Dead-On-Arrivals investigating indirect costs

Dead-On-Arrivals

investigating indirect costs

A case study for FAES Group

Final Thesis

MSc Technology & Operations Management

University of Groningen, Faculty of Economics and Business Januari 26, 2015

Joris ten Doesschate S1683799

2

Abstract

The research investigates the impact of indirect costs, concerning the handling of Dead-On-Arrival (DOA) spare parts in the tech industry. A case study is conducted at two high-tech, Original Equipment Manufacturing, companies focussing on the impact of indirect costs of Dead-On-Arrival spare-parts. DOA’s are spare parts that arrive at the customer in a non-functional state. Indirect costs include logistics, diagnostics, analytical, preventive and managerial costs. Data is collected via detailed interviews with various departments in order to build an overall DOA cost model. It can be concluded that indirect costs for DOA’s are roughly half of the direct costs and have an impact on the overall costs of doing business.

3

Content

2.2. Maintenance spare parts logistics ... 13

2.3. Activity Based Costing ... 15

2.4. Total Quality Management costs ... 17

2.5. Model building ... 18

3. Method ... 23

4. Data Collection ... 24

5. Results ... 26

5.1. Results on DOA information ... 26

5.2. Results from interviews ... 27

5.2.1. General remarks from both companies ... 28

5.2.2. Specific remarks company 1 ... 28

5.2.3. Specific remarks company 2 ... 29

5.2.4. Remarks per function for the process model ... 29

5.2.5. Quantified results ... 33

5.3. Process model update ... 36

5.4. DOA cost model results ... 39

5.5. Results on cost mapping ... 46

6. Conclusion & Recommendations ... 51

6.1. Conclusion ... 51

6.2. Recommendations ... 53

7. Discussion and suggestions for further research ... 54

8. References ... 55

4

Preface

This master’s thesis presents the results of a three month research into the indirect costs of Dead-On-Arrivals in the high-tech industry. Before I started this research, the concept of Dead-On-Arrivals was unknown to me, nor did literature give me a quick start into the subject. This is probably a reason for the relatively slow start to this research, it took a little while before I was acquainted with the subject.

During this research I have learned a number of things, both in the field of my studies, and about myself. Where the final year of the Technology and Operations Management study gave me enough knowledge and skills to pass all the courses successfully, it has also prepared me for this master’s thesis. However when starting this project I could not have suspected so many difficulties along the road.

I would therefore like to thank all the persons involved in making this research a success, and supporting me in the more difficult times. Firstly I would like to thank the FAES Group and in particular Bart van Dijck for assisting me throughout this research, introducing me into the field of Dead-On-Arrivals, introducing me to the contacts at the two case study companies, and finally thanking him for his valuable comments during our discussions on updates of my work. I would also like to thank the contacts at the two companies, Walter and Ruud, for receiving me, and introducing me to others involved in the DOA process.

I would also like to thank professor W. Alsem for providing me with valuable feedback and suggestions. Both from an academic point of view as well as based on his long experience in industry and at the university.

Finally I would like to thank my girlfriend and parents for their support during the last few months. It was nice to be able to discuss certain problems with the persons closest to me.

Joris ten Doesschate

5

1. Introduction

The occurrence of Dead-On-Arrivals (DOA’s) can ruin long term supplier – customer relations. DOA’s may reduce customer satisfaction and also spoils efforts to manage after-sale service. A DOA occurs when a critical spare part is received in a non-functional state. In other words, a DOA is a newly installed replacement part that does not function as it should and may result in a longer downtime of the customers equipment. A spare part is thus Dead-On-Arrival when it is received in a non-functional state. A DOA can also be qualified as a zero-hour failure (Petkova, Sander, & Brombacher, 2000). In certain business conditions, a machine or piece of equipment needs to be repaired as fast as possible in order to avoid long downtimes. Therefore the availability of spare parts to repair a customer’s machine is an important part of after-sales services. If however a critical part after the initial replacement appears to be a DOA part, needed for the basic functioning of a machine it could lead to a distressed production system and eventually to less satisfied customers.

In order to keep customers happy, machines need to be producing or working according to its intent and not waiting for a spare part in order to be repaired. Not only can spare part management be expensive if not managed correctly, it also is an important element for the quality of after-sale services. In high-tech industries, in which failure of parts can have huge implications on the production process, companies need to offer high availability of spare parts and rapid repairs (Goffin, 1999). Original Equipment Manufacturing companies are under increasing pressure in the after-sales services because after-sales services have been identified as a major source of increasing profitability and competitive advantage (Murthy, & Kumar, 2000). Important components of the after-sales service support according to them are warranty provisions, repair services and maintenance strategies. The importance of a good spare part management strategy could be a very good addition to this list. According to Huiskonen (2001), spare part inventory management can be considered as a special case of general inventory management, with some special characteristics; the prices of individual parts can be very high, demand for parts can be sporadic, and service requirements are very high as the effects of stock-outs may be financially remarkable.

6 DOA’s cost around 4% of the total spare part turnover and that preventing DOA’s can lead to 20% more net profit. The main causes of DOA’s were investigated and found to be; part quality, service quality and logistics. These causes will be discussed in more detail in chapter 2.1. In the research of Van Swieten (2013) it was concluded that part damage, transportation and inventory policy were the most important direct DOA cost components. It was furthermore suggested to investigate the indirect / overhead DOA cost components “The costs of overhead such as managers and DOA analysis teams are expensive and might imply substantial indirect costs for the company considered” (Van Swieten, 2013). She furthermore mentioned that indirect costs might have an even greater financial impact. Therefore the indirect costs are the focus of this research. Van Swieten (2013) also suggested having a further look at “image” costs, the impact DOA’s might have on the customer. Image costs can be hard to quantify since they include company proprietary information which is typically not distributed externally and that is why, they are not within the scope of this research. Only an indication of what image costs are will be presented in this research. A schematic overview of DOA costs can be found in figure 1.1.

Figure 1.1: The DOA cost model as presented by Van Swieten (2013).

7 large amount of money to reduce the amount of DOA’s and thus keep their customers satisfied. It is therefore important to get a better understanding on the magnitude of these costs due to the occurrence of DOA’s.

That indirect costs can have a significant impact is supported by (Cohen, Zheng & Agrawal, 1997). They investigated 9 high-tech companies in the computer industry, in which they found that, on average, 29 percent of total operating costs are in inventory investments, 8 percent is allocated to transportation (regular & emergency), and 32 percent to warehousing and scrap. The costs for inventory investments, warehousing and half (emergency) of the transportation costs can be seen as direct costs. They are very similar to the direct costs from Van Swieten (2013). According to the research of Cohen et al. (1997) the other 31 percent of total operating costs include the indirect costs such as administrative, personnel and other miscellaneous costs. However their research focused on total operating costs, not specifically on DOA spare parts. In the field of DOA’s there has hardly been any published research, it is therefore interesting to test if these numbers hold, in handling with DOA parts.

To get an understanding of the scope of this research figure 1.2 presents an overview of an Original Equipment Manufacturer supply chain.

Figure 1.2: An overview of the OEM supply chain.

8 can find out that the newly installed part is DOA. Again a new spare part will need to be ordered. This new part might need to be shipped form one of the warehouses described above, before the part can be installed a second time. The actions described before, are seen as “direct” costs by van Swieten (2013). However, a large amount of “indirect” costs can also be described. The FSE needs to fill in an extensive report about the DOA, the DOA part needs to be send to an inspection/repair shop after de-installation and from there either back to a warehouse, to the original part supplier or even be scrapped. At the OEM extensive analysis of the DOA part needs to be executed, and management gets involved in the aftermath of a DOA, through customer service and meetings.

The purpose of this research is to provide an analysis on the indirect cost components related to Dead-On-Arrival spare parts and to suggest opportunities for improvements. A full DOA cause and effect analyses will not be part of this research. This research will extend on the research into the direct costs related to a DOA by Van Swieten (2013). This research will be conducted at two different companies in the high-tech products market, these two companies are two of the three companies that were studied by Van Swieten (2013). This way the indirect costs could be compared with the direct costs from the same companies. The high-tech industry has been chosen because in this industry, failure of parts can have significant financial and process related implications on the production process (Goffin, 1999).

1.1. Research question

9 The following research question has been established; “What is the impact of indirect costs concerning the handling of Dead-On-Arrival spare parts in the high-tech industry?”

Sub-questions for this research are as followed;

 What is a Dead-On-Arrival spare-part?

 What are the different activities within the OEM organisation?

 How much time is spent in the different activities on DOA’s?

 What are the indirect costs in the DOA handling process?

 What are possible improvements to reduce the costs of DOA handling? 1.2. Conceptual model

This section describes the conceptual model for this research. The elements (boxes) from the model as shown in figure 1.3. The model presents the direct and indirect DOA costs. The pink boxes are indirect costs, and the grey boxes are direct costs. Some boxes contain both direct and indirect costs and therefor consist of both colours.

Figure 1.3: A conceptual model of DOA costs.

10 The indirect costs in this model are the costs of analyses and R&D, logistics costs, management meetings, service costs, and image costs. Indirect logistics costs, are all other transportation costs that were not included in the research of Van Swieten (2013), she only took into account the costs of emergency shipping of spare parts. Other logistics costs included, are the costs of coordinating all DOA part shipments. The indirect service costs are the costs of customer service, either through a separate department of through a field service engineer. The final indirect costs are the image costs and the costs of non-conformance to the customer. Only an indication of what these image costs are will be presented in the research.

1.3. Outline

11

2. Theoretical framework

The aim of the theoretical framework is firstly to provide a broader understanding of the concept of Dead-On-Arrivals, maintenance spare parts logistics, Activity Based Costing, and total quality costs. Secondly, these theoretical findings will be used to build an initial process model.

2.1. Dead-On-Arrival

Based on previous research done by Van Swieten (2013) at three companies it became clear that the phenomenon of Dead-On-Arrivals spoil the efforts to manage after-sale services to maintain customer satisfaction. A product is called Dead-On-Arrival when it arrives in a non-functional state. It can also be called a ‘zero-hours failure’ (Petkova et al., 2000). However, very little research has been conducted in the field of DOA’s related to spare parts management.

From the work by Van Swieten (2013), three main causes of DOA’s were found; logistics, service quality and part quality. Logistics concerns parts arriving ‘non-functional’ due to damages during transportation, stocking and packaging. DOA’s due to service quality refer to the capabilities of the field service engineer handling the spare parts. Possible DOA causes may be bad training and instructions or use of wrong tools. Thirdly DOA’s could be caused by part quality, spare parts in this industry are generally fragile products due to used materials and design, and part quality DOA’s can also occur due to lack of intensive quality inspections.

12 When a spare part does not function at all it is said to be completely dead on arrival. This is the most severe state of a DOA, mostly the spare parts are not repairable by the service engineer, and even if they are, this would be at a great cost. A completely dead part can either be due to hardware or software issues.

When a spare part is said to be cosmetically unacceptable it can still function however it may be scratched, dented, in need of cleaning or come in unacceptable packaging. Overall costs for these kind of DOA’s are lower than for completely dead DOA’s, especially if the spare part only needs to be cleaned, or the box it came in is smashed. From Van Swieten (2013) it appeared that not all of these cosmetically unacceptable parts are DOA’s, however there are costs involved before the spare part can be properly used, it cannot be used immediately and needs additional handling before installing (Papadopoulos, 1996). Therefore in this study it will be classified as a DOA.

Finally a spare part can be non-functional due to delivery issues. When a spare part is needed, but not available due to logistical issues it can be said to be DOA. Spare parts falling into this category can be a wrong product delivered, a missing part, or an extremely late delivery. Similar to the cosmetically unacceptable DOA’s, there are costs involved in this category before the correct part can be used. An item that is delivered late can be difficult to quantify as a DOA, since there is no general definition of what is acceptable as ‘late’. Based on practice a timeframe would need to be set on this. Service contracts will in a number of cases state acceptable delivery terms.

Summarising on the different levels of DOA, a general definition of a DOA can be presented as follows; “A Dead-On-Arrival refers to a spare-part, sent by a supplier or manufacturer,

which is received by the FSE or customer in a non-functional state, i.e. broken (dead), cosmetically unacceptable or not timely usable due to the deficiencies in the delivery process.”

13 spare part is either correctly or incorrectly labelled as DOA upon first installation, but during a second test no fault could be found (Söderholm, 2007).

2.2. Maintenance spare parts logistics

This section will focus on operational control characteristics of maintenance spare parts. The focus of this section is to give an insight into the classification of DOA spare parts. In general it can be argued that spare part inventory management is a special case of general inventory management, with some special characteristics; the prices of individual parts can be very high, demand for parts can be sporadic, and service requirements are very high as the effects of stock-outs may be financially remarkable (Huiskonen, 2001). There are four control characteristics considered most relevant; criticality, specificity, demand and value. These four will be discussed in more detail below.

The criticality of a part is measured by the consequences caused by the failure of the part on the process. According to Huiskonen (2001), the impact of a missing critical spare part can be a multitude of its commercial value. It can be hard to determine these costs, although in practise it is mostly evaluated by the downtime costs of the process. According to Cohen et al. (1997), part criticality can be seen as the impact of a parts failure on the operation of the product.

Spare parts can said to be either standard parts or tailored specific parts. For standard parts the availability is generally good, stocks of these parts can generally be found at different locations in the supply chain. For the more specific parts quite the opposite is generally true. These special parts are most of the time make-to-order and therefore lead times can be very long (Huiskonen, 2001).

The demand pattern of parts includes the aspects of predictability and volume (Huiskonen, 2001). For high-tech industry DOA parts there typically is a wide variety of parts with very low and irregular demand patterns. Combined with other characteristics such as high criticality and high price this generally results in an increase in the amount of safety stocks needed to cover unpredictable situations.

14 so that administrative costs, in order to get the spare part to its destination, do not increase un-proportionally to the value of these items.

Combining these control characteristics for the high-tech industry a number of suggestions can be made related to DOA’s. In the high tech industry, normally a Dead-On-Arrival spare part is high value, high criticality, the machine of the defect part has to be repaired as fast as possible, and the spare parts should be supplied immediately. Furthermore it can be said that DOA parts are usually very specific make-to-order parts with long lead times, relatively high prices and low volumes. Combining these two characteristics hints that in most cases a certain safety stock should be kept. According to Huiskonen (2001), safety stocking becomes necessary for all the parts for which lead times are longer than the time to tolerate a stock-out situation in case of failure. Looking at this from a cost perspective a company will need to choose between the costs of keeping a large enough safety stock to handle a number of Dead-On-Arrival spare parts, and the expected downtime costs of the process due to long lead times of the specific spare parts. Maintaining high stock levels however will result in significant purging costs in case a specific part is redesigned or replaced in the production process (Östlin, Sundin, & Björkman, 2009).

In order to streamline (reduce costs) the organization and management of inventories, the ABC classification (Teunter, Babai & Syntetos, 2009) can be used by companies. Although detailed analyses of the total costs of spare parts for the individual companies is not part of this thesis, the method described by Teunter et al. (2009) to optimize (minimize) costs at a certain service level provides a very good and simple method for analysis to the companies interviewed as part of this thesis.

The authors state that ranking classes of parts according to the criteria below would provide a very good option to reduce overall part costs. The following formula is used. The Cycle Service Level CSL is defined as: CSLi = 1 – (hi * Qi ) / (bi * Di).

If we apply this to the DOA parts the following can be stated:

 CSLi: cycle service level for a certain DOA or DOA group.

 hi: inventory holding costs (per item per time unit).

 Qi: average (order) quantity = 1 based on the interviews. DOA parts are typically

ordered on an individual basis.

15

 Di: demand per unit time – Needs to come out of detailed analyses.

Ranking DOA’s based on criticality according to the ratio: Penalty costs * Demand per unit time / inventory holding costs per time unit could be a great way to define cost reduction opportunities, starting at the lowest of these ratio’s.

2.3. Activity Based Costing

This section explains why an Activity Based Costing (ABC) approach is taken in this research. Firstly the difference between traditional costing and ABC will be discussed. Secondly, ABC will be analysed in more detail, and finally the ABC tree concept, for quantifying the indirect costs, will be discussed.

Traditional costing systems are known to distort the cost information by using traditional overhead allocation methods. Allocating overhead costs to products as a proportion of direct labour costs or volume of production (Ben-Arieh & Qian 2003; Rezaie et al., 2008). However, when a certain product involves high labour costs, this does not directly suggest that this product also has high overhead costs such as purchasing, planning, legal, inspection and quality costs. The above indicated traditional method, might lead to less accurate cost allocation and representation. Activity Based Costing (ABC) systems provide a more accurate cost estimation and calculation method that overcomes the weaknesses of the traditional costing systems and enables for a more accurate allocation of overhead costs to individual products and or activities (Nicholas, 1998). ABC systems are different to traditional costing systems in two ways. Lewis (1995) states that Activity Based Costing is often used as a part of total cost management. Cost pools can be defined as activities instead of production cost centres and, the cost drivers used to assign activity costs are contrasting to those used in traditional cost systems.

16 The ABC method links the usage of overhead costs incurred to the amount of production. The fundamental elements in ABC are activities, cost objects and cost drivers. The ABC method identifies the activities as any kind of task consuming resources such as products and services. Using ABC can lead to classifying activities as value-added and non-value added, and eventually provide a basis for elimination of the non-value-added activities (Gunasekaran & Sarhadi, 1998). A cost object is the output of the activity and or the place to which the activity costs are allocated. And finally, a cost driver is a measure of how much an activity is used by a cost object, this can be the number of units produced, total labour hours, hours of equipment time, or the number of orders received. These cost drivers determine which part of the total costs per activity are allocated to a specific cost object (Ben-Arieh & Qian, 2003; Nicholas, 1998).

Ben-Arieh & Qian (2003), provide an overview of the advantages and drawbacks of Activity Based Costing. They state that advantages of ABC can be providing timely cost information suitable for decision making and allowing more detailed tracking of indirect cost-to-cost objectives, as well as improving the accuracy and relevance of products costing. A possible drawback of the ABC method is that the method is not driving companies to change their fundamental views about how to organize work and to satisfy customers efficiently (Ben-Arieh & Qian, 2003).

17

Figure 2.1: The product cost tree concept as described by Rezaie et al. (2008).

2.4. Total Quality Management costs

This section will cover some aspects of Total Quality Management (TQM). The insights from this section will mostly be used in building the initial process model. TQM is an integrated management philosophy that emphasises, among others, continuous improvement, meeting customers’ requirements, and reducing rework (Ross, 1994). TQM can thus generate; reduced costs, more satisfied customers and improved overall financial performance. TQM literature provides multiple opportunities to learn about total costs. All three TQM gurus; Crosby, Deming and Juran (Powell, 1995) describe quality costs as the total of the costs incurred by:

 Investing in the prevention of non-conformance to requirements.

 Appraising a product or service for conformance to requirements.

18 More specifically they split the above in the following sub categories.

Prevention costs are the costs of all activities specifically designed to prevent poor quality in products or services. Examples are; quality planning, supplier capability surveys, and quality improvement team meetings.

Appraisal costs are the costs associated with measuring, evaluating or auditing products or services to assure conformance to quality standards and performance requirements. An example of appraisal costs are incoming and a source inspection/test of the purchased material.

Failure costs, the costs resulting from products or services not conforming to requirements or customer/user needs. Failure costs can be divided into internal and external failure categories. Internal failure costs are costs occurring prior to delivery or shipment of a product, or the furnishing of a service, to the customer. These can be costs due to scrap, rework, and re-inspection. External failure costs are failure costs occurring after delivery or shipment of the product and during or after furnishing of a service to the customer. Specific costs are; the processing of customer complaints, warranty claims and product recalls.

The sum of the costs noted above are the total quality costs. This represents the difference between the actual costs of a product or service and what the reduced costs would be if there were no possibility of substandard service, failure of products or defects in their manufacture.

2.5. Model building

In this section the process of building the initial process model will be described in more detail. Based on the suggestions by Van Swieten (2013) the possible indirect costs due to a DOA could be a multiple of the direct costs. In her study, the focus was on the direct costs of the Field Service Engineer, the spare part itself, the total inventory of spare parts and emergency transport of spare parts. However based on her conclusions, it can be assumed that there are multiple departments involved in the handling of a DOA spare part.

19 compared to the warehouse and repair rooms from figure 1.2. Their research thus hints that indirect costs can be found at the customer, at the Original Equipment Manufacturer (warehouse and repairs) and at the external supplier. As this research is conducted at the OEM, the focus is on the indirect costs concerning the OEM only.

Based on factory and corporate support costs (Cooper & Kaplan, 1988), a number of costs were found to be indirect costs. These costs are; logistics, marketing and sales, distribution, manufacturing information systems, quality management, and general administration (Cooper & Kaplan, 1988; Lambert, Cooper, & Pagh, 1998). These costs are typically made in the following departments; field service engineering department, logistics department, research or engineering investigation group, sales and/or customer relation contact, and management. In order to collect data and build a first process model, the “Service Response Time” model as designed by Cohen et al. (1997) and used by Van Swieten (2013) was expended for this research. According to Cohen et al. (1997), the service delivery process is started by a customer call, which is followed by a visit from the field service engineer to the customers site.

Figure 2.2: The “Service Response Time” model by Cohen et al. (1997)

20 repair calls are initiated by the customer. Despatchers at the original manufacturer receive the calls, on-line service is provided where this is possible. The on-line service may be compared to the remote diagnosis from the service response time from Cohen et al. (1997). If on-line service is not the effective, a technician is assigned. The technician proceeds to schedule a visit, and reports to the location to diagnose and repair the equipment. Availability and quality of the spare parts define when the customer equipment will be available again. The service call is cleared after the repairs are completed. Next the after-sales services are initiated with the billing and feedback processes. The final conclusion from Behara and Lemmink (1997) was that “The level of customer service provided will depend on how effectively the technicians are supported by their despatcher, local managers, and the company’s spare parts unit”.

To further expand the process model, the research of Gross and Pinkus (1979) was used. They present a simple overview of the repair process. When a spare part has failed, the part is removed by the FSE as discussed above. The removed part will then be transported to a repair facility, analysed, repaired and dispatched to a new spare part warehouse. Simultaneously a new spare part needed for the repair of the machine is requested from a warehouse and used if available. If there is no spare part available, the part will need to be backordered.

The final part of the model is based on the total quality management costs as described by Wood (1999) & Powell (1995). They both described that total quality management costs consist of prevention costs, appraisal costs, and failure costs. Prevention costs are included in the initial model in the form of evaluations with all departments. Appraisal costs are an essential element of the indirect costs in the process model. These are costs for product and process evaluations or service audits. Internal failure costs in the form of re-inspection/testing are included in the model, external failure costs are included in the model in the form of processing customer complaints.

21 they are interrelated (Nicholas, 1998; Damelio, 2011). Each swim lane represents a department. The reason for building a flow chart is supported by (Nicholas, 1998). He states that part of the first step in building the ABC cost tree is to prepare a detailed flow chart of every step in the process.

Based on the elements discussed above, the following initial process flow model was developed. The complete model can be found in Appendix 1. The model can be split in the following three sections: normal maintenance, direct costs and indirect costs. These will be discussed in more detail below.

The normal maintenance process contains steps or activities that take place when a machine at a customer has to stop due to a failed part. Customer service receives a notification that a machine needs to be repaired. The FSE schedules a visit, and proceeds to the location to diagnose and repair the equipment. Spare parts are send to the FSE through coordination of the logistics department. Afterwards the machine is tested, and it either works or a new diagnoses needs to be performed. The normal maintenance section can be found in figure 2.2.1.

Figure 2.2.1: The normal maintenance section of the initial DOA process model.

22 diagnoses a DOA part, checks for spare parts and continues to repair the machine until it is working again.

Figure 2.2.2: The direct costs section of the initial DOA process model.

The indirect costs related to a DOA are costs the original equipment manufacturer makes in order to analyse and evaluate the DOA. Customer service provides the customer with updates about the repair process. If a required spare part is not available in a warehouse, it will need to be produced. The research department, investigates the DOA part sent by the FSE, and decides to repair or dispose of the parts. Management is notified that a DOA has occurred. The final step from all departments is an evaluation meeting about the DOA.

23

3. Method

This research will predominantly be based on two high-tech companies. Case study research aims to provide the researcher with a set of theoretical generalisations based on observations, “constraint testing” of these generalisations, and eventually validating theory (Bonoma, 1985).

The case study companies are from the high-tech sector. This sector was chosen as after-sale services such as spare part management and maintenance services are said to be significantly more substantial in these businesses than in others (Cohen et al., 1997). This and the fact that both companies offer high quality, high value parts makes this industry very suitable for investigating the cost effect of DOA’s. The companies in this research will be two of the same companies from the research of Van Swieten (2013). The two case study companies are high-tech companies from the Eindhoven region, the Netherlands.

This research will be threefold. Firstly this research is an exploratory study. Based on previous research a DOA definition has been established. Preferably this definition will be used again, however it is tested both in literature and with the two companies. However it is not the intention of this research to provide a detailed analysis into the cause and effect of DOA’s. In chapter 2, a background study has been conducted into the Activity Based Costing method, total quality costs and maintenance spare part logistics. Chapter 2 further includes an initial mapping of the complete process of DOA spare parts handling.

Secondly, a descriptive method will be used. This part will be based on interviews at both companies. The purpose of this part is to test the correctness of the initial process model. This will be done through interviews with different people at different departments at the two companies. The purpose of this section will be to make a generalizable model fitting the processes of both companies.

24

4. Data Collection

Most of the costs related to a DOA that have not been captured by Van Swieten (2013) are based on hours worked. At first there did not seem to be an obvious reason why these costs are very strongly dependent on the costs of the spare part. However, the ABC method allocates indirect costs to a product in order to identify its true costs. So the basis for collecting the data is to look at the hours spend for the various activities and the costs for the employees executing these tasks. During the cost analysis these hours worked will then be translated in the costs of a spare part using the ABC method.

Given the various departments that will be involved in the process following a DOA it was decided to collect the data via interviews with key players in these departments. To make sure that the interviewees understood the background of the questions and since it was unclear if they would have any idea of the content of the Van Swieten study a PowerPoint presentation was prepared. The presentations first slide showed the process model and the direct costs tested / discussed by Van Swieten (2013), it also covered the current focus and the departments involved. The second slide in the presentation covered an overview of the findings by Van Swieten (2013); the cost of the spare part, the logistics costs to get the parts delivered to the customer and the inventory costs required to ensure customers could be supplied with most spare parts.

The following slides of the presentation covered questions related to the various departments that were targeted to be interviewed. These departments were the:

 Field service engineering department

 Logistics department

 Research or engineering investigation group

 Sales and/or customer relation contact

 Management / DOA responsible contact manager

25 The generic questions were:

 What do you think about the process model, is it correct, which steps are missing / incorrect?

 What is your role in the various DOA steps?

 How much time does this cost you?

 How is communication in the process organized?

 How is documentation and evaluation executed?

 How is the relationship with the customer managed?

Some of these questions could have been focussed on one or two departments only, but they were still selected to be generic and discussed with all interviewees to make sure potential discrepancies were highlighted and could be discussed directly. The generic questions were discussed because the belief was that these questions were going to have the biggest impact on the mapping the indirect costs in the process model.

The last slide covered questions about the cost of people involved in the process. This to make sure the results could be converted using the ABC method, from hours worked during the various activities into relevant cost.

26

5. Results

This section will discuss the results. Firstly the DOA definition was tested. Next the remarks from the interviews are presented. These remarks are used to quantify the costs and update the process model. The updated DOA cost model will be discussed, and this sections ends with the results on cost mapping.

5.1. Results on DOA information

This section will discuss the renewed DOA definition from one of the two companies, together with a short analysis on DOA causes.

One of the two companies had redefined their DOA definition. They say that a DOA part is a spare part that doesn’t work according to specification, at the moment it is installed, or within 24 hours after installation. Compared to the DOA definition stated in chapter 2.1:

“A Dead-On-Arrival refers to a spare-part, sent by a supplier or manufacturer, which is received by the FSE or customer in a non-functional state, i.e. broken (dead), cosmetically unacceptable or not timely usable due to the deficiencies in the delivery process.”

This company’s specific definition doesn’t specify why a part does not work to specification. However it does state that a spare part is classified as a DOA at the moment of installation, or within the first 24 hours, which was missing from the original definition. The definition there could be altered in the following manner:

“A Dead-On-Arrival refers to a spare-part, sent by a supplier or manufacturer, which is in a non-functional state, i.e. broken (dead), cosmetically unacceptable or not timely usable due to the deficiencies in the delivery process, upon installation by the FSE or customer or within the first 24 hours after installation.”

27 failure might be overlooked and a new spare part is installed. The machine is started and fails again after a couple of hours. The root cause is still the electrical fault somewhere in the machine, however the newly installed spare part might be qualified as a DOA with the indicated definition. In addition, when testing this newly installed part, a NFF could be found.

Van Swieten (2013) found three main causes of DOA’s; logistics, service quality and part quality. Based on the data from one company on DOA causes it can be said that 17% of DOA’s is related to logistics, 44% is related to service quality and 39% is related to part quality. The biggest contributor are DOA’s caused by faults made by the FSE, which are reported to be a result of less than adequate training and incorrect or insufficient instructions. This accounted for 30% of total DOA’s. It should be noted that only 2/3 of the total amount of DOA’s from 2014 had been analysed, the final numbers could therefore be a little different. Further analyses reveals that 10% of analysed DOA’s are related to procedures, 28% of the analysed DOA’s is related to material and design problems, and that an impressive 62% is related to human error and human behaviour. So even in the high-tech environment of these companies, more than 60 percent of DOA’s is still caused by human error. In addition, the other company confirmed that a number of DOA’s arise due to faulty diagnosis and repairs of the FSE. Based on these findings it would be very interesting for the companies involved to perform extensive research on the possible causes of their DOA’s.

5.2. Results from interviews

28

5.2.1. General remarks from both companies

All persons interviewed expressed the great importance of customer satisfaction. In their mind, it is far more important to keep a customer satisfied than it is to lower the costs spent on DOA analysis. If a customer suffers a number of DOA’s in a short period of time, they might lose confidence in the OEM as a supplier and end the relationship, resulting in a far greater loss. Moreover, both companies expressed that DOA’s are still not the main point of focus, normal maintenance procedures and failed parts gain still most of the companies attention. Finally both companies stated that their number of DOA’s is slowly decreasing. However the cost of further analysis might increase significantly. The most common and easy to correct DOA causes have been found and corrected first. Afterwards the less common DOA's are left, but these DOA’s are less generalizable and have more different causes. Therefore, both more research must be done and time must be spent on every single DOA. However, nobody interviewed had a clear and complete view on the total costs of DOA’s.

5.2.2. Specific remarks company 1

29

5.2.3. Specific remarks company 2

This company makes a premier differentiation between parts for which they own the design and parts which are purchased from their external suppliers. About 90% of parts are purchased from external suppliers. When a supplier part is classified as a DOA, this company only performs trend analysis but they leave root cause analysis and repairs to the external supplier. Their own parts however are fully analysed in-house. Furthermore they differentiate between quality DOA’s and logistical DOA’s. Quality DOA’s are DOA’s due to the non-functionality of the part. Logistical DOA’s are due to problems during handling, shipping or storage, or because the wrong part is sent. In total this company has about 1% DOA parts from a total of 1.2 million shipped parts, which is around 12000 DOA’s per year. Their main driver for lowering the number of DOA’s is to improve customer satisfaction.

5.2.4. Remarks per function for the process model

When looking at the initial process model from chapter 2.5 a number of changes were suggested during the interviews. These changes and the required resources, people and money, for the various process steps will be discussed per function as they are shown in the updated process model in section 5.3. A list of the major changes compared to the initial model can also be found in section 5.3.

Field Service Engineering

30

Logistics

A large enough safety stock is kept in warehouses worldwide so that parts can be shipped to a customer’s location within 24 hours. Under normal conditions, sufficient spare parts are available, so no production of new parts is needed once a DOA is found. Company 2 revealed that to reduce the risk of incorrect first investigation by the field service engineer, a number of different parts for a certain repair is shipped to the FSE, not all of which will be used. This is due to the company’s strategy to move parts around frequently as part of their goal to have the correct spare part available at any moment. Every time the unused parts need to be repacked and shipped again. Although it was quoted that some parts could be re-shipped up to 10 times, costs for two times on average has been used for the analysis.

After the machine is repaired by the FSE, both the original failed part and the DOA parts are shipped back to the factory (company 1) or a local warehouse (company 2) for final analysis or further shipment to a location where the correct test can be performed. For company 2 most of the parts are shipped from the local warehouse to an external supplier and they are returned after they are repaired. The shipment of normal failed parts is outside the scope of this research, only the indirect costs related to the DOA parts have been quantified.

Customer support

Together with the FSE, customer support does the intake (first diagnosis) of the DOA part. This is mostly about gathering information about the problem and documenting the information concerning the DOA. For company 1 there are a number of zone quality managers responsible for the follow-up with customers. They are visiting and evaluating with customers who have suffered a severe DOA. Company 2 hinted that most of their machines are connected to the internet, and that customer support can perform a first on-line analysis of the failed machine before sending a FSE with spare parts to the customer. Both companies interviewed indicated that there are a number of contacts within their customer support department working fulltime on DOA’s and the logistics required to solve these.

Research / autopsy

31 intercontinental warehouse and are analysed at this warehouse. For company 1 this is mostly done in combination with the FSE and a local service team, after which the parts are shipped to the main factory for a second analysis.

The second task is to classify all reported DOA’s, for company 1 this results in a general trend analysis. Company 2 makes a top 20 list of reported DOA. Every few weeks an update is produced, which is used for the DOA reflection day.

Based on this trend analysis it may be decided that a certain part needs to be redesigned, and all current parts stored in warehouses need to be called back to the factory in order to be adjusted (stock purging). Adjustments can be quality and/or logistical adjustments to a part due to design errors, adjustments to the packaging or the instructions for the FSE. Both companies state that safety risks for the customer are a major reason for adjustments. Company 2 makes a further differentiation between cheap and expensive parts. All expensive parts are analysed on an individual basis. For the cheaper parts a trend analyses is done based on the parts failing and when a trend has been spotted, those cheaper parts will be analysed in more detail. Both companies interviewed have test rooms for detailed analyses of electronic parts.

Given the high level of own parts and its relative high value, company 1 indicated that as a result of learnings from DOA’s, they are replacing, purging, around 60 different parts from their warehouses. This comes at a cost of on average 25.000 euro per serial number. These costs include logistics, handling, rework and administrative costs. The estimated costs of stock purging are therefore 1.5 million euro on a yearly basis. Company 2 also performs some bin checks, ranging from mandatory actions due to safety situations, to updates through service recommendations. However, not enough data was available to quantify these costs. Because a number of DOA prevention actions were reported, it was decided to split research and preventive actions in to two separate groups. These can be seen in table 5.2.

External supplier

32 For company 2, 90% of the DOA parts are shipped back to the external supplier, who will perform the analysis and conduct repairs to the part. However it can happen that the supplier contacts the analysis department of the company to discuss if and by whom the part should be repaired. Company 2 indicated that one person per business unit spends a significant part of their time tracking issues and discussing those with the supplier. These supplier analysis costs can be seen as preventive costs however they are caused by DOA’s in the first place. Therefore the costs are included in the indirect costs overviews.

Management

Both companies have a DOA reflection day every month or 6 weeks. Company 1 invites around 30 people to join this day to reflect on all DOA’s from the past period. All involved persons need to prepare, and travel to the day’s location which typically consumes a minimum of two full days per participant.

Next to these meetings, most of the other departments involved are led by a project leader and a higher manager, both of which spend a bit of their total time on dealing with DOA’s.

Costs for senior management involvement have not been included. It is obvious that when there are significant issues in the relationship with a customer they will be involved. It may also be assumed that when it is necessary that senior management gets involved, that a lot more employees throughout the company get involved. However none of these situations were addressed during the interviews.

Other results

The DOA reflection day together with the follow-up from customer support can be seen as part of the evaluation blocks from the original process model.

In addition to making sure that everybody that needs to know is being updated on any DOA situation, part of the DOA day is to come up with possible improvements for further reducing the number of DOA. Designing preventive actions is also part of the job of the research department.

33 Table 5.1 presents some specific company data, that will be used to analyse and compare the data, such as an overview of the total number of spare parts, the number of DOA’s, the percentage of total parts that is a DOA, and the average value of a single DOA part. Also the company’s revenue of spare part activities is included. The average part value and the revenues of spare part activities are based on the study by Van Swieten (2013). The total number of spare parts, the DOA percentage and the number of DOA’s per year is based on the interviews conducted for this study.

Company 1 Company 2

Total number of spare parts 67.000 1.200.000

DOA % 1,5% 1%

Number of DOA parts per year 1.000 12.000

Average DOA part value (euro) 15.164 710

Average part value (euro) 8.415 394

Revenues of spare part activities (euro)

354 million 439 million

Table 5.1: General DOA information from the two companies.

5.2.5. Quantified results

34

Company 1 Company 2

Field Service Engineer 2 FTE 7.5 FTE

Logistics 1 FTE + 1 mEuro 5.2 FTE + 2 mEuro

Customer service 7.2 FTE 1 FTE

Research / Autopsy 20 FTE 17.2 FTE

Prevention 1 FTE + 1.8 mEuro -

Supplier - 3 FTE

Management 3 FTE 2 FTE

Total FTE + Costs 34.2 FTE + 2.8 mEuro 35.9 FTE + 2 mEuro

Table 5.2: Overview of FTE and costs per function for both companies.

The results from table 5.2 will be explained in more detail below.

In order to calculate the number of full time equivalent (FTE) for the FSE, the number of DOA’s is multiplied by the amount of hours it takes to fill in a complete DOA report. Half of the reports only take up to 15 minutes to fill in, however in specific cases, it can take up to 8 hours. A calculated average of 1 hour is used for this research. The resulting total amount of hours is divided by the number of hours per year that is 1 FTE. 1 FTE is assumed to be 1600 working hours. For company 1 the number of hours and FTE is spend on this is doubled, as the FSE also does part of the analysis.

35 company ships their parts back and forth to an external supplier for repairs. It is assumed, that the external supplier pays half of the transport.

Costs for customer service for company 1 are a sum of multiple departments all of which spend a portion of their time dealing with customers. Customer contact is done through the FSE at company 2, however 1 FTE in the form of diagnostic, is involved in collecting all DOA reports.

Research costs for company 1 are based on the number of DOA’s (1000) multiplied by the average time needed for root cause analysis, which was on average 50 hours. The total hours per year was divided by the number of hours per year that equals 1 FTE. However only 20% of this time was actually used for analyses, resulting in roughly 6 FTE. Other research costs were 3 analysis teams for part quality (8 FTE), logistics (3 FTE), and trend analysis (3 FTE). For company 2 the number of FTE is based on 2 analysis teams (8 + 3 FTE) and a central autopsy team (4 FTE). Company 2 also has a software writing team aimed at DOA analyses (2 FTE).

Prevention costs are the costs of stock purging, and the clean room for company 1. The costs for stock purging were estimated at 1.5 million, the costs of the clean room are 300.000 euro per year, this includes man-hours. Not enough data for company 2 was available to specify the costs for bin checks.

As company 2 ships 90% of their parts to the original supplier, there is a coordination effort. Furthermore it is indicated that one person per business unit spends a significant part of their time tracking issues and discussing those with the supplier. A total of 3 FTE is involved in this.

36 5.3. Process model update

Based on the remarks from the interviews, the initial process models in chapter 2.5 have been altered. This results in the following process models, for which the major changes are stated with each model. The complete model can be found in Appendix 2.

Figure 5.1.1 presents the normal maintenance process, which are the activities that take place when a machine of a customer has to stop due to a failed part. When the failed part is replaced by a new part, the machine starts and the process ends at this moment. There are no DOA’s in this section. Compared to the initial process model (Figure 2.2.1) the major changes are:

 Notification of machine down comes in through the FSE at the customer support department.

That the FSE initiates the repair process was in line with the research of Behara and Lemmink (1997), however, it was found that the intake, diagnosis, is done in co-operation between the FSE and customer support.

 After repairs the machine is given free, restarted and tested.

 The costs for the return shipment part, receive parts and normal part analysis, are not part of this research. However, for completeness of the model they are part of the model.

37 Figure 5.1.2 presents the direct costs related to a DOA. Compared to the initial process model (Figure 2.2.2) the major changes are:

 When the machine does not work upon restart, a new diagnosis is performed. Only after this diagnosis it can be said that a certain part is DOA.

 New parts don’t need to be produced. They can be shipped from another warehouse. Both companies maintain such high stock levels compared to the failure rate of the components that they never have experienced that a spare-part is not available, this is in line with Huiskonen (2001), on keeping a large enough safety stock in their warehouse worldwide to handle DOA spare parts.

38 Figure 5.1.3 present the indirect costs such as research, analyses, and customer support related to a DOA. Compared to the initial process model (Figure 2.2.3) the major changes are:

 Diagnosis (intake) is done by both the FSE and customer support.

 DOA spare parts are analysed at a local warehouse and after that shipped back to the factory or to an external supplier.

 Each company took a different approach on the repair of the DOA part. One company repaired the parts at their own facilities the other company send the majority of parts to their external suppliers.

 Evaluation is not done as presented in the model. Once a month there is a DOA reflection day, and follow-up with customers is done by either customer support or the FSE.

 The research department may come up with preventive actions in the form of adjustments (redesign) or stock purges.

39 5.4. DOA cost model results

In this section the results from the interviews are used to extend the original DOA cost model (van Swieten, 2013) as indicated in the introduction of this thesis. This thesis is covering the “indirect or overhead” costs which are the results of a DOA. The indirect costs section can be split in a couple of segments. To convert people or FTE into cost, the FTE’s indicated in table 5.2 will be merged into the following 5 functions; logistics, diagnostics, analysis, preventive actions and management involvement.

 Logistics is the cost of coordinating the shipping of spare parts. Other logistics costs are costs of shipping DOA parts from the customer to the location of the analyses, and possibly to the external supplier for repairs. The parts also need to be shipped back to warehouses worldwide.

 Diagnostics costs are costs related to the Field Service Engineer and the customer support department. The FSE fills in the DOA report, and together with customer support they perform the intake. Contact with the customer is also part of these costs.

 Analysis covers trend analyses and autopsy and it are typically costs generated by higher educated employees. Also the costs to work with suppliers are included in this segment. These costs are the research / autopsy costs from table 5.2. The software team of company 2 are also seen as analysis costs.

 Preventive action costs are costs related to redesign of parts or software. The costs for the clean room from company 1 are also included here. The costs to purge stocks are also included. Furthermore the issues with the supplier, aimed at preventing new DOA’s are also included under preventive actions.

40 These 5 functions can be seen as the activities from the ABC cost tree from Rezaie et al. (2008), figure 2.1. As described above the main activities can be split in more individual cost objects. A schematic overview is presented in figure 5.2.

Figure 5.2: The indirect DOA costs can split in activities and cost objects.

In order to convert FTE into costs, typical salary costs for employees in the various functions were assumed. These costs include typical overhead costs such as social security, pension, housing and all other benefits to the employees. This resulted in the following assumed employee costs per year:

 Logistics / FSE: €75.000

 Diagnostic: €100.000

 Analysis / Preventive: €150.000

 Management: €200.000

Together with the total number of FTE per category, these employee costs can be seen as the cost drivers and unit costs from the ABC cost tree from Rezaie et al. (2008), figure 2.1.

During the interviews it became very clear that “image” costs are a very strong driver for both high tech companies. Typically included in image costs are;

41 example, assume a machine is down for 4 hours, and on average half of the DOA’s have an impact for the customer (DOA targets company 1, chapter 5.2.2), the downtime costs for the customer is 4 * 50.000 euro * 50% = 100.000 euro. Thus on average this results in 25.000 euro per hour. The assumption for 4 hours downtime is based on figure 5.1.2, it may take a while to perform the diagnosis, pick the spare parts and repair the machine. No specifics for customer downtime were given for company 2 given that their “lost profit” cannot be easily quantified given their specific applications.

 Customer satisfaction is the second element of image costs. These costs are hard to quantify and will surely not be released by any of the companies. But one can understand that in a highly competitive market as the high tech industry the potential loss of an essential customer could have serious impact on the profitability of the company involved. During most interviews this understanding was expressed in many cases.

Figure 5.3: The DOA image costs can be split in two categories.

42 The results on indirect costs discussed above leads to the following indirect costs overview as shown in table 5.3. Per function, the typical employee costs and the number of FTE are presented. Combining these, results in the total indirect costs per function. Other logistics costs and purging costs are only based on a cost total, and thus presented separately. Diagnostic costs consist of costs for the FSE and customer support. Analysis costs consist of costs for trend analysis and autopsy. The bottom row of the table presents the total FTE and total indirect costs spend on DOA analysis within these two companies.

Typical cost

(kEuro/Yr.)

Company 1 FTE Cost (kEuro)

Company 2 FTE Cost k(Euro)

Logistics coordination 75 1 75 5,2 390 Shipment 1.000 2.000 FSE 100 2 200 7.5 750 Customer Support 100 7,2 720 1 100 Total Diagnostic 100 9,2 920 8,5 850 Trend 150 3 450 3 450 Autopsy 150 17 2.550 14 2.100 Total Analysis 150 20 3.000 17 2.550 Preventive actions 150 1 150 3 450 Purging costs 1.800 Management 200 3 600 2 400 Total 34.2 7.545 35.9 6.640

43 To be able to compare the two companies, the indirect costs per function have been calculated as a percentage of total indirect costs. In this diagram the costs for logistics and other logistics, and the costs for preventive actions and purging costs have been combined again.

Figure 5.4: Relative indirect DOA costs per function.

When the same data is plot in a “100% bar” diagram the results are as follows;

Figure 5.5: Relative contribution of each function. 6% 7% 38% 13% 36% 8% 26% 40% 12% 14% Management Preventive actions Analysis Diagnostic Logistics

Indirect DOA costs per function as percentage of total

indirect DOA costs.

Company 1 Company 2 14% 36% 12% 13% 40% 38% 26% 7% 8% 6% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Company 1 Company 2

Indirect DOA costs per function as percentage of total

indirect DOA costs.

44 It can be seen from figure 5.4 and 5.5 that the percentage of indirect costs spent on management, analysis and diagnostics are very similar for both companies. With analysis, covering both trend analyses and autopsy, having the highest impact. These are typically costs generated by higher educated employees. Furthermore it can be seen that the prevention costs for company 1 are very high. This is due to the clean room and the large amount of high value stock purges. Purges that were necessary due to a part re-design which was the result of DOA analyses. Company 2 has on average a lot more parts, which get shipped around a lot more than the products of company 1, resulting in higher logistics costs.

When looking at the total quality costs as discussed in chapter 2.4, the costs are split into four categories; prevention, appraisal, internal failure, and external failure costs. A few interesting conclusions can be drawn. The DOA reflection meetings can be seen as prevention costs in the form of quality improvement team meetings, as these days not only focus on reflection but also on plans to lower the amount of DOA’s. However, because management involvement has been taken into account as an indirect cost in this research, and the DOA reflection day is specifically for management, these costs are under management.

The ‘clean bench’ from company 1 can been seen as an appraisal cost. Although this is still a pilot test, the purpose of the clean bench is to inspect the incoming parts. It can be interesting for this company to reflect on this in the near future, and decide on whether to continue with the clean bench.

45 Figure 5.6 is the updated DOA cost model. The direct cost (black) boxes are part of the research of Van Swieten (2013) and are left outside the scope of this research. The indirect costs can be split in indirect/overhead costs and image costs. Only an indication on image costs is presented in this research, further investigating image costs can be part of further research in the field of DOA’s. Given sensitivity of cost data, it will be more likely that image costs will be used for further analyses within the companies to reduce DOA and / or inventory costs. The ABC classification method described by Teunter et al. (2009) can provide an excellent tool for this. The colour of the image boxes is therefore made grey.

The indirect/overhead costs can be split into 5 activities based on the ABC cost tree model from Rezaie et al. (2008); logistics, diagnostics, analysis, preventive actions, and management involvement. These can be split into more specific cost objects as discussed in the first part of this chapter and presented in table 5.3. Not enough data was available to quantify the costs for improvements and customer contact by management, therefore these boxes are grey as well. The only relation between direct and indirect costs are the costs for logistics. The direct costs are emergency transportation of spare parts, once a DOA has occurred. The shipment of other spare parts, and the return shipment of DOA parts, to a warehouse or the external supplier, has been included as indirect costs.

46 5.5. Results on cost mapping

In the previous section the indirect/overhead costs have been analysed in great detail. However, the purpose of Activity Based Costing is to allocate indirect costs to a product to identify its true costs (Ben-Arieh & Qian, 2003). Therefore the indirect costs of the functions (activities) discussed before, are translated to a percentage of its average part value (table 5.1).

Figure 5.7: Relative indirect costs of DOA parts expressed as its average part value.

Figure 5.7 shows the indirect costs of the two companies expressed as a percentage of its average part value. The average value of parts is based on the findings from Van Swieten (2013), table 5.1. She discovered that the average DOA value of parts from company 1 is around 15.000 euro, and that the average DOA part value for company 2 is around 700 euro. Calculating the expenses for the five functions as a percentage of its average DOA part value, it can be seen that all costs except the logistics costs are similar. The high costs for logistics for company 2 can be explained by the fact that company 2 has a lot more parts, which get shipped around a lot more. The reason for this is company’s 2 strategy to have multiple spare parts available and that company 2 sends 90% of their DOA parts to their external suppliers for analysis, whereas company 1 does this analysis themselves. Apart from the high logistics costs, the cost for analysis remains to have the highest impact.

7% 28% 6% 10% 20% 30% 13% 5% 4% 5% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Company 1 Company 2

Indirect costs of DOA parts expressed as a percentage of its

average part value