Using Six Sigma To Improve Replenishment Process in A Direct Selling Company
Using Six Sigma To Improve Replenishment Process in A Direct Selling Company
Using Six Sigma To Improve Replenishment Process in A Direct Selling Company
Abstract
Purpose – The purpose of this study is to propose a Six Sigma project to improve the replenishment process in the logistic center of Amway Taiwan
Company.
Design/methodology/approach – This study elucidates the define, measure, analyze, improve and control (DMAIC) methodology of Six Sigma, using
qualitative and quantitative tools. Numerous improvement actions are implemented. A control plan is developed to sustain and enhance the
performance of the replenishment process.
Findings – The project has substantially benefited the logistic center, including reducing replenishment frequency and urgent shipping rate, decreasing
cost of transportation, lowering the inventory of service centers and shortening the replenishment planning time.
Research limitations/implications – The methodology that is employed in this study can be adopted by the logistic centers of other direct
companies. However, they can modify the detailed contents and tools according to their organizational conditions.
Practical implications – This study demonstrates a project in which Six Sigma is adopted to improve the replenishment process of the logistic center
in Amway Taiwan. It describes how the project was specified, how the tools were employed in the different phases and how the improvement actions
were implemented. In conclusion, the key benefits of, and experience gained from this project are emphasized.
Originality/value – This paper will be of interest to academic researchers and practical managers. The structured methodology of Six Sigma integrated
logistic techniques was applied in a leading direct sales company. The project in the logistic center of Amway Taiwan provides a good reference for
other companies that plan to implement Six Sigma management.
Keywords Six sigma, Distribution management, Direct selling, Supply chain management
3
Using Six Sigma to improve replenishment process Supply Chain Management: An International Journal
Chun-Chin Wei, Gwo-Ji Sheen, Cheng-Ting Tai and Kuo-Liang Lee Volume 15 · Number 1 · 2010 · 3 –9
4
Using Six Sigma to improve replenishment process Supply Chain Management: An International Journal
Chun-Chin Wei, Gwo-Ji Sheen, Cheng-Ting Tai and Kuo-Liang Lee Volume 15 · Number 1 · 2010 · 3 –9
approximately NT$400,000 per year after the project KPIVs’ failure modes were discussed by using the Failure
completion. Especially, all service centers will offer correct Mode and Effect Analysis (FMEA). The Risk Priority
and sufficient products to IBOs, but the inventory of centers Number (RPN) of each cause of a KPIV was calculated by
cannot increase. multiplying the severity, the frequency of occurrence and the
current detection ability ratings. Table IV presents the FMEA
3.2 Measure phase before improvement.
The aim of the measure phase is to narrow down the
problems which the project team wants to solve to a few vital 3.3 Analyze phase
few input factors (Raisinghani et al., 2005). The input The team thought that lack of consideration of the sales
variables (Xs) of process mapping were identified by using a characteristics of different products, the time of the month
fishbone diagram and associated with six principal and the ability of service centers to draw up replenishment
considerations, man, machine, material, method, measure plans would be major causes of excess inventories or
and mother nature (namely 6M). The replenishment process shortages. They tried to classify products based on their
comprised five sub-processes with 30 steps. The five sub- average sales volumes in the preceding three months. After
processes included; making the replenishment plan, statistical simulating, they divided all products into three
modifying the replenishment plan, outputting replenishment categories:
reports, picking correct items and volumes and delivering 1 Product category A: General items. The sales volumes of
them to service centers, and receiving goods by service these items are stable. About 200 items are in this
centers. category. Based on average sale volumes in the preceding
To clarify the inputs that affect the KPOVs, a Cause and three months, these products represented 90 percent of
Effect (C&E) matrix was constructed. The KPOVs were rated the total sales volumes of all items.
according to their importance, while the inputs were scored in 2 Product category B: Slow-moving items. Sales of these items
terms of their influence on the KPOVs. Particular key process are variable. More than 400 items are in this category.
input variables (KPIVs) with high priority scores were Based on average sales volumes over the preceding three
selected. Table III presents parts of the C&E matrix. months, they represented less than 10 percent of all sales.
The failure modes of these KPIVs, effects and causes of the 3 Product category C: New products. A new product is one
KPIVs’ failure modes and current control plans against the that has been on sale for no more than seven days.
5
Using Six Sigma to improve replenishment process Supply Chain Management: An International Journal
Chun-Chin Wei, Gwo-Ji Sheen, Cheng-Ting Tai and Kuo-Liang Lee Volume 15 · Number 1 · 2010 · 3 –9
Table IV FMEA
Current
Input Failure mode Effect SEV Cause OCC controls DET RPN Actions
Completeness System cannot Replenishment plan 10 Replenishment rules are 7 N/A 10 700 Redesign replenishment
of information provide sufficient cannot meet the too simple, so information system
replenishment information for good requirements of provided by system is
system offered replenishment service centers incomplete
planning
Workload of planners Department of Inventory
is heavy as Management provides
replenishment plan requirements for
needs to be artificially redesigning system
modified
Reference Reference period Frequency of emergent 10 System does not consider 10 Planning 7 700 Redesign replenishment
period for cannot reflect future shipping is high variation in sales with time based on system
calculating demand of month in calculation of sales of
replenishment replenishment volume previous
quantities week
Replenishment Forecast demands from
volumes do not meet historical sales data
demand
Available space Planners do not Planners limit 7 Available space for stock at 7 N/A 10 490 Discuss and establish
for stock at know available replenishment service centers is neither mechanism for
each service space for stock at volumes defined nor reported monitoring available
center service centers space for stock at
service centers
Product type, period of the month and service center were concluded that the sales volumes varied significantly with the
chosen as the factors for a three-factor Balanced Analysis of various product types, periods and service centers.
Variance (Balanced ANOVA). The product types were The main effect plots (Figure 1) also demonstrate that the
divided into the aforementioned categories A and B. Since three main factors significantly affected sales volumes. The
product category C concerns new products on which no effects of the interaction between the product type and the
historical sales data are available, it was not considered. The time of the month factors and between the product type and
sale period had three levels: the service center factors were significant. Based on these
1 the first ten-day period; results, they had to establish replenishing rules according to
2 the middle ten-day period; and the sales volume of each product type in each period within
3 the last ten-day period of a month. the month and for each service center.
All seven service centers were considered. The team collected
historical sales data from the last 12 months for analysis. 3.4 Improve phase
Balanced ANOVA was carried out at a 5 percent significance A number of improvement actions were implemented:
level and 5 percent power by a statistical software, Minitab. 1 Replenishing products based on different rules for the A, B and
Table V presents the test results. Since all p-values for the C categories. As the sales volumes of category A products
three main factors were significantly lower than a, the team are stable, the team decided to calculate the inventory
6
Using Six Sigma to improve replenishment process Supply Chain Management: An International Journal
Chun-Chin Wei, Gwo-Ji Sheen, Cheng-Ting Tai and Kuo-Liang Lee Volume 15 · Number 1 · 2010 · 3 –9
level of such products by inventory-holding-days. The need to ship to service centers daily. They attempted to
sales volumes of category B products are uncertain. reduce the frequency of replenishment to the service
Accordingly, the team decided to calculate the inventory centers. Finally, the logistic center replenished to service
level by Each (EA). The reorder point (ROP) and safety centers three times a week.
stock for each item at all service centers were established 5 Generating a low inventory warning report. The low
by simulation. The inventory level of each item at each inventory warning report lists items whose inventory
service center after replenishment may not exceed six levels are lower than safety stock levels. The
weeks of sales volume for that product. The replenishment system can provide low inventory warning
replenishment quantity of a new product (product reports on a specific item or all items, and for one service
category C) is determined by the forecast made by the center or all service centers.
sale department. After seven days of sales, the new 6 Investigating trends in new product sales. The team collected
product is classified as belonging to category A or B. The historical sales data on the first seven days of sales for a
replenishment system was redesigned to suggest variety of new products to investigate trends in those data.
automatically replenishment plans based on the rules for They established some decrease rates for sales of various
categorizing products. product types during the first seven days of sales.
2 Investigating the space available for stock at service centers. Replenishing planners can use these decrease rates of
The team decided that the available stock volumes of the the same product type to predict the demand for a new
racks plus the volumes provided by each warehouse equals product and modify the replenishment volumes based on
the available space for stock at a service center. This space actual sales conditions. The team also initiated a training
was calculated in units of inventory-holding-days. The program that covered information system functions, basic
usage rate of inventory space at a service center is the statistical tools and plots, new product sales trends and
present average number of inventory-holding-days of all the characteristics of products and service centers.
products divided by the available space for stock at the
service center. The total available space for stock and the 3.5 Control phase
maximum volume of stock at each service center will be The control plan helped to ensure that the improved
reviewed every six months. replenishment process actions became institutionalized.
3 Setting the reference period for calculating replenishment Table VI shows parts of the control plan.
volumes. The new replenishment rule for an item is based The project took approximately seven months after the
on its average sales volume in the preceding three months. initialization date. Some benefits have been obtained from this
For example, if planners want to replenish an item to a Six Sigma project:
service center from 4/1 to 4/7, they can determine the .
Service centers were replenished three times a week. The
replenishing volume from its average sales volumes over Finance Department estimated the annual transportation
the corresponding periods in the preceding three months fee savings to be NT$550,000 dollars.
1/1 to 1/7, 2/1 to 2/7 and 3/1 to 3/7 at the service center. .
The frequency of replenishment of service centers was
4 Reducing frequency of replenishment. Since the shipping time reduced, but number of abnormal shipments did not
from the logistic center to all service centers is less five increase, but decreased. The urgent shipping rate
hours and all service centers had ample available space for decreased from 0.75 percent to 0.51 percent within
stock, the team believed that the logistic center does not three months after the project had ended.
7
Using Six Sigma to improve replenishment process Supply Chain Management: An International Journal
Chun-Chin Wei, Gwo-Ji Sheen, Cheng-Ting Tai and Kuo-Liang Lee Volume 15 · Number 1 · 2010 · 3 –9
.
Although directors of service centers did not record and customers, the team can understand their requirements and
calculate the frequency of IBOs’ complaints, they enhance their satisfaction.
indicated that IBOs were satisfied with the improved The Six Sigma project facilitated effective problem
replenishment condition. definition; allowed for the use of data and systematic tools
.
The average time taken by a planner to make a rather than emotion and conjecture during critical decision-
replenishment plan of a service center was reduced from making. The team eliminated and controlled the correct
60 minutes to less than 40 minutes. sources of variation, and thereby enhanced process
.
The planners only made the replenishment reports three performance. Then, Six Sigma provided a good platform
times per week, but they needed to make the reports six between managers and employees for managing consistent
times every week prior to improvement. The workload of improvements.
planners was markedly lightened. Top management support is very important for
.
The number of shipping errors were reduced from 5 times implementing Six Sigma to provide a long-term
to 2 times, even to 0, per month. commitment and resources. In this project, a formal kickoff
meeting that was attended by top management announced the
start of the project. They were involved in the project by
4. Lesson learned providing sufficient budget, time and people, attending some
From this Six Sigma project, Amway Taiwan learned some important meetings and reducing the conflicts between the
critical lessons. project team and the service centers.
Six Sigma begins and ends with customers. This project In the future the requirements of environment and
began with the definition of customer requirements and made customers will become stricter and the replenishment
efforts to reduce the gap between the logistic center and process would need to change. This project established
customers (service centers and IBOs). The team members many useful documents of the replenishment process, like
included customers, the directors of service centers, and kept process flow, IO map, FMEA form and control plan. These
on communicating with them. Through dialogue with documents offer a good database for process knowledge
8
Using Six Sigma to improve replenishment process Supply Chain Management: An International Journal
Chun-Chin Wei, Gwo-Ji Sheen, Cheng-Ting Tai and Kuo-Liang Lee Volume 15 · Number 1 · 2010 · 3 –9
management. The process owner and manager can effectively Kwon, I.W. and Suh, T. (2005), “Trust, commitment and
modify the replenishment process by using the database to relationships in supply chain management: a path analysis”,
satisfy the requirements of customers. Supply Chain Management: An International Journal, Vol. 10
The project team, facing the resistance of employees against No. 1, pp. 26-33.
the implementation of Six Sigma, had to understand who and Kumar, M., Antony, J., Singh, R.K., Tiwari, M.K. and Perry,
what would be influenced by process changes. Education and D. (2006), “Implementing the lean sigma framework in an
training helped people to understand better the techniques of Indian SME: a case study”, Production Planning & Control,
Six Sigma that were adopted. A clear communication plan Vol. 17 No. 4, pp. 407-23.
and channels were required to overcome resistance, and to Rahman, S. (2006), “Quality management in logistics:
enable managers, employees, and even customers (service an examination of industry practices”, Supply Chain
centers) to be educated on the benefits of Six Sigma. Management: An International Journal, Vol. 11 No. 3,
The organized use of project management tools is pp. 233-40.
imperative. In this project, the schedule was delayed by Raisinghani, M.S., Ette, H., Pierce, R., Cannon, G. and
about one month since the Department of Information Daripaly, P. (2005), “Six Sigma: concepts, tools, and
Management was busy on other projects. Fortunately, the applications”, Industrial Management & Data Systems,
progress of the project was tracked by periodic meetings. The Vol. 105 No. 4, pp. 491-505.
project was well documented to ensure that outputs would be Tang, L.C., Goh, T.N., Lam, S.W. and Zhang, C.W. (2007),
delivered and objectives were reached in every phase. “Fortification of Six Sigma: expanding the DMAIC
The successful DMAIC approach described herein has toolset”, Quality and Reliability Engineering International,
been applied to other processes in the logistic center of Vol. 23, pp. 3-18.
Amway Taiwan. The integration of Six Sigma with other Yang, H.M., Choi, B.S., Park, H.J., Suh, M.S. and Chae, B.
existing management practices, like ISO 9000, can maximize (2007), “Supply chain management six sigma:
the positive effects of Six Sigma. The most important thing is a management innovation methodology at the Samsung
that quality concepts must be embedded into the design of Group”, Supply Chain Management: An International
process and that just monitoring quality at an operational Journal, Vol. 12 No. 2, pp. 88-95.
level does not suffice. Zylstra, K. (2005), “Distribution made lean”, Industrial
Engineer, Vol. 37 No. 1, pp. 31-5.
References
Furterer, S. and Elshennawy, A.K. (2005), “Implementation
Corresponding author
of TQM and lean six sigma tools in local government:
a framework and a case study”, Total Quality Management, Cheng-Ting Tai is the corresponding author and can be
Vol. 16 No. 10, pp. 1179-91. contacted at: andy_dai@amway.com