Harvard Business School
9-693-035
Rev. December 7, 1992
American Connector Company (A)
“Looks bleak.” Those were the words Denise Larsen, the Vice President of Operations at
American Connector Corporation said calmly as she read the report on a competitor’s plans to build a
new electrical connector plant in the United States. The competitor, DJC Corporation of Japan, had
become a dominant supplier of electrical connectors in its home market in recent years, after building
what was rumored to be one of the most efficient connector plants in the world. However, despite its
success in Japan, DJC was barely a contender in the U.S. market. The company had no plants in the
U.S. and only a small sales force there. Larsen knew that this could all change quickly. As she
explained to her assistant, Jack Mitchell, a recently graduated M.B.A.:
There have been rumors the last few years that DJC would build a new plant here to
launch an attack on the U.S. market. But with the market so crowded with
competitors and burdened with excess capacity, no one took them seriously here.
Either way, we figured that we still had a cost advantage. But if your report is right,
and if DJC can operate a plant here like the one they have in Japan, I think DJC could
quickly grab some market share here.
Larsen was worried because she felt American’s position was particularly vulnerable at the
moment. She was chiefly concerned with the company’s connector plant in Sunnyvale, California
since it had been struggling with a series of operating problems during the past year. Costs at
Sunnyvale were increasing while quality seemed to be deteriorating. In the past month, she and
Andrew Li, the new plant manager there, had discussed ways to improve the plant’s performance.
Now she wondered whether the DJC situation called for a completely new manufacturing strategy.
The Electrical Connector Industry in the Early 1990s
Electrical connectors were devices made to attach wires to other wires, attach wires to outlets,
attach wires, components or chips to PC boards, or attach PC boards to other boards. A connector
typically had two main body parts—a plastic housing and metal socket pins or terminals. The
housing was usually made of a plastic resin such as a polyester, nylon, or polycarbonate. The metal
pins could be made and plated with a number of different metals, ranging from tin to gold. Exhibit 1
illustrates several basic product designs.
The connectors were used in a variety of product applications, including military and
aerospace electronics, industrial electronics, telecommunications equipment, computer and office
equipment, automobiles, and consumer electronics and appliances. Each application—often each
Professor Gary Pisano prepared this case with assistance from Research Associate Sharon Rossi. The case was prepared as
the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation.
Data have been disguised for purposes of confidentiality.
Copyright © 1992 by the President and Fellows of Harvard College. To order copies or request permission to
reproduce materials, call 1-800-545-7685 or write Harvard Business School Publishing, Boston, MA 02163. No
part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in
any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the
permission of Harvard Business School.
1
This document is authorized for use by Michael Fleytman, from 8/7/2012 to 12/31/2012, in the course:
BADM 375: Business Process Management – Agrawal (Fall 2012), University of Illinois at Urbana-Champaign.
Any unauthorized use or reproduction of this document is strictly prohibited.
693-035
American Connector Company (A)
producer—called for different connector specifications. In 1990, there were over 700 standard
connector product lines in North America alone. Standard designs were those which had been
established by the International Institute of Connectors and Interconnect Technology (IICIT), the
National Electronics Distributors Association (NEDA), or by end use industries. Other designs were
custom-produced, usually on a one-year contract with a single connector company and were often
accepted as industry standards after the contract expired. Since each type of connector had its own
set of specifications, suppliers which produced many types of connectors (sometimes hundreds or
thousands of models) for different industries were finding it difficult to meet the increasing number
of specifications. In 1992, attempts were made to standardize product specifications among the many
different industry associations.
Because connector types were made of different materials and varied from low- to hightechnology, they varied dramatically in price, as well. For example, a simple connector such as a
phone jack sold for only a few cents, whereas a custom designed connector such as one used in
military electronics sold for several dollars. The cost of connectors used in any product typically
counted for 2% or less of the cost of the end product.
In the 1970s, the U.S. connector industry1 had experienced very rapid growth as firms built
up capacity to meet the growing demand (particularly for computer applications). But when demand
slowed in the mid to late ‘80s, there were too many suppliers and too much capacity. Price
competition intensified as more offshore producers entered the U.S. market.
In the 1990s, the U.S. connector industry was characterized as a hostile environment. There
were more than 900 suppliers and sales continued to slacken. In 1991, sales were down 3.9% from the
previous year, while the ten industry leaders were on average down 7.9%. The abundance of
suppliers gave customers leverage to demand reduced prices, improved quality, and faster delivery.
At the same time, many customers were working to reduce the number of suppliers they did business
with. OEMs that previously dealt with 10 or 12 connector vendors had reduced their suppliers to as
few as four. Pressures spurred a trend of mergers and acquisitions in the industry and analysts
predicted that the number of connector suppliers in the U.S. might drop to 400 or fewer by the end of
the 1990s.
Electrical connectors were very engineering intensive products and were critical to product
performance. As electronic circuitry became more miniaturized and operated at higher speeds, new
connectors had to meet more demanding requirements for space, weight, cost, quality, reliability and
performance.
In 1991, worldwide sales of interconnect products totaled roughly $16 billion. The top ten
worldwide leaders accounted for $6.67 billion, but the total industry (1,200 competitors) was very
fragmented. The dominant company in the industry was AMP, Inc. It held 16% market share with
sales of $2.6 billion in 1991. The second tier of companies consisted of six other companies, each of
which had sales in the $500 million to $800 million range. DJC and American Connector were among
the second tier companies worldwide. Companies in the third tier had sales in the $250 to $500
million range. In total, there were 28 firms with sales greater than $100 million.
1The "connector industry" included three types of interconnect products:
connectors, cable assemblies and
backpanels. While American Connector and DJC Corporation produced all types of interconnect products, the
Sunnyvale and Kawasaki plants manufactured only connectors.
2
This document is authorized for use by Michael Fleytman, from 8/7/2012 to 12/31/2012, in the course:
BADM 375: Business Process Management – Agrawal (Fall 2012), University of Illinois at Urbana-Champaign.
Any unauthorized use or reproduction of this document is strictly prohibited.
American Connector Company (A)
693-035
Profile of DJC Corporation
The DJC Corporation produced a variety of electrical connectors used in computers,
telecommunications, and consumer electronics. For these applications, DJC produced four basic
types of connectors: wire-to-wire, wire-to-outlet, item-to-board, and board-to-board.
There were several aspects characterizing DJC’s competitive strategy in most of its product
lines. First, the company cultivated and maintained close links with the major computer,
telecommunications, and electronics companies and distributors in Japan. These relationships
represented an important entry barrier in the Japanese connector market. Secondly, the company’s
design strategy emphasized simplicity and manufacturability over innovation. Early DJC product
designs were based on reverse engineering of other companies’ designs, including those of American
Connector. As one former manager of DJC explained:
In 1965, we copied other companies’ products. Americans made good products and
the U.S. market was the most advanced in the world. Our R&D was geared entirely
toward analyzing U.S.-made products, copying improvements as they were made.
By 1975, our quality was as good as theirs. After that, the production process became
the basis for competition.
However, DJC’s design strategy went beyond simply copying U.S.-made connectors. The
company paid very careful attention to customer and user needs in adapting American designs to the
particulars of the Japanese market. For example, DJC connectors were designed for maximum
compactness since this feature was very important to Japanese OEMs, particularly those producing
consumer electronics. DJC also adapted the designs to economize on raw materials (which were
nearly twice as expensive in Japan as they were in the United States) and to simplify manufacturing.
Features which did not add perceived value to customers (such as color-coded housings) were
eliminated.
Finally, and perhaps most importantly, DJC viewed highly efficient manufacturing as
absolutely critical to its competitive strategy. The company historically relied upon manufacturing as
the major means to achieve their overall profit goals. As one former Managing Director of DJC
described it:
In electrical products, high quality is a prerequisite for success. With large established
competitors fighting for a maturing market, a low cost position becomes necessary
for long term success. Manufacturing excellence is the source of both and has
therefore been at the heart of DJC.
The importance of manufacturing to DJC was reflected in the organization of the company
(see Exhibit 2). For example, Mr. Okada, the head of production, was responsible for the operations
of four domestic factories and reported directly to the company president, Mr. Esaka. In addition, the
balance of power between manufacturing and the sales/marketing division was clearly tipped in
favor of manufacturing. For example, sales/marketing had little power to alter production schedules,
product mix or lead times. As one former manager at DJC explained: “Sales sometimes needs an
unscheduled delivery, but manufacturing just does not allow it. There isn’t even any debate.”
DJC’s President, Mr. Esaka, was considered to be a dominant, hands-on leader. He was
hand-picked by DJC’s founder to become the company’s president in 1971. Many within DJC viewed
Esaka as the decision maker within the company. As one former executive put it: “Our strategy is
pure and strict, driven by Esaka himself. People may bicker back and forth…but everyone knows
they are personally responsible to achieve the goals set out by the president.”
3
This document is authorized for use by Michael Fleytman, from 8/7/2012 to 12/31/2012, in the course:
BADM 375: Business Process Management – Agrawal (Fall 2012), University of Illinois at Urbana-Champaign.
Any unauthorized use or reproduction of this document is strictly prohibited.
693-035
American Connector Company (A)
The Kawasaki Plant
During the early 1980s, the Japanese connector market experienced increased labor and raw
material costs, a rising yen, and increased import penetration. Amidst these conditions, top
executives at DJC were concerned that the company might not be able to maintain its historical 50%
gross margins in the connector business. This concern led Mr. Esaka to formulate his vision of a plant
which could achieve “the ultimate rationalization of mass production.” His vision called for a highly
automated, continuously operating plant which could meet the following three goals: First, the plant
must achieve asset utilization of 100%. Secondly, yield on raw material must reach 99%. Finally,
customer complaints could not exceed 1 per million units of output. While cost was not stated as an
explicit goal, everyone understood that if the above three goals were met, the plant would be one of
the lowest cost producers in Japan.
DJC chose to build its “ultimate” new plant in Kawasaki, Japan; the plant was completed and
began operating in 1986. Management chose the Kawasaki site for several reasons. From a logistical
point of view, Kawasaki offered the advantage of being located close to the major Japanese electronics
companies. In addition, perhaps more importantly, Kawasaki was near the major raw material
suppliers. This was particularly important because it was anticipated that most raw materials would
be delivered from vendors on a daily or weekly basis. The Kawasaki area also had an ample supply
of relatively young, highly skilled workers.
The Kawasaki plant was designed to produce a maximum of 800 million connectors per year,
assuming 100% utilization. Initially, the plant produced only 80% to 90% of this volume. About 75%
of this output was sold in Japan and 25% was sold in developing Asian markets outside of Japan. The
plant operated 24 hours a day, seven days per week, 330 days per year. The main advantage of
running the plant on a nearly continuous basis was that it avoided start-up and shut-down costs.
To successfully operate the plant on a continuous basis, as well as meet Esaka’s yield and
quality goals, Kawasaki management carefully integrated decisions and policies related to the
product and process technology, workforce, production control, quality, and organization.
Plant Layout The Kawasaki plant was organized into four large cells, each of which was
responsible for producing one of the four general types of connectors (wire-to-wire, wire-to-outlet,
item-to-board, and board-to-board). With the exception of plating, all of the processes needed to
manufacture a complete connector were located in each cell. Plating was organized separately in
order to fully utilize the high fixed cost equipment and to protect the rest of the factory from exposure
to corrosive chemicals and noxious fumes.
Each cell contained anywhere from two to six production lines, with each line consisting of
terminal stamping, housing molding, assembly, and packaging. Each production line was
responsible for producing a specific family of the cell’s products. Successive processing stages in each
line were located close to one another and in a straight line in order to minimize materials handling
steps and to reduce as much as possible the distance work-in-process had to travel. For example,
each plastic molding press in a cell was located only a few feet from the line’s assembly operations.
Operations were synchronized so that completed housing parts flowed almost continuously (via
small bins) between each molding and assembly line. Because molding and terminal stamping
equipment had shorter cycle times than assembly equipment, these processes were run below their
top speeds in order to synchronize parts fabrication and final assembly. Assembly operations were
almost completely automated. After assembly, connectors were inspected and transported a few
yards to the cell’s packaging area. In packaging, connectors were sealed individually in plastic on
strips containing 2,000 units.2 Each strip or “tape” was loaded onto a large reel. The central shipping
2Reels of 1,500 units were considered standard in the industry.
4
This document is authorized for use by Michael Fleytman, from 8/7/2012 to 12/31/2012, in the course:
BADM 375: Business Process Management – Agrawal (Fall 2012), University of Illinois at Urbana-Champaign.
Any unauthorized use or reproduction of this document is strictly prohibited.
American Connector Company (A)
693-035
department was responsible for packing each customer’s order for shipment. The Kawasaki plant
delivered to many customers on a daily basis and some of the largest customers received shipments
every few hours.
Product Technology Product design reflected the goals of continuous and reliable
operations and the need to economize on raw materials. Before commencing production at Kawasaki,
product designs were thoroughly analyzed to determine ways in which the product might be made
easier to manufacture and use less material. For example, product design of most connectors was
standardized to reduce the number of product variations. In 1991, the plant produced only 640
different stock-keeping-units (SKU’s), a relatively small number for a plant its size. The limited
number of product variations, it was believed, reduced the costs and complexity associated with
shorter production runs. To economize on the use of raw material, designers adapted some types of
connectors to use pins plated with tin rather than gold. Though gold was the most reliable and
durable material, tin was far cheaper and worked well in low power applications. To further simplify
production and reduce costs, DJC packaged its connectors only on tape and reels. This packaging
was particularly suited to customers with automated production environments, and it did not
inconvenience customers with manual operations. DJC’s engineers undertook extensive valueengineering to identify and implement cost saving design changes which did not compromise
product quality or performance. These design changes are discussed in Exhibit 3.
Process Technology Process design activities reflected several basic principles. First,
while the plant was to be highly automated, significant resources were devoted to what DJC called
“pre-automation.” Pre-automation referred to the activities required to make the production process
suitable for highly reliable automation. It reflected the philosophy that a production process could
only be automated after it was completely understood, properly designed, and properly laid out. To
automate before might mean automating a process which was inherently inefficient or unreliable.
During pre-automation, process flows were carefully analyzed to determine ways in which the
process could be streamlined and inventories eliminated. Worker movements and motions were also
studied to identify ways in which the process could be made more efficient. Pre-automation activities
also included specifying raw material quality and process tolerance levels.
There were several examples of how pre-automation problem-solving affected the process.
The warehouse was centrally located to simplify material flows and to economize on space. The
amount of warehouse and floor space was intentionally limited so that there would be no room for
excessive raw material or in-process inventories. To simplify material flows, each injection mold for
plastic parts had a dedicated press and each press was dedicated to a single assembly line. Each
assembly line was laid out in a continuous straight line from stamping to packaging. This made it
possible for one operator to run two assembly lines. Only after the process had been “preautomated” would steps be taken to implement automation.
The second principle guiding process design was the notion that it was better to use an old,
reliable process than a new, less reliable one. Rather than taking chances with new technology, the
plant relied on continuous improvement of existing proven processes. Reliable process technology
was considered absolutely essential to keeping the process running smoothly, without inventory, on a
continuous basis. To further ensure smooth runs, processes were generally operated below
maximum speed.3 Emphasis was also placed on maintaining equipment with the goal of eliminating
unscheduled downtime.
A third principle guiding process design at Kawasaki was the emphasis on absolute
reliability in upstream molding processes. According to one former DJC executive, “DJC views
molding as the most critical part of the manufacturing process and has focused its efforts there.” Its
3For example, an assembly line in the Kawasaki plant ran at 200 units/minute. By comparison, a similar line in
American Connector’s Sunnyvale facility ran at 500 units/minute.
5
This document is authorized for use by Michael Fleytman, from 8/7/2012 to 12/31/2012, in the course:
BADM 375: Business Process Management – Agrawal (Fall 2012), University of Illinois at Urbana-Champaign.
Any unauthorized use or reproduction of this document is strictly prohibited.
693-035
American Connector Company (A)
molding group included experts in polymer physics and many former employees of mold
manufacturers. Several policies and decisions reflected DJC’s desire to make molding virtually
faultless. Molds were subjected to rigorous repair and maintenance schedules. For example, every
mold received full maintenance every six months in addition to daily basic maintenance. Molds were
also replaced relatively frequently to reduce the chance of failure and to allow mold technology to be
upgraded. The average life of a mold at Kawasaki was three years. Taking into account purchase
costs, maintenance costs, and costs of repair, Kawasaki’s average annual cost per mold was $29,000.4
The plant achieved mold yields in excess of 99.99%.
A fourth element of Kawasaki’s approach to process technology was its reliance on in-house
technology development. DJC’s strong in-house process engineering competence was a result of both
historical conditions and of strategic choice. One former employee described the historical conditions
shaping the company’s in-house process engineering expertise: “Because we were a small company
and couldn’t afford to buy equipment, we built a lot of it ourselves. There were extensive workshops
in every factory.”
During its early years, the company developed ideas about manufacturing by looking at the
operations of emerging Japanese role models like Toyota. The decision to develop technology inhouse was also strategically motivated. In Japan, contractual agreements with equipment vendors
preventing resale to competitors were not common. Thus, DJC worried that its ability to achieve a
competitive edge in process technology would be severely limited if it relied too much on equipment
vendors. Indeed, part of this strategy was shaped by observing American Connector’s experience
with equipment vendors. As one DJC engineer recalled: “American Connector didn’t develop their
(connector) assembly machine themselves, they asked an equipment manufacturer to help them
design the machine. That same equipment company then offered to sell us the identical machine.”
Thus, while the Kawasaki plant might buy standard equipment from vendors, it would make all
proprietary design modifications in-house. In addition, Kawasaki designed all of its molds in-house
and manufactured about half of them in-house as well (the most complex molds were all produced
in-house). Kawasaki’s goal was to eventually build 100% of its molds in-house.
A final element of Kawasaki’s technology strategy was the inter-functional coordination of all
its technology development activities. The plant’s “Technology Development Division” was
responsible for coordinating and managing the activities of the product planning section, the
materials section, process engineering, and the molding technology group. Each section was assigned
a specific set of goals. For example, it was the job of process engineering to design and modify new
equipment and to identify opportunities for automation. However, it was the job of the Technology
Development Division to ensure that all of these sections were working in concert to achieve a
consistent set of explicit goals. These goals included: efficient resource utilization, design quality and
manufacturability, smooth manufacturing introduction, shorter development cycle, and continuous
process improvement.
One example of how this worked in practice was the development of a new resin to improve
connector durability. The materials planning group solicited input from the product planning group
on customer needs and requirements. Through this contact, the materials group learned that a more
durable connector would help differentiate DJC’s product in the customer’s eyes. This led to a
discussion with members of the R&D group about possible material breakthroughs which might
create a more heat-tolerant and damage-resistant connector. At the same time, the materials group
also held discussions with the process engineering group to le…
