ... 45
Appendix III. Kraljics purchasing portfolio model Company A ... 48
Appendix IV. Product turnover and EOQ of sugar paste ... 49
Appendix V. Sales in lead time and safety stock sugar paste ... 50
Appendix VI. Difference between expected inventory levels and actual inventory levels sugar paste ... 51
Appendix VII. Comparison production Site X and Site Y ... 52
Appendix VIII. Purchase costs sugar paste 2016 Company A ... 55
Appendix IX. MRP view sugar paste ... 56
Appendix X. Inventory sugar paste 2016 Company A and Company A Site Y ... 57
Appendix XI. Production sugar paste 2016 ... 59
Appendix XII: Inventory levels fruits ... 60
Appendix XIII. Purchase costs fruits 2016 Company A ... 61
Appendix XIV. Product turnover and EOQ of fruits ... 62
6
List of tables
Table 1. Categories ERP implementation according to Parr & Shanks (2000) ... 12
Table 2. Distinction criteria coordination and control ... 13
Table 3. Case selection ... 15
Table 4. Kraljics purchasing portfolio model ... 17
Table 5. Components and definitions according to Stadtler & Kilger, 2005 ... 18
Table 6. Average amount of kg raw material needed in Site X for one production ... 31
Table 7. SKUs overlap Site X and Site Z ... 33
Table 8. Average amount of raw material in kg in one production order for Site X and Site Z ... 36
Table 9. Interview questions planner Company A Site Y ... 44
Table 10. Interview questions purchaser Company A ... 44
Table 11. Interview questions Planner Company A ... 44
Table 12. Interview questions IT manager Company A ... 44
Table 13. Product turnover and EOQ of sugar paste ... 49
Table 14. Sales in lead time and safety stock sugar paste ... 50
Table 15. Difference between expected inventory levels and actual inventory levels sugar paste ... 51
Table 16. Purchase costs sugar paste 2016 Company A ... 55
Table 17. Inventory sugar paste Company A 2016 in KG ... 57
Table 18. Inventory sugar paste Company A Site Y 2016 in KG ... 58
Table 19. Production sugar paste 2016 ... 59
Table 20. Inventory levels fruits ... 60
Table 21. Purchase costs fruits ... 61
Table 22. Product turnover and EOQ of fruits ... 62
Table 23. Sales in lead time and safety stock fruits ... 63
Table 24. Difference between expected inventory levels and actual inventory levels fruits ... 64
Table 25. Production fruits 2016 ... 65
List of abbreviations
BoM………... Bill of Material
BPR……… Business Process Reengineering EOQ……… Economic Order Quantity ERP………. Enterprise Resource Planning KPIs……… Key Performance Indicators MRP………... Material Requirements Planning MTO………... Make-To-Order
7
List of figures
Figure 1. Hypothesis internal deliveries between sites ... 15
Figure 2. Inventory levels ... 19
Figure 3. Steps analyzing inventory ... 20
Figure 4. Information- and goods flow between sites ... 22
Figure 5. Purchase process ... 23
Figure 6. Process Site X – Site Y ... 24
Figure 7. Process Site X – Site Z ... 26
Figure 8. Inventory levels sugar paste ... 28
Figure 9. Product turnover sugar paste ... 29
Figure 10. Reorder point sugar paste ... 29
Figure 11. Minimum inventory, average inventory and expected inventory for sugar paste ... 30
Figure 12. Production sugar paste 2016 ... 32
Figure 13. Inventory levels fruits ... 33
Figure 14. Product turnover fruits ... 34
Figure 15. Reorder point fruits Site X ... 34
Figure 17. Minimum inventory, average inventory and expected inventory for fruits ... 35
Figure 18. Production fruits Site Z 2016 ... 36
Figure 19. Production fruits Site X 2016 ... 36
Figure 20. Kraljics purchasing portfolio model Company A ... 48
Figure 21. Comparison production 303085 ... 52
Figure 22. Comparison production 303090 ... 52
Figure 23. Comparison production 303084 ... 52
Figure 24. Comparison production 303086 ... 52
Figure 25. Comparison production 303091 ... 52
Figure 26. Comparison production 303092 ... 52
Figure 27. Comparison production 303101 ... 53
Figure 28. Comparison production 303102 ... 53
Figure 29. Comparison production 303087 ... 53
Figure 30. Comparison production 303088 ... 53
Figure 31. Comparison production 303096 ... 53
Figure 321. Comparison production 303097 ... 53
Figure 332. Comparison production 303103 ... 54
Figure 343. Comparison production 303057 ... 54
8
1. Introduction
An Enterprise Resource Planning (ERP) system is a company wide information system that integrates all aspects of an organization, such as human resources, accounting, sales, purchasing, manufacturing, distribution and supply-chain management. Information is accessible through one database and one application (Bingi, Sharma, & Godla, 1999). Correctly implementing ERP systems can be very costly and an exhaustive process, but it can also be a competitive advantage (Bingi et al., 1999).
One of the difficulties of the implementation of ERP systems is deciding between a single-site ERP system implementation and a multi-site ERP system implementation. Deciding between single-site or multi-site ERP systems depends on the desired degree of individual site autonomy (Umble, Haft, & Umble, 2003). In the last decades organizations have changed from single-site organizations to multi-site organizations in an attempt to improve their competiveness (Aissani et al., 2012). Multi-site organizations can be defined as an organization which is organizing and segmenting their activities into several sites (Galliano & Soulié, 2012). Different sites need to be managed in order to meet the long term goals of the organization (Aissani et al., 2012). Organizations should choose either to implement one multi-site ERP system or multiple single-site ERP systems, in which each location has its own system.
The aim of this thesis is to provide criteria that are important for selecting either a multi-site ERP system or multiple single-site ERP systems. In literature a lot is written about ERP implementation, however information about critical issues when deciding between a multi-site ERP system or multiple single-site ERP systems are lacking. When information about this issue is available, the information is outdated or there are already new issues.
Studying the topic of single-site ERP implementation versus multi-site ERP implementation is interesting because as many organizations are multi-site organizations, a decision has to be made between multi-site ERP implementation and multiple single-site implementations. Besides this, existing literature lacks an overview of criteria which can support by choosing the right ERP implementation.
10
2. Theoretical Background
In this chapter the definitions will be given for multi-site ERP implementations and multiple single-site ERP implementations. The criteria which a company must satisfy to successfully implement a multi-site ERP system which are mentioned in existing literature will also be discussed. The literature was searched by using the scientific database Web of Science and Google Scholar. The keywords which were used to find literature are: ‘Multi-site ERP’ and ‘Multi-site ERP implementation’.
2.1 Characteristics of ERP
ERP systems can be explained as software programs that are used by companies to integrate and coordinate information in every division in the business. ERP programs are designed to help organizations to manage company-wide business processes, use one database and shared management report tools. ERP can be used in the divisions of Marketing and Sales, Supply Chain Management, Accounting and Finance and Human Resources. (Monk & Wagner, 2012) For successfully implementing an ERP system, it is important that the ERP implementation team and top managers are able to assess organizational fit of the target ERP system (Hong & Kim, 2002).
2.2 Multi-site ERP implementation
In the last decades most organizations have changed from single-site companies to multi-site organizations in an attempt to improve their competiveness (Aissani et al., 2012). Multi-multi-site organizations can be defined as companies which are organizing and segmenting its activities into several sites (Galliano & Soulié, 2012). For a multi-site organization ERP systems can be useful, as they can be used as a tool to reach more control and coordination between the different sites, and aim to improve the visibility, transparency and efficiency of operations. Multi-site organizations are trying to implement a multi-site ERP implementation in order to reduce costs and complexity instead of implementing multiple single-site ERP systems (Rahimi & Møller, 2013). A multi-site ERP implementation can be explained as the following; at multiple sites, the same system is implemented and working. This means that information between those different sites can be shared. When multiple single-site ERP systems are implemented information between different sites cannot be shared through the ERP systems. However, it should be noted that it is possible that for example the purchasing department uses a centralized system, and the sales department can have a lot of local autonomy. It is important to keep in mind that not a whole company is centralized, but specific departments can be centralized and other departments are not centralized. Coordination and control in ERP implementations
Misalignments in ERP projects can be viewed from the level of coordination and ERP is seen as coordination technology (Daneva & Wieringa, 2005). When there is a lack of coordination, it can lead to misalignments in ERP projects. Clemmons and Simon (2001) are focusing on coordination, but also control is mentioned. The definition of coordination according Clemmons and Simon (2001) is; ‘The level of integration between different sites of an organization’. In which integration according to the Oxford Dictionaries is explained as; ‘Combining (one thing) with
another to form a whole’ (Integrate n.d.). Lee, Siau, and Hong (2003) define enterprise integration
11 Simon, 2001). Coordination is about linking different activities of an organization in order to reach a common goal (Van De Ven, Delbecq, & Koenig, 1976). Malone and Crowston (1990) are stating that coordination is about the management of dependencies to link all the different activities together. Daneva and Wieringa (2005) are explaining different ways of coordination; the definition of goals, processes, common meanings of key information entities and the interpretation of information. Effective coordination can be reached through structuring and creating standards in order to minimize interdependencies, or through extensive communication which are managing the interdependencies or through mediation by an entity which is able the solve conflicts (Gosain, Lee, & Kim, 2005). Control can be explained as ‘the adherence to a goal or target through the exercise
of power or authority’ (Clemmons & Simon, 2001). Control can reduce uncertainty, which leads
to more predictability and the common goal of the organization is reached by different parts of the organization (Clemmons & Simon, 2001). These two processes, coordination and control, can be seen as key processes for integration. It can be concluded that coordination is about linking all the different activities together within a company in order to reach a common goal, and control is about reaching the common goal using power or authority.
2.3 Criteria multi-site ERP implementation
A problem which arises with multi-site organizations is how to manage the relationship between these different sites. Markus et al., (2000) divides the different ways of relationships into five categories; total local autonomy, headquarters control only at the financial level, headquarters coordination of operations, network coordination of operations and total centralization. Accordingly, the categories are ranging from the category ‘total local autonomy’ in which no integration in ERP systems exists between different sites to integration of ERP systems from a centralized place in the category ‘total centralization’. According to Markus et al., (2000), the desire or need for centralized control plays a role in deciding whether to implement a single-site ERP system or a multi-site ERP system.
Besides centralization, the notion of coordination, i.e. the degree of process and product consistency across sites plays an important role (Umble et al., 2003). If processes or products are comparable it would be easier to implement a multi-site ERP system than in a situation in which processes and products are totally different. However, it should be noted that the level across different departments can differ, i.e. the purchasing department can be centralized and the sales department can have a lot of local autonomy. This is important to consider, that not a whole company is centralized, but specific departments can be centralized and other departments not. Another issue is deciding between corporate standardization, in which it is easy to move people, products and information between the sites, or local optimization, in which according to Umble et al., (2003) the focus is more on efficiency and reducing costs. Moreover, it is easier to adapt to local needs when the focus is on local autonomy. In terms of coordination, a decision should be made what is in line with the goal of the organization.
12 system rather than the other way around. Implementation takes about 6-12 months. The characteristics of the category ‘Middle-road’ are that an organization exists of multiple sites, and most of the time only core ERP modules are implemented. Business process reengineering (BPR) plays are role here, but not as extensive as in the category ‘Comprehensive’. Business process reengineering is defined as; ‘the implementation of deliberate and fundamental change in business
processes to achieve breakthrough improvements in performance’ (Grover, Jeong, Kettinger, &
Teng, 1995) Implementation of the ERP system can take approximately 3-5 years. The category ‘Comprehensive’ represents most of the time multinationals, with multiple sites. Full ERP systems are implemented, with a possibility of implementing industry specific modules. BPR is high in this category. Options are either implementing ERP by ‘module-by-module’ in which step by step a module is implemented and then moved to the next module, or either ‘full ERP’ in which everything is implemented at once. Implementation of such a system can take up to seven years. Parr and Shanks (2000) mentioned also several characteristics which play are role when implementing ERP systems. As mentioned before, BPR plays an important role, as it is a necessary feature of ERP implementation. Besides this the physical scope, i.e. single-site or multiple-site, technical scope, i.e. how much needs the ERP software to be modified, module implementation strategy and resource scope, i.e. time and budget. Linking this to the problem, deciding between single-site implementation or multi-site implementation, single-site implementation can be considered when there is no desire to make any modifications to the ERP system, and the current business processes will be aligned to the ERP system. For a multi-site implementation, BPR is high, and the need for modifications of the ERP system is high. Parr and Shanks (2000) divided the implementation into three categories, however, in reality it can be difficult to determine whether an implementation is either ‘Vanilla’ or ‘Middle-road’ or ‘Comprehensive’ beforehand. Therefore, it can be difficult to determine which factors needs to be included for successfully implementing an ERP system. Another difficulty with this distinction of these three categories is that it can be difficult to place an ERP system implementation into one group, as an implementation can belong with certain aspects to one category and with other aspects more to the other category. The distinctions between groups can be vague.
Table 1. Categories ERP implementation according to Parr & Shanks (2000)
Vanilla Middle-road Comprehensive
Criteria One site Multi-site Multi-site
Users <100 BPR BPR Implementation takes 6 - 12 months Implementations takes 3 – 5 years Implementation takes up to 7 years
When organizations have different sites in different countries, the factors national culture and government/corporate policies play an important role in successful implementation (Yen & Sheu, 2004).
13 for centralization (Markus et al., 2000), the degree of process and production consistency across sites, corporate standardization versus local optimization (Umble et al., 2003), BPR, the physical scope, the technical scope, the module implementation strategy and the resource scope (Parr & Shanks, 2000). All these factors relate to the control and coordination processes which are mentioned by Clemmons and Simon (2001), these factors are shown in Table 2.
Table 2. Distinction criteria coordination and control
Author Control Coordination
Markus et al., (2000) Total local autonomy vs. total centralization Umble et al., (2003) Corporate
standardization vs. local optimization
Process and production consistency
Hong and Kim (2002) Organization fit
Parr and Shanks (2000) BPR
Physical scope Technical scope
Implementation strategy Resource scope
Yen and Sheu (2004) Government &
corporate policies
National culture
Daneva and Wieringa (2005) Definition of goals
Processes
Common meaning of key information Interpretation of information
Malone and Crowston (1990) Management of dependencies
Lee and Kim (2005) Structuring and creating standards
Extensive communication Mediation by an entity
These factors are stated in literature; however, a clear framework is missing when deciding between a single-site or multi-site implementation. As is shown in Table 2, most literature is from early 2000s, which is not very up-to-date, and new literature is missing. Especially when an organization should decide whether to implement different single-site ERP systems or one multi-site ERP system. This leads to the following general research question:
14
3. Methodology
In this chapter the methodology will be discussed. First it is explained what methods will be used, then the setting of the research will be explained, unit of analysis is mentioned and a case selection is done. Then the method of data collection and how the quality of the research is assured is explained. This chapter ends with a description of how the data will be organized and will be analyzed.
In this research is sought for the criteria to choose either for a multi-site ERP implementation or an integrated single-site implementation or multiple single-site implementations. Currently, little scientific literature is available about multi-site ERP implementation, and the literature which is available, is most of the time more than 15 years old and outdated. A case study is suitable to look for factors which are important for multi-site ERP implementation. A case study is used when a lot of different variables can influence decision whether to use a multi-site ERP implementation or a single-site ERP implementation. The relations can be complex, and an in-depth and context research needs to be executed. Therefore, a case study is suitable for this research. The case study is used for theory building, and it is an exploratory research in which is looked if some defined independent variables will influence the dependent variable. In this research is sought for the linkages between the variables and why these linkages and relationships exists. For this case study two companies are needed, to research a multi-site ERP implementation between sites and to research a single-site ERP implementation.
3.1 Setting of the research
As mentioned, the research question is about factors which influence the choice for either a multi-site ERP implementation or a multiple single-site implementation. To make this comparison, a multi-site ERP implementation and a company that has a single-site ERP implementation was needed. As I am doing my master thesis project at a company which is producing bakery ingredients. Due to confidentiality the company is described as ‘company A’ in this thesis. Company A, which has both a multi-site ERP implementation and a single-site ERP implementation, it is suitable for the research. The purchasing department is crucial for Company A, and they want more insight in their inventories. Therefore, the focus will be on the purchasing department and on their inventories.
3.2 Unit of Analysis
The Unit of Analysis is derived from the research questions, as it is intended to find how certain criteria influence multi-site ERP implementations & multiple single-site implementations. The cases which were analyzed are ERP implementations.
3.3 Case selection
The focus for the case selection was on a company with multi-site ERP implementations and a multiple single-site implementations, the explanation of these definitions are mentioned in the theoretical background.
15 Table 3. Case selection
Case 1 2
ERP implementation Site X - Site Z
ERP implementation Site X - Site Y
Criteria Multi-site Single-site
3.4 Hypothesis
In this thesis, the general research question is; ‘What are the criteria to choose for either a
multi-site ERP implementation or multiple single ERP implementation?’ This is researched at the
company Company A, as it has a multi-site ERP implementation for locations Site X and Site Z and a single-site ERP implementation at the site Site Y. As mentioned before, the company wants more insight in their inventories, as the inventories levels could be too high.
The hypothesis linked to the general research questions is; the route Site X – Site Y is performing worse than the route Site X – Site Z by internal deliveries. This is outcome is expected as Site Y is not included in the multi-site ERP system, and less information is available. For internal deliveries the inventory levels are analyzed, and it is expected that the inventories for the sites Site X and Site Z are better managed and are performing better. The relation is shown in Figure 1.
Figure 1. Hypothesis internal deliveries between sites 3.5 Data collection
16 Data was collected by comparing the turnover of products of the multi-site implementation and the two single-site ERP implementations. A selection of products was made based on the Kraljics purchasing portfolio model, which is based on the products’ profit risk and supply risk. The focus should be on products which have a high supply risk or profit risk.
When this was done, the analysis of the optimal order quantities were compared of the two different ERP implementations and was sought for similarities and differences. Also, was sought for explanations for similarities and differences between the results. The explanations were found by conducting semi-structured interviews. The interviews were conducted with people who were related to the ordering process and purchasing department. The interview questions can be found in Appendix I. The interview summaries can be found in Appendix II.
When explanations were found, it was checked if these findings also hold for other departments of Company A Lastly, the findings of this research were compared to the findings of expert interviews.
3.6 Guarantee quality of the research
The chain of evidence was maintained by explaining step by step what was done in the research and how conclusions were made. First the relevant variables were mentioned and explained, then was explained why these variables were chosen for the case selection, and how the cases were selected. The next step was data collection, which was explained extensively. The findings of the case study were compared with expert interviews to be able to generalize the findings of the case study.
17
4. Theory inventory management
In this chapter, some background information will be given about inventory management as the focus of this research is on inventory management and purchasing of Company A To be able to draw conclusions about their inventory levels, some theoretical background about inventory management is needed. In paragraph 4.1 the Kraljics Purchasing Portfolio Model is explained, 4.2 is about the product turnover, in 4.3 the different components of inventory will be explained. When this is done the EOQ will be explained in paragraph 4.4 and the time-phased order point is explained in paragraph 4.5. This chapter ends with a model which steps should be taken to analyze inventory levels.
4.1 Kraljics Purchasing Portfolio Model
The Kraljics Purchasing Portfolio Model can be explained by two factors, the profit impact and the supply risk. The profit impact exists of the volume purchased, percentage of total purchase costs and/or the impact on product quality or business growth. The supply risk is explained by availability, number of suppliers, competitive demand, make-or-buy opportunities, storage risks and substitution possibilities. This leads to the purchasing portfolio model which is shown in Table 4. (Caniëls & Gelderman, 2005; Kraljic, 1983).
Table 4. Kraljics purchasing portfolio model
Profit impact Supply risk
Low High
High Leverage items Strategic items
Low Non-critical items Bottleneck items
Strategic items can often only be purchased from one supplier, which leads to a large supply risk. Here the purchasing strategy is to either accept the situation and the power of the supplier or find a new partner. Bottleneck items have a high supply risk as suppliers have a large power position for the products, the purchasing strategy is accepting the dominance of the supplier, and trying to reduce the negative effects of this position, or otherwise finding a new partner if this is not possible. Leverage items can in general be obtained of several suppliers, this item has a large share in the cost price of the final product. The buyer has a lot of options for negotiation, as small price changes lead to large savings. The purchasing strategy here is to exploit the buying power, or otherwise try to develop a strategic partnership. Non-critical items have a low value per unit and a lot of suppliers can be found for this. The purchasing strategy is aiming for standardization and bundling of purchasing requirements, or individual ordering is advised. (Caniels & Gelderman, 2005)
4.2 Product turnover
18 4.3 Components of inventory
Holding inventory is about finding the trade-off between the costs for holding inventory and the benefits such as high service level. To be able to making decisions about the inventory, it needs to be known what functions inventory has. In Table 5 the different components and definitions of inventory are stated according to (Stadtler & Kilger, 2005).
For different products, or Stock Keeping Units (SKUs), inventory can be hold. a SKU is defined according to the business dictionary as the following; ‘Warehousing item that is unique because of
some characteristic, i.e. brand, size, color, model, and must be stored and accounted for separate from other items.’ (stock keeping unit, n.d.) What also needs to be considered whether a product is
produced Make-To-Stock (MTS) or Make-to-Order (MTO). MTS means that inventory is kept for a product. Most often the MTS strategy is applied at the starting phase of a product where demand is unknown. When demand is known, the MTO strategy is applied. With MTO, only with a customer order product the product will be produced. This means, in principle, that the inventory level of this SKU is zero. (Gupta & Benjaafar, 2004)
The safety stock can be calculated with the following formula: 𝑆𝑎𝑓𝑒𝑡𝑦 𝑠𝑡𝑜𝑐𝑘 = 𝑧 ∗ 𝜎𝐷√𝐿𝑇 Where;
z = safety factor depending on the service level σD = standard deviation in demand per period LT = lead time supplier expressed in the same period
Table 5. Components and definitions according to Stadtler & Kilger, 2005
Component Definition
Production Lot-sizing Stock Stock is used to cover demand between two consecutive production runs
Transportation Lot-sizing Stock One shipment should cover the demand until the next shipment arrives due to the fact it is cheaper to ship a full truckload
Inventory in Transit Inventory that is transported in-between, dependent on transportation time and demand Seasonal Stock Inventory is held to buffer future demand peaks
which exceeds the production capacity
Work-in-process Inventory (WIP) The process of transformation from raw materials to finished goods
Safety Stock Safety stock exists for protection against
19 4.4 Economic Order Quantity
When the components and reasons of a certain stock level are known, it is possible to determine the optimal order quantity. This can be done with the Economic Order Quantity (EOQ). The formula is the following:
𝑄∗= √2𝐷𝐾 ℎ Where;
Q* = optimal order quantity D = annual demand quantity
K = fixed cost per order, setup costs h = annual holding cost per unit
The formula gives an optimal amount which should be ordered considering the demand, order costs and holding costs. Q should be the maximum inventory level. When Q is divided by 2, the average inventory level is shown.
4.5 Time-phased order point
Materials Requirements Planning (MRP) is managing inventory of companies with the help of time-phased inventory levels (Verwijmeren, van der Vlist, & van Donselaar, 1996). MRP is explained as a computer-based planning tool, which looks in the future what is needed for a finished good. For MRP is a master production schedule, Bill of Materials (BoM), inventory status data and lead time information needed. MRP generates requirements for components which are needed for a finished good (Burcher, 2015). With the help of a BoM the total requirements for a component in a certain period is determined (Verwijmeren et al., 1996). When the requirements are known, the current inventory levels are compared with the total requirements of a component, and based on that it is decided whether new materials need to be ordered or not. The Time-Phased Order Point (TPOP) is an approach to use in deciding when to order new components when demand is not stable. Time-Phased means that it is checked when inventory levels of a certain SKU drop below the reorder point (R), which is also shown in Figure 2, and this means that a new order has to be placed for a certain SKU. The order quantity is ideally the number which is calculated with the EOQ formula, as it leads to the lowest costs.
20 Concluding from the above, the steps which to be taken to be able to draw some conclusions about inventory levels, are the following; first the products must be segmented according to the Kraljics Purchasing Portfolio Model. The next step is to measure the stock levels of these important products. The following step is calculating the product turnover. this can be done by using the KPI product turnover. When product turnover is low of a certain product, it needs to be analyzed why the product turnover is low. This can be done by analyzing the certain components of inventory. When this is known, the optimal order quantity can be determined. The steps which need to be taken are shown in Figure 3.
21
5. Set-up ERP systems
In this chapter the set-up of the ERP systems within Company A are explained. The goods and information flow between the sites is also shown in this chapter.
The sites Site X and Site Z are part of Company A, and are in the same ERP system. The system is Exact Globe Next, version 410 with the use of Synergy.
Exact is an ERP system which main customers are mainly in the small and medium enterprise sector. Exact Globe is a system which integrates business processes, such as logistics, financials, product data management and customer relationship management. Synergy is an application which gives real-time information, such as financials, and relationship and knowledge management. With this application it is possible to provide employees company-wide real time information. (Jansen, Brinkkemper, Ballintijn & van Nieuwland, 2005)
Site X and Site Z fall both on under ‘administration 100’ which is also shown in Figure 4. The reason that both sites are included in the same ERP system is that it is easier for financial reporting when these sites are in the same ERP system. If the sites were separated, it would take more effort to consolidate the financial reporting. The site Site Y is not included in the same company, Company A, as Site X and Site Z, but is Company A Site Y B.V. Site Y in not a part of Company A is since the companies have different pension plans, and it would cost a lot of money to merge. The ERP system that the site Site Y uses is Exact Globe Next version 410. Site Y falls under ‘administration 104’ which is also shown in Figure 4. The dotted lines represent information flows, and the other lines are representing goods flow.
Each customer sends an order to Company A, also if the products are produced at the site in Site Y. Company A sends an invoice back to the customer. The products will be delivered from the sites which produces the products. The products will be stored in the warehouses of each site after production. From the warehouse of the site the products will be delivered to the customer. This process is also shown in Figure 4.
When an order is placed for Site Y at Company A, there would be problem sending the order to Site Y, as it is in a different system. However, this problem is solved by using a module of Exact, E-Intercompany Logistics. The module is described as ‘optimizing logistics between multiple sites
with a central warehouse. A benefit of this module is that a purchase order only once has to be entered, changes are updated everywhere at once, and it is possible to see stock levels form all sites.’ (Exact, 2017) With this module the orders are sent to Site Y which means that it is not
22
23
6. Processes
In this chapter the different processes from receiving an order to shipping an order to a customer and how the different sites are related to each other will be described. First the route Site X – Site Y is described and then Site X – Site Z is described. After this a cross-case analysis was executed.
6.1 Purchasing process
For purchasing the raw materials for the sites Site X and Site Y, the following products are overlapping; sugar, dextrose, milk powders, dyes, whey powders, lactose, almonds and packaging materials such as buckets. As mentioned before, the sites Site X and Site Y are not in the same ERP environment. This means that the total usage for raw materials which are both used at these sites needs to be combined and then an overall contract can be made. When this is done, the usage can be split up again in and put in the different systems. This process is shown in Figure 5. Different product numbers are used at the different sites; this is because of traceability of the raw materials. For purchasing raw materials, the products sugar, milk powders and some fruits are overlapping for Site X and Site Z. As Site X and Site Z are working in the same ERP system, it would be convenient for the purchaser to purchase these raw materials as they could have the same product number. However, the raw materials do not have the same product numbers. The reason why the sites Site X and Site Z are using different product numbers for the same products is because the sites want to keep track of their own raw materials, and are afraid that if the product number is the same, control and traceability is lost. The product numbers for Site X are starting with a ‘3’ and for Site Z the product numbers are starting with ‘13’. For the purchaser, this means when setting up a purchasing contract, total usage needs to be combined. As total usage is determined, price agreements will be made. When this is done, the total usage will be split again and usage per site will be filled in the systems, this process is shown in Figure 5. A difficulty when the total usage needs to be combined manually, it needs to be checked manually if the agreed amount in the contract is satisfactory or not.
24 6.2 Site X – Site Y
In Site Y large amounts of sugar paste is produced and this is processed further in Site X, where the large amounts of sugar paste are split up in smaller amounts for products for the customers. The finished products are MTO, however for the raw materials a minimum inventory level is agreed.
In Figure 6 the process of producing sugar paste is described. It starts with Site X receiving an order for sugar paste. Then the planner in Site X starts planning the production for sugar paste. Based on MRP and a minimum order level it is checked if enough raw materials are available, an image of MRP is shown in Appendix IX. For some products the minimum order level is zero due to the unpredictable demand of the product. When for these products an order is received, the planner of Site Y is contacted. For the other raw materials the planner is contacted when the raw materials are below the minimum order level. An email is sent to the planner in Site Y. The planner in Site Y checks when it is possible to produce the sugar paste. Then the planner in Site X receives an answer, and sends a purchase order to Site Y with the right date. The next step is that Site Y receives the purchase order and transportation will be arranged for the sugar paste. After this, Site Y plans the production and the sugar paste will be produced. For each color or sugar paste, at least 1000kg needs to be made in order to be able to start production. When the sugar paste is produced, it has to be stored 5 days in Site Y before Site X is able to process the sugar paste. After this time, the sugar paste will be transported and received by Site X. This process from sending an ordering e-mail from Site X to Site Y until receiving the sugar paste in Site X, takes two to three weeks.
When enough raw material for production is available, the sugar paste is handled in Site X. At least 250 kg of each color needs to be handled to start production. When the goods are finished, products will be consolidated and collected and the order is transported to the customer. Overall Site X has three weeks for the process receiving the order till transportation of the product. As receiving sugar paste takes already two to three weeks, not much time is left for production.
Figure 6. Process Site X – Site Y 6.3 Site X – Site Z
25 When Site X receives an order, production will be planned. The next step is to check whether the inventory levels of the needed raw materials are still above a certain level. If the raw materials are below a certain level, the materials will be ordered, transported and received by Site X, the lead time for this is seven days. If the inventory levels are above a certain level, production is executed and the products are collected and consolidated and the order will be shipped to the customer.
If Site Z receives an order, production will be planned for this order. First the availability of raw materials under the product number of Site Z is checked. When enough raw materials are available, production is planned. When not enough raw material is available, it is checked if Site X has the raw materials available. As Site X has enough raw materials available for Site Z, they will be reserved for Site Z and are transported to Site Z. The items will be either booked from Site X to Site Z, with changing the product numbers from ‘3…’ to ’13…’ or they just keep the product number with ‘3…’. There is not a clear distinction made between when keeping the product number the same or changing it to a Site Z product code.
26 Figure 7. Process Site X – Site Z
6.4 Cross-case analysis
Analyzing the two cases, it is noted that at all three sites different product numbers were used for the same raw materials. This leads to more complexity and the purchaser need to be aware that different product numbers are used for the same raw materials. It could be more convenient to use the same product number for the same raw materials, as then it is immediately clear how much raw materials is used. However, this only holds when the sites are operating in the same ERP system. This is not yet the case, as Site Y is not included in the same ERP system as Site X and Site Z.
27
7. Inventory analysis
In this chapter an inventory analysis will be executed. The first step is analyzing the situation Site X – Site Y in 8.1, in 8.2 the situation Site X – Site Z is analyzed. This chapter ends with a cross-case analysis.
The steps which were mentioned in Figure 3 were followed. First the product categories were categorized according to the Kraljics Purchasing Portfolio Model which is shown in Appendix III. Then the inventory levels were analyzed. The next step was calculating the product turnover, then the expected inventory was split up in different components. The last step was calculating the optimal EOQ.
The focus is on the category leverage items, when the products are categorized according to the Kraljics Purchasing Portfolio Model. These products have a large profit impact and a low supply risk. For example, when a change needs to be made, this can be easily done within this category when comparing this category with the categories strategic items or bottleneck items. Products within the category leverage items were interesting to analyze, because the products have a large impact on the profit, but the buyer has still options for negotiation. The categories non-critical items and bottleneck items are not as interesting, as their profit impact is low, and more profit related to impact can be made in the category leverage items. In the category strategic items, it is more difficult to make changes, as most of the time the buyer has to accept the situation. Thus, the category leverage items is most interesting to focus on for now, as in this category improvements can still be made which also have a large impact on profit.
In chapter four the components of inventory were explained. The components were; production lot-sizing stock, transportation lot-sizing stock, inventory in transit, seasonal stock, WIP and safety stock. The components production lot-sizing stock, seasonal stock and safety stock were considered for analyzing the case of Company A. The others were not relevant for the following reasons; transportation lot-sizing stock and inventory in transit were not considered as transportation times between sites are very short, as products can be transported within a day. The category WIP was not considered, as the products which were analyzed were raw materials.
7.1 Site X – Site Y
28 7.1.1 Inventory levels
In Figure 8 and Appendix X the inventory levels of sugar paste are shown. At the beginning of the year a peak in inventory levels of sugar paste is seen, and this also holds for the months August and September. These peaks were representing large orders of a customer which has a promotion in the beginning of the year and in the second part of the year.
Figure 8. Inventory levels sugar paste 7.1.2 Product turnover
The next step was analyzing the product turnover of sugar paste. The sales of sugar paste, which are shown in Appendix IV were divided by the average inventory which leads to the product turnover which is shown in Appendix IV and Figure 9. Two products have a product turnover below 5. However, their purchasing costs are only 2% of the total purchasing costs for sugar paste as shown in Appendix VII. The product turnover of 303090 is 22, and the purchasing costs of this product are relatively 39% of the total purchasing costs for sugar paste.
-5000 0 5000 10000 15000 20000 25000 30000 35000
Jan Feb March Apr May June July Aug Sep Okt Nov Dec
KG
Inventory levels sugar paste 2016
29 Figure 9. Product turnover sugar paste
7.1.3 Components of inventory
After analyzing the inventory levels and product turnover of sugar paste, the different components of inventory of sugar paste were analyzed. Seasonality was not a component here, as Site Y can deliver the product throughout the year. The lead time for sugar paste is 2-3 weeks. Assumed was that the demand is stable, thus the reorder point should be the demand in these 2-3 weeks + the safety stock. The demand in the 2-3 weeks for sugar paste is shown in Appendix V. For calculating the safety stock, it was assumed that the service level is 96%. The safety stock is also shown in Appendix V. In Figure 10 the minimum inventory levels which are now agreed are shown, and the expected reorder point is shown. It is seen that for some products, the minimum inventory is lower than the demand lead time + safety stock.
Figure 10. Reorder point sugar paste
0 5 10 15 20 25
Product turnover sugar paste
Product turnover 5 0 5000 10000 15000 20000 25000 30000 30 3086 30 3087 30 3084 30 3090 30 3095 30 3091 30 3101 30 3085 30 3097 30 3088 30 3096 30 3102 30 3103 30 3092 30 3057 KG
Reorder point sugar paste
Minimum inventory Demand lead time
30 7.1.4 EOQ
The EOQ is calculated based on ordering costs (K) of €20, - per order and holding costs (h) of €0,15 per kg. The demand (=sales) is shown in Appendix IV. The optimal Q* is also shown in Appendix IV. The average optimal inventory level is Q/2 which is shown in Appendix VI. In this table are also the actual average inventory levels and difference between the optimal and actual inventory levels shown. In Figure 11 are the minimum inventory, average inventory and expected inventory shown. The average inventory almost always exceeds the expected inventory, except for product number 303057. For the first four products in the figure, the expected inventory is much lower than the actual average inventory.
Figure 11. Minimum inventory, average inventory and expected inventory for sugar paste 7.1.5 Explanations inventory levels
In Figure 8 two peaks in the inventory levels of sugar paste were shown. This was because at the beginning of the year and in the month September a large order of a customer of sugar paste is received. The minimum inventory levels which are needed to be available in Site X for sugar paste were based on agreements which the sales department made with customers. The planning department has adapted for some products the minimum inventory level, as the turnover of the product was highly unpredictable. This holds for the products 303057, 303091 and 303088. In Figure 11 is shown, that for most products the average inventory was still higher than the minimum inventory levels. Often the minimum inventory levels were also higher than the expected inventory levels. 0 5000 10000 15000 20000
Reorder point, average inventory and expected
inventory for sugar paste
31 In Figure 12 the production of sugar paste throughout the year is shown. In the month February a small increase in production is seen, this also case in September and November. When comparing this to the inventory levels in Figure 8, a peak is seen in Februari and in August and September. In Appendix VI are the production amounts per month for Site X and Site Y shown. It is expected that the pattern of Site X and Site Y were the same, and that the peaks in Site X were about two to three weeks later. However, the peaks of Site X were for most productss lower than the peaks of Site Y. In Table 6 is the average amount of kg raw material needed in Site X for one production shown. It is shown that three products, 303057, 303088, 303102 are below 1000 kg, which is the minimum order amount at Site Y for sugar paste.
32 Figure 12. Production sugar paste 2016
7.2 Site X – Site Z
Analyzing the Kraljics Purchasing Portfolio Model in Appendix III, the category fruit was classified as a leverage product. Profit impact was high for this category and supply risk was low.
7.2.1 Inventory levels
In Figure 13 and Appendix XII the inventory levels of fruits are shown. For the different products, some peaks in the year are seen. However, not a clear overall pattern can be seen in the inventory levels. 0 10000 20000 30000 40000 50000 60000
Jan Feb March Apr May June July Aug Sep Oct Nov Dec
KG
Production sugar paste Site X 2016
33 Figure 13. Inventory levels fruits
7.2.2 Product turnover
The next step was analyzing the product turnover of fruits. First it was important to mention that for the same raw material different product numbers are used in Site X and Site Z, this can be seen in Table 7. As mentioned before, the products which starts with a ‘3’ are for Site X and the number which are starting with ‘13’ are for Site Z. The product turnover was calculated through the following; first the average inventory was calculated, which is based on the inventory levels of fruits shown in Appendix XII. The inventory levels were measured at the end of every month. The sales of fruits were divided by the average inventory which leads to the product turnover which is shown in Appendix XIV and Figure 14. One product, 135205, has a product turnover below 5. This product has a purchase costs of relatively 5% of the total purchase costs of the named fruits, as shown in Appendix XIII. product 135211 has a product turnover of 117, this product represents for 4% of the total purchasing costs of the analyzed fruits. Calculating the product turnover for product 135207, was not possible as there was no average inventory.
Table 7. products overlap Site X and Site Z
Description Itemcode
IQF Aardbeien 15-25 MM 135211 IQF Aardbeien 15-25 MM Senga 321302 IQF Donkerrode Kersen Lutowka 135207 IQF Donkerrode Kersen Lutowka 321307
0 10000 20000 30000 40000 50000
Jan Feb March Apr May June July Aug Sep Okt Nov Dec
KG
Inventory levels fruits 2016
34 Figure 14. Product turnover fruits
7.2.3 Components of inventory
After analyzing the product turnover and the inventory levels of fruits, the different components of inventory were analyzed. Seasonality must be considered here, as fruits are not available throughout the year, and the finished products of Site Z are mostly seasonal. However, a contract for fruit is made around June/July and then the batches fruits are delivered throughout the year, thus it should not have an impact on the inventory levels. The lead time for fruits is one week. Assumed is that the demand is stable, thus the reorder point should be the demand in one week + the safety stock. The demand in the one week of fruits is shown in Appendix XV. For calculating the safety stock, it is assumed that the service level is 96%. The safety stock is also shown in Appendix XV. In Figure 15 the minimum inventory levels are shown, and the expected reorder point is shown. For product 321307 the reorder point is larger than the minimum inventory. For the Site Z items, there should not be a reorder point, as these items are MTO, and these are excluded in Figure 15. For all raw materials the demand lead time + safety stock are lower than the minimum inventory.
Figure 15. Reorder point fruits Site X
0 20 40 60 80 100 120 140 135211 321302 135205 321304 135208 135206 135209 321307 135207
Product turnover fruits
Product turnover 5 0 1000 2000 3000 4000 5000 6000 321302 321304 321307 KG
Reorder point fruits Site X
Minimum inventory
Demand lead time
35 7.2.4 EOQ
The EOQ was calculated based on the assumptions that the ordering costs (K) €20, - were per order and holding costs (h) was €0,15 per kg. The demand is shown in Appendix XIV. The optimal Q* is also shown in Appendix XIV. The average optimal inventory level is Q/2 which is shown in Appendix XVI. In this table is also the actual average inventory levels and difference between the optimal and actual inventory levels shown. In Figure 16 are the minimum inventory, average inventory and the expected inventory shown. For products 135205, 135206 and 321307, the actual average inventory is much higher than the minimum inventory levels and the expected inventory levels.
Figure 16. Minimum inventory, average inventory and expected inventory for fruits
7.2.5 Explanation inventory levels
The production of Site X is throughout the year steady. The production in Site Z is very seasonal and starts to increase in the second part of the year. Production also depends largely on how many orders come in, which fluctuates a lot, and when an order comes in, these are mostly large, a lot of raw material is needed. What is happening now, is that Site X is holding a large inventory, as Site Z gets an order, Site Z can use the raw materials that Site X is keeping on stock. In
Table 8 the average KGs of raw materials which are needed for production are shown. Site Z needs on average more kg of raw material for one production than Site X. Site Z is also using raw materials of Site X, product numbers; 321302 and 321307. In Figure 16 can be seen that product 321307 has a large average inventory level, compared to 135207 which average inventory level is 0. The minimum inventory levels are based on safety stocks + lead time calculation which was done 3-4 years ago. As it is already 3-4 years ago that the calculation was made, it could be possible that these numbers do not fit the situation now anymore. Looking at the production of Site Z in Figure 17, production is fluctuating more when comparing to the production in Site X, shown in Figure 18. 0 5000 10000 15000 20000 25000 135211 321302 135205 321304 135208 135206 135209 321307 135207
Reorder point, average inventory and expected
inventory for fruits
36 Figure 17. Production fruits Site Z 2016 Figure 18. Production fruits Site X 2016
Table 8. Average amount of raw material in kg in one production order for Site X and Site Z Site + Product No Average KG Site Z 2928 135205 3020 135206 2970 135208 3054 135209 2992 135211 2532 321302 2506 321307 2962 Site X 408 321302 555 321304 207 321307 395
7.3 Cross – case analysis
The two cases, Site X – Site Y and Site X – Site Z are two different cases. Site Y is the supplier of Site X for the sugar paste, and Site X and Site Z are using the same raw materials, i.e. fruits, for some finished goods. Analyzing the inventory levels of both cases, it cannot be said that the inventory levels are lower or higher in one case. For product turnover holds the same, in both cases have some products a low product turnover, but overall the turnover of the products is not very low. When the actual average inventory is compared with the expected inventory, it is seen that for both cases most of the time the actual average inventory is higher than the expected inventory. Thus, when comparing the cases with each other, it is difficult to find large differences regarding the inventory levels.
0 5000 10000 15000 20000 25000 30000 35000 Jan Feb Mar ch Apr May Jun e Ju ly A u g Se p O ct N o v De c KG
Production fruits Site Z 2016
135205 135206 135208 135209 135211 0 10000 20000 30000 40000 KG
Production fruits Site X
2016
37
8. Discussion
In this chapter the results which were found are compared and discussed with the current available literature which was discussed in chapter 2, the theoretical background. Limitations of this research will also be discussed in this chapter.
8.1 Reflection on results
Multi-site ERP implementation are executed to reduce costs and complexity for multi-site organizations. In this thesis the inventory levels of Company A were analyzed, in relation to multi-site ERP implementation. It was expected that the multi-sites with a multi-multi-site ERP implementation were performing better regarding internal deliveries than the site without the multi-site ERP implementation. Inventory levels of these sites were compared to be able to conclude if the inventory levels are too high or not.
It cannot be said that the sites with the multi-site ERP implementation, Site X and Site Z, were performing better regarding the internal deliveries and their inventory levels.
In the theoretical background it is explained that control and coordination are important aspects regarding multi-site ERP implementation. For control an important factor is corporate standardization versus local optimization. The focus of Company A, is local optimization. Regarding coordination, the design of processes along the different sites is important. For Company A, the sites are organized differently and there is not much overlap in processes or products. It is interesting even though there is not much overlap and corporate standardization, Company A, chose to implement a multi-site ERP system. While theory is stating that these factors are important for implementing multi-site ERP systems.
As Site Y is in a different ERP system, it was expected that sharing information would be more difficult. However, at Company A this is solved by using a module. The complexity of sharing information between sites which are using different ERP systems is solved by adding this module. However, it needs to be considered that both Site X and Site Y are using the same ERP system, Exact Globe Next. Adding a module when the sites are using different ERP systems, i.e. Exact and SAP, implementation of this module could be more difficult.
In terms of administrative processes, it easier to have all sites in the same system. When raw materials have to be transferred from Site X to Site Z, the inventory has to be booked from one warehouse to the other warehouse. When raw materials need to be transferred from Site Y to Site X, a sales order and purchase order should be generated. This means more administrative work, compared to booking raw materials from one warehouse to the other warehouse.
38 should be based on the safety stocks. What also needs to be considered is that the safety stock calculations were already made 3-4 years ago, but they were not updated afterwards. It is important to continuedly analyze and update these safety stocks, as demand patterns can change this may lead that the safety stocks are incorrect for the situation now.
Regarding the process of sugar paste, it takes Site Y 2-3 weeks to go from order to delivering to Site X. As Site X has an agreement with its customers to have its product ready for transport within three weeks. If delivering raw materials from Site Y to Site X already takes three weeks, Site X is not able to deliver to product within three weeks. A consequence is that to still fulfill the agreements of having a finished good within three weeks, inventory levels of the raw materials could be higher. The minimum amount of sugar paste which Site X can order at Site Y is 1000 kg, and in Site X the minimum amount of sugar paste to start producing is 250 kg, thus this could mean that of the order 1000 kg, 750 kg is stored as raw material, which could lead to higher inventory levels.
The orders of Site Z are MTO and demand is varying throughout the year. Inventory levels should be zero for these products. For Site X the demand of fruit is more stable throughout the year. For these products the minimum inventory level is higher than the demand lead time + safety stocks, which means that these inventory levels could be decreased. Site X should not keep inventory as a buffer for Site Z, as demand is more fluctuating and uncertain in Site Z. Now is the site Site X holding inventory for Site Z, which leads to large inventory levels, as Site Z needs large amounts of raw material for their production process.
8.2 Managerial implications
The fact that Site Y is operating in a different ERP system regarding inventory levels, does not seem to lead to a worse performance in inventory levels. Site Y can be treated as a supplier and through extensive communication of the planners of both Site X and Site Y the process of ordering sugar paste runs smoothly. The process of transferring orders is solved by adding an extra module to the ERP systems. However, an interesting finding was that in the multi-site ERP implementation of Site X and Site Z, sometimes issues were caused with reserving the raw materials, which led to shortages for production, while in theory this process would be smoother than the single-site ERP implementation.
39 8.3 Implications for future research
The method for researching this subject has several limitations which will be discussed here. The research is conducted at one company, which had both the multi-site and single-site ERP implementation. An ERP implementation is a very complex situation, and it is different for every company. The difficulties which were found in this company could not be an issue for other companies. However, it is interesting for future research to analyze the difficulties which were found and research how other companies cope with these issues. For future research it is important to research different companies, i.e. different sector or size, and check what kind of difficulties they interfere with. Researching more companies could lead to improve the generalizability of the findings.
The sites which were researched had all different operating processes in producing bakery supplies, and they did not have much overlap in their products and processes, it could be interesting for further research to find companies which have more overlap in their processes or products and how this is related to multi-site and single-site ERP implementations.
For future research it could be interesting to research other departments within in a company, this thesis is mainly focused on the purchasing department and inventory management. This is only one part of a company, in other departments, i.e. finance, sales, ERP implementations could be organized differently in terms of ERP implementation.
40
9. Conclusions
In this chapter the conclusions of this research will be given. The following research question played a central role in this thesis;
‘What are the criteria to choose for either a multi-site ERP implementation or multiple single
ERP implementations?’
This research question was researched by the following hypothesis; the route Site X – Site Y is performing worse than the route Site X – Site Z by internal deliveries. This is because Site Y is not included in the multi-site ERP system, and less information is available.
The hypothesis does not hold, the multi-site ERP implementation of Site X and Site Z are not performing better regarding the internal deliveries and their inventory levels when this is compared to Site X and Site Y. For the multi-site ERP implementation factors such as control and coordination did not play a role in choosing for a multi-site ERP implementation for Company A The expected issue was that it would be difficult to share information between the sites Site X and Site Y was easily solved by adding an extra module to Exact which communicates between the two different systems. Even though regarding the inventory levels multi-site ERP does not have an influence in this company, the administrative processes will become easier when multi-site ERP is implemented for all the sites.
