A Practical Research to Improve The Process
of Material Return Application at Gracesemi
University of Groningen
Faculty of Economics and Business
MSc Business Administration
Specialization Operations & Supply Chains
Supervisors: Dr. Michiel Hillen Dr. Manda Broekhuis
Grace Semiconductor Manufacturing Corporation
Sales and Marketing Department
Sales Administration Unit
Supervisor: Edward Wang
Author: Chunchun Fang
Student Number: S1797158
Tel: +31(0)644801591
Preface
The thesis project is the final part of my master program of Business Administration- Operation and Supply Chain at University of Groningen. This research is conducted at Grace Semiconductor Manufacturing Corporation. During this project, many people have helped me with research design, data collection, report writing, etc. I would like to take the opportunity to show my appreciation to them. Without your help, this thesis could not be possible.
First I would like to thank my supervisor at Gracesemi Edward Wong, who gave me the opportunity to carry out research and continued to support me with data and information when I was writing thesis in the Netherlands. Furthermore my thanks go to staff at Gracesemi who shared with me their thoughts and ideas. What impress me most is no matter how busy they were, they tried to help me as soon as I came to them.
Also my gratitude goes to my supervisors of the faculty of Economics and Business of the University of Groningen. My first supervisor, Dr.M.W Hillen kept giving me advices before and during the research and shared with me his insights, understanding and knowledge of process re-engineering. His remarks and constructive criticism are the most helpful suggestions for thesis improvement. Besides I would like to say thanks to my second supervisor Dr. H. Broekhuis for her understanding, knowledge and feedback.
Finally I would like to thank friends and classmates in Groningen, with whom I had very good times. Also I would like to say “thank you” to my family, friends and colleagues in China and the USA for their support and advice.
Management Summary
The research project is conducted at the Sales Administration (SA) team, a sub-unit of Sales &MKT at Grace Semiconductor Manufacturing Corporation1. Gracesemi is a pure foundry that focuses on make-to-order (MTO) manufacturing of semiconductor wafers. Due to the strategic market expansion plan put forward in 2005, followed by several successful business projects, Gracesemi has been short-capacitated ever since. (Gracesemi faces order shortage now due to global economic crisis). Short lead-time promise to customers to some extent leads to unacceptable quality of products— sometimes some steps of inspection are skipped in order to meet deadline--which will be returned to Gracesemi via the Return Material Application (RMA) process coordinated by SA team.
Since the system and logistic process are still local standard, responsiveness and service quality of RMA case handling are far from what international customers expect from Gracesemi. The team actually missed target of RMA handling several times within one quarter in 2006, 2007,2008 and first half year of 2009. The region the most severely suffering delay and poor service quality is Japan, the reason of which is Japanese customers keep higher quality standard and set much stricter reject criteria. A research question is developed as what are causes of current bad RMA responsiveness and what are solutions for improvement.
A conceptual model was built trying to explain presumed relationships between different factors, which finally cause poor RMA handling. The model, built under the theory of Mirror Effect, Cost and Quality model by Slack is verified by observations, and semi-structured interviews.
The research conducted at Gracesemi focuses on the following scope: 1) Process flow
2) The level of staff satisfaction 3) Knowledge retention
Data analysis mainly focuses on history data regarding responsive time of each RMA phase, interview reports.
The improvement approaches suggested are the follows:
Firstly, optimize the process by paralleling and re-sequencing activities with purpose of eliminating any slack within whole system. This project relies heavily on accurate ERP data as well as IT support.
reasonable workload, (SMIC2 model); the second is to increase their salaries to compensate the effort devoted to their jobs (HHNEC3model). Besides, take advantage of HR resource to organize relevant training and take good care of training materials. It is suggested that only one senior sales administrator be appointed the training specialist for answering all questions to new hires, so that others could be free of disturb.
Last yet importantly, pay more attention on maintaining information achieved and take good care of skilled employees in order to take good advantage of the skills and knowledge they have.
The investigation result implies that corporate culture is very important. Compared with international companies with a history of hundred years, Gracesemi is rather at its childhood forming its own culture and working style. It is my own experience during investigation that there are culture conflicts between different departments and management styles, since Gracesemi recruited many professionals abroad from Intel, TSMC, Charted Semi and Infineon, etc. How to capitalize on the diversity and establish unique corporate culture that benefits Gracesemi remains a question for future research.
2 SMIC: Semiconductor Manufacturing International Corporation 3
Table of Contents
Chapter 1 Introduction………...6
1.1 Grace Semiconductor Manufacturing Corporation………..6
1.2 Sales & MKT department………6
1.3 Sales Administration Team………..7
1.4 Management problem………..8
1.5 What is RMA………...8
1.6 Report Structure………...9
Chapter 2 Problem Identification……….10
2.1 Problem Exploration………..10 2.2 Internal Measurement………10 2.2.1 Performance Objective………..11 2.2.2 Performance Status………...13 2.3 External Measurement………...15 2.4 Problem Statement……….16
2.4.1 Objective and Research Question………..………..16
2.4.2 Research Boundaries………16
Chapter 3 Research Design………...17
3.1 Introduction………17
3.2 Theoretical Framework………..17
3.2.1 Mirror Effect………...17
3.2.2 Findings on Relation between Employee Benefits and Work Motivation...19
3.2.3 Cost and Quality Curve………...19
3.2.4 Lean concept………...20
3.2.5 Value Chain Analysis……….21
3.3 Conceptual Model………..21
3.3.1 Variables and Relations………..22
3.3.2 Variable Operationalization and Data Collecting Ways………24
3.3.3 Sub-questions………..24
3.4 Research Strategy………...25
3.5 Data Gathering Methods………25
3.5.1 Archiving Data………26
3.5.2 Semi-structured Interview………...26
Chapter 4 Results and Diagnosis...27
4.1 Introduction……….………..27
4.2 Process………..27
4.2.1 RMA Verification process………..27
1) Process Flow………..27
2) Lead-time analysis………28
4.2.2 Coordination with Imp&Exp………..35
1) Activity Introduction……….35
2) Problem Analysis………..36
4.2.3 Manufacturing and Credit Note………..37
1) Activity Introduction……….37
4.2.4 Bottleneck Identification………...………….37
4.2.5 Summary………...………….38
4.3 Analysis on Employee Satisfaction………...……….38
4.3.1 Staff Compensation……….38
4.3.2 Internal Support………..40
1) Software Support……….40
2) Cooperation Support………40
4.4 Analysis on Knowledge Retention………...…...41
4.4.1 Document Maintenance………..41
4.4.2 Staff Turnover……….41
4.5 Conclusion………...……...42
Chapter 5 Remedy Approach……….……...44
5.1 Introduction………44
5.2 Process Re-engineering………..44
5.2.1 Future version of RMA process……….44
5.2.2 Action Plan……….47
5.3 Employee Satisfaction………...48
5.4 Knowledge Retention………48
5.5 Summary………...49
Chapter 6 Conclusion and Future Study………..50
6.1 Introduction………50
6.2 Conclusion……….50
6.3 Limitation………...50
6.4 Reflection on Future Study………51
Chapter 1 Introduction
This research has been conducted at Grace Semiconductor Manufacturing Corporation, a Shanghai based foundry. In order to give an overview of the organization, a company introduction will be provided, followed by introduction to Sales & MKT department and Sales Administration unit. Thereafter the management problems as well as the report structure will be discussed.
1.1 Grace Semiconductor Manufacturing Corporation
Grace Semiconductor Manufacturing Corporation (Gracesemi), a Shanghai-based wafer fabrication company started operation in Sep., 2003. Since its establishment, Gracesemi has been sticking to the objective of becoming a leading foundry in China by supplying high quality and advanced value-added technology solutions to their domestic and international customers.4 Unlike other integrated design manufacturers (IDMs) in semiconductor industry such as Intel and Toshiba, which incorporate design, manufacturing, testing & assembling, Gracesemi is a pure foundry that focuses on manufacturing. Both fabless design houses and IDMs are customers to Gracesemi. A majority of 1370 employees are operators who work in the fab on a 3-shift base. Due to high value of its product (wafer5), operation models are make-to-order (MTO) and engineering-to-order (ETO); no back-up wafers are available in stock for any form of compensation. Company organization is illustrated in Figure 1:
Figure1 : Organization Chart of Company
Each department has its own function. RMA process is a project involving all departments.
1.2 Sales & MKT department
Sales & MKT department is a vital part for the manufacturing company. There are clear role distinctions between different units within this department. A general organization picture is provided in Figure 2.
Overseas sales offices are mainly responsible for developing potential customers; once a customer starts to run business with Gracesemi, it is Account Manager (AM)’s job to take good care of the customer from all perspective. Customer Engineers (CEs) are responsible for technical support. The Sales Administration unit is involved in logistic coordination for order-related business. Central planning balances capacity allocation
4
Quoted from website page: http://www.gracesemi.com/enhtml/company.jsp?page_type=10&bid=10&id=9
among customers and tries to come up with solution when there is capacity constraint. Customer & Product Service is in charge of turnkey service of back-end when customers have circuit probe (CP) testing request, which Gracesemi could hardly run the test.
Figure 2: Organization Chart of Sales & MKT
1.3 Sales Administration Unit
This unit comprises of 10 persons in total. For each sales region, sales administrators are assigned based on sales volume. Table 1 shows the assignment status.
Region Sales Admin. Team Manager
North Asia (JP/North Korea) 2 people
South Asia (Taiwan/China Mainland) 3 people
Europe 1 person
US 3 people
1 person
Table 1: Sales Administrator Allocation Status
The Sales Administration unit is responsible for order-related activities, and RMA
handling (mentioned in following sections) in general. Daily routines are indicated below: 1.Report monthly delivery status
2.Execute order administration process, from order acceptance, manufacturing till shipping. Coordinate between customer and manufacturing team for requests and emergencies.
3.Monitor monthly supply and demand planning, translating customer forecast into capacity booking.
1.4 Management Problem
This research is designed to solve management problems. By talking with unit manager, the problems of RMA-related coordination are discovered. Since the operation models at Gracesemi are MTO or ETO, no back-up wafers are available at inventory. If the wafers are detected unqualified by customer, Gracesemi can’t deliver compensation wafers immediately from inventory; instead Gracesemi has to trigger the process of Return Material Application (RMA) and solve the quality issue as soon as possible. The SA unit is responsible for RMA coordination within the company. The manager of SA unit mentioned RMA actually negatively impact the annual KPI result of Sales & MKT. Besides, customers sent emails to complain RMA handling, which is also confirmed during the interview with a senior sales administrator.
1.5 What is RMA?
Return Material Application (RMA) occurs when Quality departments of customers pinpoint wafers that failed to pass certain criteria after circuit probe (CP) testing in their fabs or labs, and claim either economic make-up by Credit Note or compensation wafers. According to trade terms, Gracesemi has to cover all expenses occurring during the process. Since Gracesemi only starts compensation action after unqualified wafers have been shipped back to the warehouse, the time period for closing an RMA case is very long. Figure 3 illustrates the period.
Generally speaking, RMA handling incorporates phases of verification, shipping, Import & Export, and an average of 2-month manufacturing. The involvement of internal stakeholders such as Quality department, Planning, and process engineering for verifying the rationality of return application as well as logistic coordination by sales administrator normally requires time and energy.
Figure 3: Timeline of RMA process for a specific case
Phase-1 Verification
Verification part aims to decide whether Gracesemi should be responsible for RMA. The process goes through ERP system and has Sales & MKT, Quality and TD involved.
Phase-2 Imp & Exp (batch return)
This process is triggered only when the Quality department announces Gracesemi is responsible for the quality issue. Activities include forwarder selection, coordinating import and export as well as tax return. (In order to encourage export of hi-tech products, Chinese government refunds VAT of exported products. If these products are returned, Gracesemi has to pay back the government). This phase is only for returned wafer of large quantity. Sample return is not included in this phase.
Phase-3 and 5 Shipping
Once a forwarder is selected and tax return is settled, sales administrator has to keep an eye on shipping and at the same time feedback to internal engineers with regards of delivery status.
Similarly when compensation wafers are shipped, sales administrator has to inform customers about delivery status.
Phase-4 Manufacturing/Credit Note
If customers accept compensation by wafers, then SAs release manufacturing instruction. The Planning department monitors manufacturing status and makes sure cycle time of 1.8 days per layer is strictly kept.
If the deal needs to be closed by credit note, then sales administrators have to submit application for issuing credit note.
Since Phase 3,4 and 5 are fixed with time frames, our research will mainly focus on phase 1 and pay some extra attention to Phase 2.
1.6 Research Structure
A preliminary research will be conducted to explore problems. With short conversation with sales administrators, root causes are assumed and indicated in Chapter 2. A comprehensive introduction to research design is presented in Chapter 3. By doing inspection and interview, the causes are further proven in Chapter 4. Remedy approaches are discussed in Chapter 5.
Preconditions
Chapter 2 Problem Identification
2.1 Problem Exploration
In order to further explore problems, conversations with all sales administrators and their manager are conducted. According to their description, I summarized their opinions about RMA, which is indicated in Table 2.
Internal stakeholder Problem perceived relating to RMA
Specific problem Common problem
1.feel frustrated when RMA request deserve no priority in other department
Junior sales
administrator 2.time conflict between learning RMA handling and busy routine. Unskillful in RMA coordination extend lead-time
1.interrupted to answer questions for juniors regarding RMA problem repeatedly. Can not concentrate which sometimes leads to errors 2.Lot of system errors requires rework which decrease service quality and increase lead-time 3. Tremendous key-in work due to the system is not intelligent enough, which leads to high level of errors and low morale
Senior sales
administrator
4. Long response time from other internal department leads to long lead-time
1.Not easy to hire qualified people within short time
Manager of SA unit 2.high staff turn-over leads to problem to knowledge retention and increase the cost on training
Customers complain, sometimes they threat
to upgrade to
executive level and ask for penalty
Table 2: problems perceived by staff in Sales Administration team
It provides direct information regarding RMA from an operational level. The problems perceived by internal employees are verified later in an official interview with ex-employees, senior employees and their manager. There are not only qualitative perception, but also quantitative judgment which will be mentioned in sections 4.2.
2.2 Internal Measurement
Inspection of archive in the document library reveals that Gracesemi conducts quality control within Sales and MKT. They defined standard operation procedure (SOP), quarterly business review (QBR) and annual key performance indicator (KPI) review. RMA performance is an important aspect of the annual review.
Standard operation procedure (SOP) is a set of instructions having the force of a directive, covering those features of operations that lend themselves to a definite or standardized procedure. Currently each activity has its own SOP to ensure its quality.
QBR
Quarterly business review (QBR) is held immediately after each quarter is finished. It aims to review business performance of each department. For Sales & MKT, the RMA review is one of the most important aspects.
KPI
Key performance indicators (KPIs) are measures used to help an organization define and evaluate how successful it is, typically in terms of making progress towards its long-term organizational goals. RMA handling is one of the KPIs that measure performance of Sales &MKT department at the end of a year.
The lead-time target of RMA handling is used to measure how well current performance is. The result of eliminating the gap between the lead-time target (see Figure 13) and lead-time performance (see Figure18) is the ultimate objective of this research project. By going further with this research, an introduction to the gap is provided.
2.2.1 Performance Objective
The target of RMA lead-time is set at management level based on the benchmark of industry (TSMC6). A general process flow of an RMA case is shown as in Figure 4 prior to target presentation for better understanding of the situation.
6
Figure 4: RMA process flow
Table 3 is presented in order to link activities of the RMA process flow with the 5 phases indicated in Figure 3. Please note that not all phases have their counterparts in Figure 4, such as Phase 5, and not all activities mentioned in process flow belong to RMA phases, for example, Activity 10: return sample to customers is not included in any phases.
Finance unit conducts Activity 11, which happens immediately after verification and by default the judgment is “yes”. Therefore we include Activity 11 in Phase 2.
Phase Description Activities (described with number in Figure 4)
1 Verification 1,2,3,4,6,7,8,9 (Activity 4 involves Imp & Exp,
but it is still a part of verification) 2 Import & Export (batch return) 11,12
3 Shipping 5,13
4 MFG/Credit Note 14,15
5 Shipping Not indicated in process flow
Table 3: Link between RMA phases and Activity within process flow
Phase-1, 2 and 4 have a set lead-time target, while for Phase 3 and 5, according to SA manager the lead-time is determined by external forwarders. The targets of key activities of each phase are shown in Table 4.
Item Activity Phase Target
1 1 1 Max. 1 working day
2 From 2 to 8 (activity 4 involves Imp & Exp) 1 Max. 10 working days
3 12 2 Max. around 20 days, acceptable within 45 days
4 14 and 15 (credit note) 4 Max. 2 weeks, quite flexible
5 14 and 15 (compensation wafer) 4 Cycle time<=1.8day per layer
Table 4: Target of some activity
Item 1--sales administrators conduct it individually. They receive claim from customers and upload information to ERP system. There is no record of the time between claim receipt and uploading, as the two modules are not integrated yet. According to the interview with senior sales administrator, the job is well done.
Item 2--it is clearly recorded in ERP system since it is an online verification with email notice to action owner. The email contains a short instruction as well as a link to database. The owner only needs to click the link to start a task and later send to the next owner by clicking the “confirm” button.
Item 3--it is coordination via email and telephone according to sales administrator. It has no processing time recorded in ERP system.
Item 4-- it is a little bit uncontrollable on lead-time since it involves Chief Financial Officer (CFO). Since CFO is busy with business traveling and meetings, it happened sometimes CFO delayed applications.
2.2.2 Performance Status
For each item mentioned above we have specific ways of data collection, which are listed as follows:
Item 1—Conversations with several sales administrators provide information for analyzing. According to them, customers are required to send claim to a corresponding sales administrator while copy another who is a backup. Thus the person in copy list would remind the person in charge if customer’s claim were not handled within 24 hours. So far case of more than 24 hours responsive time has never occurred.
Item 2—Current verification status is recorded in ERP system. According to data, average time spent on closing verification of a product during the 1st quarter of 2009 is 19.1 days, 9.1 days more than the target.
Item 3—The only way to find out the lead-time is by investing back and forth emails which indicating the start and the finish of activity 12, see Figure 4. Interviewing senior sales administrator about coordination status shows that the lead-time performance fails to meet the target quite often, which is proven by investigation on RMA status of a Japanese customer from Jan 1st, 2009 to Apr. 27, 2009. The average lead-time is 20.13 days, which is 0.13 days more than target. Detailed lead-time of 14 cases is shown in the figure below where process time of RMA case fluctuates along the target.
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 1111 2222 3333 4444 5555 6666 7777 8888 9999 10101010 1111 121111 121212 13131313 14141414 Time Target
Figure 5: Lead-time of Item 3(Phase 2)
Item 4-- This step could go through paper application only, because CFO’s signature is officially mandatory to this type of business deal. Record of process time could be tracked from the secretary who takes care of all paper document sending and returning history between Sales & MKT and other departments.
According to her description, 3 out of 5 applications from Jan to Mar 2009 are retrieved within 2 weeks while the other 2 requires more than 2 weeks. There are even some applications held at CFO’s assistant for around one month during Dec. 2008. The reason is “CFO went back to United States”. Another explanation from the Finance about extending lead-time is that they concerns a lot about activities that may impact cash flow and more time is needed to justify credit note application. Thus the lead-time of this item
is difficult to predict. Actually Gracesemi accepts this kind of unstable lead-time given the reasons mentioned above.
Item 5—According to data from the ERP system, the performance of current wafer manufacturing hits the target of normal lot 1.8d/layer, hot lot 1.2d/layer and super hot lot 0.8d/layer.
Figure 6: Wafer manufacturing status
The gap between objectives and performance
The biggest gap appears at phase 4, which is credit note application. However since company does not recommend the approach of issuing credit note, most of compensation is realized by starting wafer production. Therefore it is important to identify the bottleneck during the whole process. Bottleneck contains the most wastes that badly impact the lead-time. And the bottleneck identification will be introduced in Chapter 4.
2.3 External Measurement
As we mentioned above, Japanese customers have the strictest quality requirement among all customers. They adopt a quantitative method to score RMA performance. Take Company S for example, they review RMA responsiveness as an important aspect of total responsiveness. According to their scoring system, responsiveness is measured by efficiency (email feedback) and effectiveness (case closed with satisfaction) to all business including technical communication, order administration, RMA, etc. Though total score is not very bad, it is worthwhile identifying causes for point losing. The figure below is quoted from Company S-Grace QBR (Sep.26 2007) documents for tentative measurement of total responsiveness based on email feedback from Gracesemi. According to the definition of Company S, the 1st responsiveness represent email feedback of RMA initial report (IR) and the Finish represents email notifying the decision on compensation either by wafers or credit note.
Company S strictly watches the performance indicator of responsiveness. Under great pressure of scoring and rating from customers, Gracesemi has to consider ways to eliminate negative factors.
0.80 1.20 1.80 0.00 0.50 1.00 1.50 2.00 Super Hot Hot Lot Normal (p ri o ri ty )
Figure 7: Responsiveness Evaluation of Company S
2.4 Problem Statement
2.4.1 Objective and Research Question
Based on the summary of preliminary data gathering, the objective of this research is to provide insight into effectiveness improvement on RMA handling within the Sales Administration team, in order to improve customer satisfaction by shortening total responsive time on RMA issues, and at the same time maintain high service quality. Response lead-time and service quality are the two most important indicators of effectiveness.
In order to fulfill the research objective, it is of great importance to identify factors that lead to poor performance. It is necessary to identify these negative factors so as to make right diagnoses. Therefore this objective is interpreted into a main research question: What are the causes of current RMA poor responsiveness and what are improvement solutions? In section 3.3.3 there are four sub-questions helping further develop the main question.
2.4.2 Research boundaries
The research is conducted at SA unit and confined to RMA business activities. Research scope is process improvement, employee satisfaction and knowledge retention. The reason for choosing these 3 aspects is explained in 3.2.
Research limitation is that all data are from a specific company within several given time frames, so remedy suggestions applies only to Gracesemi under certain circumstances.
Chapter 3 Research Design
3.1 Introduction
This research is designed based on problem exploration within guidance of theory in management operation, and human resource management. The basic idea of this research is to verify the presumed causes negatively impacting RMA handling. Presumed causal relationships are summarized based on the Mirror Effect and Slack’s quality and cost model. In addition, on-site visit to the company is another important source to build cause and effect relation. The verification is conducted through interview and inspection. In this chapter, a theoretical framework is introduced followed by a conceptual model, which is the core of whole verification process.
3.2 Theoretical Framework
As mentioned in Chapter 2, the level of customer satisfaction is negatively impacted by RMA responsiveness. The lead-time and service quality are two indicators of RMA responsiveness. The research will focus on how to avoid missing the lead-time target and at the same time maintain good service quality. According to problems summarized in Table 2, three aspects deserve further investigation. They are process improvement (see problem 4 perceived by senior SA), employee satisfaction (see problem 1 perceived by junior SA) and knowledge retention (see problem 2 perceived by the manager of SA). Relevant literatures are reviewed and presented in this section. Mirror effect is adopted to explain the relation between customer satisfaction and employee satisfaction. It is believed that at customer encounter point, if the employees are treated well, the customer could be satisfied as well. It is important to include compensation policy into research since according to the findings motivation close relate to compensation. Part of knowledge retention is measured by staff turnover rate, which has a direct relation with cost. A model of Quality and cost curve appears in this section, trying to illustrate the present situation at Gracesemi. Lean thinking plays an import role at eliminating wastes during the whole process. Thus the literature of lean thinking drives process improvement towards a right direction. An introduction to value chain analysis (VCA) helps to get an idea about the tool which is used in process analysis part.
3.2.1 Mirror Effect
Figure 8: The Satisfaction Mirror
(Source: James L. Heskett, W. Earl Sasser. Jr. and I. eonard A. Schlesinger, The Service Profit Chain, The Free Press. New York. 1997, p. 101)
According to Fitzsimmons J.A in Service Management 6th edition (2007), there are six steps that indicate the process of how internal quality finally impacts profitability and growth through mirror effect.
Internal quality drives employee satisfaction. “Internal quality describes the environment in which employees work and includes employee selection and development, rewards and recognition, access to information to serve the customer, workplace technology, and job design.”
Employee satisfaction drives retention and productivity. It is easy for a company to maintain a high level of employee loyalty and productivity if the level of employee satisfaction is also high.
Employee retention and productivity drives service value. The high customer perception of value is realized high quality service provided by skillful and professional employee.
Service value drives customer satisfaction. “Customer value is measured by comparing results received to the total costs incurred in obtaining the service ” Customers are satisfied when they receive excellent service by minimal costs. Customer satisfaction drives customer loyalty. Customers are expected to repeat
purchase when they are well satisfied at the first time.
Customer loyalty drives profitability and growth. Customer. Stable customer base make it possible that business runs at a low cost since company are familiar with customer demands. In the meantime, company could distribute more resource on innovation and expansion.
Better Result Low Cost High
Customer satisfaction
More Familiarity with Customer Needs and Ways of Meeting Them
Greater Opportunity for Recovery from Errors
Higher Employee Satisfaction
Higher Productivity
Improved Quality of Service Strong Tendency to
Complain about Service Errors
The satisfaction mirror sheds some lights on the relationship between employee satisfaction and customer satisfaction. It is of great importance for a company to know what its staff want and tries to improve employee satisfaction level. Otherwise the company will suffer ultimately.
It is assumed that customer satisfaction is positively impacted by employee satisfaction at Gracesemi. The theory mentioned above helps to verify this assumption. Besides according to 6-step process, internal quality has important impact on employee satisfaction and employee satisfaction actually impact knowledge retention and service quality. Compensation is an important factor that positively impact employee satisfaction which is supported by the findings in 3.2.2.
3.2.2 Findings on Relation between Employee Benefits and Work Motivation
Hong et al (1995) conducted research on the impact of employee benefits on employees’ work-motivation. According to their findings, employee benefit program, especially monetary benefit program do have great impact on work-motivation. Further investigation shows different genders have different benefit demands; female employees focus on leaving-taking system such as maternity leave while male employees tend to emphasize training and development. Besides, “Unmarried employees, more than married employees, perceive that employee benefits have a greater impact on job performance.”(Hong et al 1995) Employees with low motivation are inclined to quit. This research indicates the relation between employee benefit program and work motivation, which well explains SA’s complaint on salary in the following chapter. Accordingly remedy approaches in Chapter 6 will take the findings into consideration.
3.2.3 Cost and Quality Model
Cost and quality relationship is developed based on a premise that certain level of quality failure is acceptable. With inspiration of TQM, relation between cost and quality is adjusted as the figure below. Cost of error will get closer to Cost of quality along Quality effort. This model shows that total cost decreases when more quality effort is input. More quality effort actually minimizes total cost. The model which suites current situation at Gracesemi is Figure 9.1 as there are certain defect level allowed at Gracesemi and they are trying to find an optimal staff turnover rate actually minimizes total cost. It is suggested in the future to decrease total cost, as more quality effort is input, which is illustrated in Figure 9.2.
Figure 9.2: Quality and cost curve
The model of Quality and cost curve shows the relation between total cost and quality effort. Quality effort is required in order to decrease total cost. With the guidance of this model, remedy approaches propose to input more quality effort so that total cost is expected to decrease.
3.2.4 Lean Concept
Lean concept, first introduced by Toyota Motor Company in the early 1970s, is a set of thinking that aims at optimizing operation by squeezing out wastes every step of the way. Waste here is defined as “any activity that does not add value”(Slack, Operation Management). According to Toyota, there are seven types of waste, which applied to both service and production.
Over-production. It is common practice in push production that producing more than is immediately needed by the next process. It is necessary when capacity utilization is concerned.
Waiting time. Task has to be waiting for processing due to capacity constraint or large batch consideration.
Process. The process itself could be waste since some process is added because of poor product design. The process actually has little value to the product.
Inventory. The inventory is aimed to be eliminated since it increases cost and add no value to the product.
Motion. “An operator may look busy but sometimes no value is being added by the work. Simplification of work is a rich source of reduction in the waste of motion”
Defective goods. Defective goods have little value to customer and should be eliminated.
The investigation followed focuses on how to eliminate wastes. In addition, another 2 key issues that also define lean philosophy are discussed as follows:
Involvement of staff. An organization’s culture is important in supporting these objectives through an emphasis on involving staff. By cultivating leaning thinking, employees are encouraged to enrich their jobs with extra-related activities and to be multi-tasking. “The intention is to encourage a high degree of personal responsibility, engagement and ownership of the job”. In RMA verification cases, strong sense of ownership is very helpful to minimize wastes.
Continuous improvement. It to some extent resembles to six sigma, and has its unique characteristics. Lean objectives of zero waste and perfect quality are often expressed as ideals. It emphasizes the practice on the way in which a company moves closer to its ideal targets.
3.2.5 Value Chain Analysis
Michael Porter (1985) in his book “Competitive Advantage: Creating and Sustaining Superior Performance” put forward the concept of value chain analysis. Based on his concept, a value chain is a chain of activities that add value to products or final outcomes. The value chain has a close linkage with cost.
The model of value chain analysis (VCA) figures out all activities in a business process and then identify the activities that creating real business value or customer value. By filtering those non-value added activities, a cost-effective chain is formed.
The VCA is applied extensively in the analysis of process part. Valued-added activities remain in process of new version.
3.3 Conceptual Model
first.
Figure 10: Conceptual Model
The above-mentioned relationships are explained below. RMA handling responsiveness
Responsiveness of RMA means that performance of RMA handling meets the target of cycle time and quality. Cycle time target is set at management level and quality target is defined as no further complaint on service attitude and email response. (JP survey shows email respond is good)
Process execution
Process execution is one of the two most important aspects impacting RMA handling performance. This concept deals with goods flow process. To be specific, process execution incorporates circulation of hardcopies of internal applications, sample delivery between Gracesemi and customers, batch return of unqualified wafers and restarting of compensation wafers. An RMA case is finally closed when compensation wafers are shipped.
It is assumed that process execution impact RMA performance in a positive way. In other words, if process is executed efficiently and effectively, RMA performance is excellent accordingly.
Information delivery
Information delivery is another important aspect impacting RMA handling performance. It mainly refers to information flow that facilitates the smooth running of goods flow. Information is required to deliver precisely and in time. There are 3 different information types, namely technology information, business information and logistic information.
Compensation
+
The relation between information delivery and RMA performance is positive too. In other words, good information delivery makes good RMA performance possible.
Communication with (external) stakeholders
This variable refers to communication between SA team and customers. Interactive activities are conducted via emails and telephones. Therefore responsiveness of email is an important indicator of customer service. Besides, quick and accurate RMA data uploading is another indicator.
The information exchange between SAs and customers has direct effect on information flow. Accurate information at customer site guarantees a good start of information flow at Gracesemi.
Training
Variable training refers to internal trainings organized by HR or SA unit. Trainings aim to improve professional skills of SA and enrich their knowledge about RMA.
This variable tries to explore the causal relation between training and information delivery performance. The relation is assumed positive, that is good training to sales administrators help to facilitate information flow.
Knowledge retention
Knowledge retention is a concept that incorporates archives and skills possessed by employees. Archives include hardcopies of RMA claim, Credit Note application, and training material. Skills possessed by employees are another form of knowledge retention. It could only pass down from a senior SA to a junior SA. It is assumed that good knowledge retention has positive effect on information delivery. If data could be easily tracked from archives and at the same time, employees handle RMA skillfully, information flow runs smoothly.
Staff turnover
Staff turnover describe the status that SA unit gains and loses employees. High turnover can be harmful to RMA handling since the employees leave are skilled and they usually have high productivity.
It is assumed that staff turnover has both direct and indirect impact on information delivery. The direct impact refers to the situation where some information permanently lost during handover. It is negative relation. The indirect impact on information delivery is through knowledge retention and training. It is assumed that high staff turnover decreases knowledge retention level and increase training frequency and quality.
ERP system support
ERP system provides technical platform for conducting all business activities. Good ERP system support to convey information efficiently and accurately and to track documents immediately. The causal relation between information delivery and ERP system support is positive. In other words, ERP system positively impacts information delivery.
Employee satisfaction
fewer days off, and stay loyal to the company.” It impacts process execution in a positive way, since if employees are happy with their job and stay at their post for a long time, working efficiency could be improved due to familiarity. However the relation between employee satisfaction and staff turnover is interpreted as high employee satisfaction results in low staff turnover.
Cooperation with other departments
Department cooperation is very important to a MTO company. The cooperation guarantees smooth running of each case.
It is assumed that seamless cooperation between different departments facilitates process execution and increase employee satisfaction level.
3.3.2 Variable Operationalization and Date Collecting Ways
Welker et al (2008) explain that operationalizing concepts means you have to make variables measurable. In other words, variables need to be measured and translated into specific, measurable, observable and recordable aspects. Generally speaking, variables fall into two categories, quantitatively measurable and qualitatively measurable. With quantitative measurement, data are collected from ERP or some archived records and used in analysis. With qualitative measurement, information is collected from interviews that are more like an opinion or a judgment.
Process execution
As mentioned above, process execution is about goods flow. Four types of goods flow are measured in the following ways:
Circulation of credit note application
Data from secretary’s desk (see Table 9)
Quantitative
Sample delivery Interview Qualitative
Batch return Interview Qualitative
Compensation wafer MFG Data from MFG (see Figure 6) Quantitative
Table 5: Data collecting ways of process execution
Information delivery
There are there types of information delivered company-wide. They are measured in the following ways:
Technology information Via emails, not coordinated by SA No discussion
Business information Conveyed by AM No discussion
Logistic information Records in ERP system and Emails Quantitative
Table 6: Data collecting ways of information delivery
Communication with stakeholders
There are two important aspects in communication, email responsiveness and upload data to the ERP system. Table shows the details.
Email responsiveness Record in ERP, interview and customer materials
Quantitative + Qualitative
Data upload Interview Qualitative
Knowledge retention
Two forms of knowledge retention are discussed in the thesis. One is the skill possessed by experienced employees and the other is archived document. The table below shows the details:
Skill Interview Qualitative
Archives Inspection Qualitative
Table 8: Data collecting ways of knowledge retention
Training
The information of training status is collected from the interview with sales administrators and their manager. The analysis is conducted based on their judgment.
Staff turnover
Staff turnover is quantitatively measured by staff turnover rate. Staff turnover rate is applied to measure hiring and leaving status.
Staff turnover rate is defined as Total number of leavers over period x 100% Average total number employed over period
ERP system support
The variable is measured in a qualitative way. The interview is conducted with senior SA in order to get their remarks about current ERP systems.
Employee satisfaction
This variable is measured in a qualitative way. Gracesemi has conducted a survey regarding employee satisfaction. The raw data are kept confidential at HR. The interview with the manager shows that SA are not satisfied with compensation, workload and poor support from other departments.
Cooperation with other departments
The variable is measured in a qualitative way. Data are collected from interviews with ex-employees and senior SAs.
Compensation
The variable is measured in a quantitative way. Data are collected from the interview with SA and then compared with average figures in Shanghai and at competitors.
3.3.3 Sub-questions
As mentioned in Chapter 2, the main question is what are the causes of current RMA poor responsiveness and what are improvement solutions. Based on assumed relations indicated in conceptual model, sub-questions are developed as follows:
1.How does staff turnover impact RMA handling performance directly and indirectly (by training and knowledge retention)?
2.How can staff satisfaction be improved with regards to internal corporation, and IT support, in order to improve customer satisfaction?
Among these sub-questions, Question 1 will be explored extensively since according to management problem, the current process does unnecessarily waste much time and an improved version of process could take effect as soon as it is launched. The other questions though are also important, will not be investigated as detailed as question 1.
3.4 Research Strategy
“According to Van der Velde et al. (2004), the classification of research strategies is based on two dimensions, namely the degree to which the researcher intervenes and the degree to which the researcher wants to make general valid conclusions. ” (Welker et al, Research Process, 2008) Figure 11 illustrates the classification of research strategies. The research conducted at Gracesemi falls into the category of case study. Accordingly, data gathering methods are devised based on this research strategy. The purpose of Figure 11 is to relate different research strategy with their data collecting methods. To be specific, the most effective ways for case study are archives and interviews, which are further introduced in the following section.
Figure 11: Classification of research strategies based on two dimensions - adapted
from Runkel and McGrath, 1972 (Van der Velde et al., 2004, p.76)
Chapter 4 Results and Diagnoses
4.1 Introduction
This chapter focuses on analysis of available data and information with regards to process improvement, employee satisfaction and knowledge retention. Data for process analysis are collected from ERP systems. And data for employee satisfaction analysis are mainly from interviews with ex-employees and senior SA. Data for knowledge retention analysis are from the interview with manager and document inspection report. At the end of this chapter, all diagnosed problems are presented.
4.2 Process
Problems are diagnosed with the model of Michael Porter’s Value Chain Analysis. This analysis firstly follows the sequence of a real RMA case (see Figure 3) and then goes deeper into each phase. RMA verification process is the first to be examined (See Figure 4, Activity 1,2,3,6,7,8), followed by Phase 2(Imp & Exp) and Phase 4(MFG/Credit Note). Verification analysis shows there are sequence problems between CE and AM. Besides lead-time target of each unit is not properly set. Phase 2 and Phase 4 both have the problem of unstable and poor lead-time performance. Since Phase 3 and 5 are both about shipping, over which Gracesemi has little lead-time control, we will not analyze them in the chapter. Based on analysis results, the phase containing the most wastes is identified as bottleneck in 4.2.4.
4.2.1 RMA Verification Process
Verification process is of great importance to Gracesemi, since it makes sure that only the valid RMA claim could get compensated, which to some extents minimizes financial loss to the company. The verification loop involves Sales & MKT, Quality, TD, and Manufacturing.
1) Process Flow
The verification process starts with online claim and ends with final report. Activity owners are from several departments and units. Figure 12 shows the detailed workflow.
Figure 12: Verification Workflow
Information Flow
Goods Flow
SA: Sales Administrator CE: Customer Engineer AM: Account Manager
QE: Quality Engineer TDE: Technology Development Engineer
MFG/PIE: Manufacturing Department/ Process Integration Engineer
Activity 1: claim goes to CE for double check from technology perspective
Activity 2: claim goes to AM for checking from business perspective
Activity 3: claim goes to QE; IR generated
Activity 4: IR goes to SA
Activity 5: claim goes to TDE if pilot wafers are concerned; otherwise it goes to MFG/PIE
Activity 6: feedback from TDE or PIE
Activity 7: FR goes to SA to close the verification
2) Lead-time Analysis
As mentioned above, lead-time target of verification is 10 working days, which should be less than 14days (two weekends are allowed to add) as recorded in the ERP system. The breakdown of target is indicated in the pie chart below:
Tar get Br eakdown
1 1 2 4 2 CE AM QE PI E/ TD QE r epor t
ERP data regarding Customer S reveal a serious responsive problem. RMAs uploaded into the ERP system between 1st Jan 2009 and 1st May 2009 shows only Quality department hits target while others fail to meet target. (See Figure 14) Process Integration Engineer (PIE) and Technology Development Engineer (TDE) process with the longest verification time while CE performs poorly with operation time 158% more than target.
2.58 1.82 1.47 6.13 1.84 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 CE AM QE PIE/TD QE report Ave.Time Target STD
Figure 14: Performance overview
Further investigation within CE, AM and PIE/TDE reveals that there are some delays that cannot be avoided. Table 3 shows investigation results:
RMA No CE wd AM wd QE wd PIE/TD wd QE report wd Total Cycle
time Detail Information
1 0.79 6.05 1.05 2.15 1.77 15.8 days Due to handover between AMs
2 0.75 5.91 2.29 11.87 3.99 34.8 days
3 0.07 4.71 1.04 6.16 1.92 17.9 days Due to handover between PIEs
4 21.01 0.74 2.99 6.17 0.88 45.8 days Due to CE technical discussion
5 3.97 0.84 2.07 5.83 1.00 20.7 days Due to CE technical discussion
6 0.09 0.72 0.29 0.98 1.92 6 days 7 0.69 1.21 0.83 4.97 1.10 12.8 days 8 0.81 0.21 0.72 5.23 2.93 13.9 days 9 4.91 1.07 2.68 5.20 2.04 23.9 days 10 0.94 0.81 0.95 4.21 1.89 12.8 days 11 0.90 0.89 1.16 2.84 2.40 12.2 days 12 0.13 0.67 1.29 2.96 1.70 11.1 days 13 0.94 0.91 1.13 3.83 1.20 12 days
14 0.12 0.71 2.14 15.90 1.03 27.9 days Due to May day holiday
Table 9: Detail performance review and delay reasons
Abnormal numbers (shaded in yellow) are considered as outlier and are excluded from further study. Other data are used for sketching unit performance.
····Investigation on CE unit
Figure 15: CE performance
As shown in figure 15, the only RMA verification that badly impacts CE’s performance happened at the end of Feb. There is no technical reason for the delay.
···· Investigation on Process Integration Engineer(PIE)/Technology Development
Engineer (TDE)
The same inspection is conducted with PIE/TDE data. Further study shows that the delay of RMA 3 and 14 are caused by handover between owners and holiday reason. Thus they are excluded from performance appraisal. According to the figure below, target-hit rate of PIE/TD units is less than 50% and average lead-time reaches 4.67 days that is 0.67days more than target.
Besides, verification within this unit suffers from unstable processing, which is indicated by the highest standard deviation value among units (STD=3.04). Unstable process will result in long waiting time based on the equation:
(
)
(
)
OT M OT CV WT ⋅ + ⋅ − = 2 1 1 2ρ
ρ
WT: waiting time, ρ: Capacity utilization,
CV: coefficient of variation (ratio of STD to mean)
Once RMA cases increase dramatically, waiting time (waste) increase accordingly if process stability remains unchanged.
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 1111 2222 3333 4444 5555 6666 7777 8888 9999 10101010 11111111 12121212 PIE/TD WD TARGET
Figure 16: PIE/TD performance ····Investigation on Account Manager (AM)
AM needs to review RMA claim from a business perspective. There is a gap between target and actual performance (average time=1.82 days while target=1 day). The only special case that should be excluded from data analysis is RMA No.1 because an AM hands over his job to another. According to Figure 17, there are two severely delayed cases, which are RMA No.2 and 3. The investigation so far shows no business issue causes the delays.
0.00 1.00 2.00 3.00 4.00 5.00 6.00 1111 2222 3333 4444 5555 6666 7777 8888 9999 10101010 11111111 12121212 13131313 AM WD Target Figure 17: AM performance
A more realistic performance review of each unit is provided below:
0 1 2 3 4 5 CE AM QE PI E/ TD QE Repor t Ave. t i me Tar get STD
Figure 18: Overall performance status
Based on data analysis above, main problems of each unit are summarized below:
1) CE: Low priority of RMA results in severe delay of one case, which is not acceptable by Gracesemi and customers.
2) PIE/TDE: Most of verification fails to meet target; besides unstable processing time is another problem.
3) AM: Similar to CE unit, low priority of RMA results in delay.
Lean concept and target review
Further investigation shows for CE and AM unit, value-added activities only take small part of lead-time target. However TDE/PIE suffer from tight lead-time target. It is difficult for them to finish tasks within the set time frame.
Considering RMA case at Gracesemi, 4 types of waste exist in the whole RMA process. They add little value as well as more lead-time.
Waiting time. During verification step, RMA claim is not immediately treated until the owner settles other tasks that have less lead-time target. Besides, during arrangement for returned wafers, there is long waiting time for a big batch.
Transportation. Applications for credit note have to get signature from CFO. The delivery of application from one department to another is transportation waste that needs to be dealt with.
Process. Verification procedure is designed based on current process. It is of great significance to reason whether the sequence of current process is optimal. For example, verifying whether paralleling AM and CE verification generates better result may significantly help improve effectiveness.
Defective goods. Low quality of verification and coordination—error data for example—definitely increase total lead-time. If verification errors occur, Gracesemi has to rework the whole process. Total lead-time is increased accordingly.
PIE/TDE fail to meet target most of the time. Thus a target review is needed to verify whether the target of each unit is reasonably set. Long cycle time in CE and AM department ask for further investigation in these two departments, where lean concept is introduced to eliminate waste and improve cycle time.
Lean thinking, as its name implies, is a mindset that focus on removing waste and increase customer value. Waste refers to all activities that don’t add customer value. According to Taylor et al. (2005), Leanness means developing a value stream to eliminate all waste including time, and to enable a level schedule. Whether an activity creates business value is the only criterion of deciding whether to remove or keep the activity. The criterion is taken from SOP where standard operation procedure is described. Before we move forward to slack identification in CE and AM unit, value chain analysis (VCA) is introduced for better understanding the investigation in CE and AM unit. CE unit verifies RMA claim from a technical perspective. The difference between CE and PIE/TDE is that CE verifies claim based on customers’ data and makes a go-or-kill decision according to agreement made with customer before. Activities that create value to verification process within CE unit are:
1) Check technical agreements for information such as agreed reject yield for mature products and certain criteria for pilot products.
2) Double check customer data to guarantee no error in raw data.
3) Contact customers via email or phone if there is any confusion about customers’ data
4) Download data for record purpose 5) Click the button to finish verification
in ERP system, CE can check information by entering product ID in searching bar. Raw data double check is to prevent double count of one same wafer. It occurs sometimes when there is system error. Occasionally CE needs to contact customers if they have questions about data, which definitely increase verification cycle-time. However customers usually understand and accept this kind of verification delay. Based on the survey result of the actual time CE spent for conducting the above-mentioned activities, a summary is provided in Table 6.
Item Description Actual time
1 Technical agreement checking Around 15 minutes
2 Customer raw data double checking Around 30 minutes
3 Customer Contact Ranging from 1 hour to 4 days
4 Download for record purpose Around 15 minutes
5 Confirmation of verification 1 minutes
Table 10: Actual production time for activities (CE)
According to the table, normal verification (without customer contact) requires only 1 hour. However the target set for them is one working day, which makes them feel loose and usually postpone verification assignment and give priority to other duties.
AM unit verifies RMA claim from business perspectives. Value-added activities are identified as follows:
1) Check compensation terms in purchase order.
2) Contact customers if there is any inconsistency between claim and previous contract. Settle issue as soon as possible
3) Download data for record purpose 4) Click button to close verification
Compensation terms are about how to compensate unqualified wafers. In other words, it is realized by wafers or credit note. Usually unqualified mature products are compensated by wafers since there are stable demands at end market while pilot wafers ask for compensation by credit note as demand changes constantly. The ERP system well supports PO tracking function which helps save a lot of time. When there are inconsistencies between contract and claim, AM have to contact and negotiates with customers, trying to settle them as soon as possible. According to survey regarding actual time spent on these activities, we summarize result as follows:
Item Description Actual time
1 Compensation terms checking Around 30 minutes
2 Customer contact Ranging from 1 day to 4 days
3 Download for record purpose Around 15 minutes
4 Confirmation of verification 1 minutes
Table 11: Actual production time for activities (AM)
Again there is a gap between this slack-free verification and target, which requires further scrutiny.
set and he believes staff tries the best to meet the target. In contrast, PIE/TDE ask for more time on their activities. 4 working days are quite tight most of time.
Investigtion shows that current target breakdown to some extent impacts verification lead-time negatively. Especially in CE and AM unit, the loosening target lower the priority of RMA verification and many unnecessary wastes are added at employee’s own will. A new version of target breakdown is presented in Chapter 5.
The VCA is also applied to whole process re-engineering in order to: 1) Understand current RMA process status
2) Identify key waste(s), issue(s) across whole process (discussed in 4.5) 3) Develop future state version of RMA process (discussed in 5.2.1)
4) Put forward action plans to achieve future state version (discussed in 5.2.2) Last but not the least, sales administrators complain about the newly launched ERP system of RMA verification. According to them, the system is not mature enough. Sometimes it crashes down and it takes time to recover, which delays RMA verification process and badly impact their mood for further processing.
4.2.2 Coordination with Imp & Exp
Once verification result justify customers’ RMA request, SA immediately starts to coordinate with Imp &Exp department in order to prepare for wafer shipping back. In this section the lead-time of Phase 2 is examined. Beginning with introduction to activities in Phase 2, the section continues with data flow analysis and finally points out the role of coordinator, which should have been taken by SA is actually taken by Imp & Exp department. Lead-time is unnecessarily increased since all data stop by Imp & Exp without adding any value (Imp& Exp don’t process the data; they just pass them to the next one). In addition, Imp & Exp face no pressure from customers; therefore they are more procedure-oriented rather than customer-oriented.
1) Activity Introduction
Sales administrators need to cooperate with Imp & Exp for the below four occasions: 1) Once Quality department needs a sample, sales administrators have to co-work with Imp & Exp for importing sample wafer. Since customs inspection is much simpler to sample delivery, usually it takes 1 day from airport to Gracesemi.
2a) If Final Report doesn’t justify RMA claim, sales administrators needs to arrange shipping back to the customer. Imp & Exp is involved again for preparing documents for customs claim.
2b) When Gracesemi justifies the RMA claim, sales administrators are informed to arrange wafer import with support from Imp & Exp. It is a complex process since many stakeholders are involved in order to deal with documents preparation for tax return. Lead-time of tax return may range from 10 days to 30 days.
3) Sales administrators are responsible for compensation wafers shipping. They cooperate with Imp & Exp for shipping documents and customs declaration.
Outside forwarder decides the lead-time of sample delivery and compensation wafers shipping. Both deliveries have a fixed lead-time. Therefore the research doesn’t include shipping.
Figure 19: Actor activity diagram (AAD)
2) Problem Analysis
As shown in AAD, Imp & Exp department has the most activities. It seems Imp & Exp is coordinating the whole process and keeps sales administrator informed when necessary. Once more explicit data needed, Imp & Exp comes back to sales administrator for additional data. Obviously there are time wasted in back and forth communication via Imp & Exp. Actually Imp & Exp should have not been the coordinator of RMA cases. Since Imp & Exp faces no direct pressure from customer, which makes them more procedure-focused rather than customer-oriented.
Figure 20: Data flow
As shown in figure 20, data provided by sales administrator first arrive at Imp & Exp and later Imp & Exp transfers the data to Finance unit, which handles business with outside stakeholder. There is no value added at Imp & Exp. The only thing that is added is transferring time.
SA Imp & Exp Finance Outside Stakeholder
