Harvard Business School

9-399-011
Rev. April 9, 2001

Acer America: Development of the Aspire
In early 1998, Stan Shih, CEO of Taiwan-based personal computer (PC) manufacturer Acer, Inc., was reviewing the first estimates of 1997 year-end results. With revenue of $6.5 billion from own brand and sales to original equipment manufacturers (OEMs) such as IBM, the company was now acknowledged to be the third largest PC manufacturer in the world. Although the performance was respectable in the wake of a dramatic drop in memory chip prices that had plunged the company’s semiconductor joint venture into losses, Acer’s extraordinary growth period of the mid-1990s was clearly over. (See Exhibit 1.) The ever-restless CEO was wondering how to re-ignite the fire. Shih was convinced that Acer’s mid-1990 successes were due at least in part to the revolutionary “client-server” organizational structure he had introduced in 1992. The concept was inspired by the network computer model, where “client” computers—the strategic business units (SBUs) and regional business units (RBUs) in Acer’s organizational metaphor—were capable of complete independence but could also take on the “server” role, adding value for the entire network. To Shih, proof of the client-server structure’s potential had come with the 1995 introduction of the Aspire multimedia home PC. Created by Acer America Corporation (AAC), Acer’s U.S. marketing subsidiary and one of Acer’s five RBUs, this new product confirmed Shih’s belief that major initiatives with global potential could be led from any part of the organization without centralized headquarters control. But Aspire’s difficult development experience and its less-than-successful global rollout had also highlighted some of the deficiencies in the client-server model. Business unit independence had resulted in problems in communication, project ownership, product proliferation, and transfer pricing—and, in the end, had led to Aspire’s $100 million of losses in the U.S. alone. Shih realized he had to find a way to balance independence with control, but did not want to sacrifice the employee initiative and entrepreneurial spirit he believed the client-server organization had released. Reviewing the lessons from the Aspire, he wondered what changes might be necessary to Acer’s radically different strategic and organizational concepts if the company was to grow the Acer brand from its current position as number eight to one of the world’s top five PC brands.

Professor Christopher A. Bartlett and Research Associate Anthony St. George prepared this case as the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation. Certain data have been disguised to preserve confidentiality, but important relationships have been retained. Copyright © 1998 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

399-011

Acer America: Development of the Aspire

Acer’s Growth and Expansion1
In 1976, with capital of $25,000, Stan Shih, his wife, and three friends established a company they called Multitech to commercialize microprocessor technology in Taiwan. The culture Shih established within the young company was built on three strong foundations—a commitment to entrepreneurial initiative (which translated in mistakes being viewed as tuition for learning, for example), a cost consciousness and financial conservatism that Shih described as a “poor man’s philosophy” (financing was largely through employees, for example, who typically took equity instead of market salaries), and a sense of family that was based more on delegation to responsible sons and daughters than on the traditional Chinese company’s nepotism and paternalism. In the mid-1980s, with a product strategy that had finally focused on personal computers, the company (known as Acer after 1987) began expanding internationally pursuing a strategy Stan Shih described as the “Dragon Dream.” Determined to take on the large established PC companies—“the nobility”—he coopted his employee-owners, supplier-partners, distributors and mass-market customers into a partnership he described as a “commoners’ culture.” This loose alliance not only allowed Acer to reduce its own capital investment and risk, but also obtain committed partners with knowledge of their local market. To sidestep “the nobility,” Acer pursued a strategy based on the Asian board game, Go. Using its limited resources to build strength in the corners, the company initially entered smaller developing markets before finally building positions in Europe and the U.S. In the late 1980s, however, intense competition caused industry prices to drop about 30%, and Acer’s gross margins fell from about 35% in 1988 to about 25% the following year. A public stock offering supported its continued aggressive expansion, but unfortunately, several costly acquisitions caused the company to become overextended. To bring discipline to the company, the loosely organized entrepreneurial organization was restructured into strategic business units (SBUs) responsible for product development and manufacturing, and regional business units (RBUs), responsible for marketing and distribution (see Exhibits 2a and 2b). But losses continued, totaling $26 million on sales of $1 billion in 1991. As price wars continued to erode Acer’s margins to around 20% in 1992, Shih realized that Acer’s business model had to change. Shipping from centralized manufacturing in Taiwan not only delayed new product time-to-market and increased inventory, it also exposed Acer to widespread import duties on hi-tech products. A team of engineers solved the problem by creating “Uniload” assembly system whereby carefully configured palettes of snap-together parts and components were shipped from Taiwan by air or sea depending on weight, value added and price volatility, to assembly centers around the world. Lower value-added and less critical components were sourced locally. Shih labeled this new way of doing business the “fast food” model because of its similarity to McDonald’s hamburger assembly approach of delivering products “hot and fresh” close to the customer. The radical business model change reflected Shih’s belief that value added in the PC industry was rapidly migrating away from assembly, upstream to component design and software development, and downstream to branding and distribution. It was a concept he captured his “Smiling Curve” diagram (Exhibit 3). Concurrent with the establishment of dozens of offshore assembly centers, Shih continued to delegate greater decision-making authority. Under an organizational concept he described as the “client-server” model, RBUs were given greater freedom to configure the products to fit their local markets. Coupled with the long-established norm of delegation and the more recent emphasis on profit responsibility, this concept operated under the slogan “every man is lord of his castle.” The change was not tension-free, however. As RBUs established a greater degree of independence, SBUs felt their role was shifting from product designers and manufacturers to components suppliers.

1 For a detailed history, see “Acer, Inc: Taiwan’s Rampaging Dragon,” HBS. No. 399-010.

2

Acer America: Development of the Aspire

399-011

As the company evolved towards a federation of locally responsive units close to their markets yet linked by the emerging Acer brand, Shih dubbed the emerging management philosophy “Global brand, local touch.” By encouraging business units to enhance “local touch” through local public listing, Shih’s hope was to achieve “21 in 21,”—21 locally owned subsidiaries by the twentyfirst century. By early 1998, two RBUs and three SBUs were publicly listed companies. By 1998 Acer had grown to 23,000 employees in 129 companies with 17 manufacturing plants and 30 assembly facilities in 24 countries. Through the client-server structure, Acer sales units were located in 44 countries, with particular strength in the developing country markets. Although Acer was only the eighth largest brand for PCs in the world overall (more than half its sales were still to OEM customers), it was the number one brand in 12 countries in Asia/Pacific, South America, and the Middle East, and in the top five in over 30 countries (see Exhibits 4a and 4b).

AAC and the Birth of the Aspire
Established in 1977 as a U.S.-based sourcing company, Acer America (AAC) was converted into a sales office in 1986. Acquisitions of two network computer manufacturers (Counterpoint and Altos) and a customer service organization (Service Intelligence) plunged the subsidiary into losses beginning in 1988. Losses continued in the early 1990s as industry-wide price cuts severely reduced margins. But through cost controls and the inventory reductions made possible by the Uniload process, AAC recovered. In 1994, AAC recorded its first profit in six years. (See Exhibit 5.) Hopes for further recovery were pinned to a new multimedia home PC named Aspire, a product local management believed could help implement Shih’s desire to build a strong global brand.

Local Inspiration: Creating the Concept
In 1994, as the world began to take note of developments in the Internet, the World Wide Web, and new audio, telecom, video, and computing technologies, several companies saw the potential for an advanced multimedia home PC. Among those sensing this opportunity was Michael Culver, AAC’s Director of Product Management. Although the company had been selling the Acer Acros, a slightly reconfigured version of its commercial desktop PC, through retail channels since 1990, its share of the huge U.S. home PC market was miniscule. With the newly granted authority to create local products, the 29-year-old MBA who had joined Acer 2½ years earlier leaped into action with his vision to create “the first Wintel-based PC that could compete with Apple in external design, ease-of-use features, and multimedia capabilities.” To test his ideas, Culver put together a project team which began running focus groups to examine market needs. One clear finding of this research was that, in addition to enhanced multimedia capabilities, consumers wanted a home PC that had a different look and feel than the standard putty-colored, boxy PCs that sat in offices throughout the world. The team also discovered it would have to move fast: product cycle time for PCs was 6 to 9 months and shortening. Not having local design capabilities (AAC had a staff of 20 engineers, mainly focused on software design and product testing) and believing that Acer’s SBU staff did not have the appropriate skills, Culver and his team looked to external design firms for help. Frog Design, a local Silicon Valley firm that had designed everything from bicycles to consumer electronics appealed to the Acer team because of its reputation for “thinking outside of the box.” At this point, Culver went to AAC President Ronald Chwang for $200,000 to fund the design phase and approval to go ahead with the project. “Especially since this was a decision to let an RBU design its own product for the first time, the process was incredibly informal,” related Culver. “It literally took place in one 20-minute discussion in the hallway in late November.”
3

399-011

Acer America: Development of the Aspire

The Aspire product management team and the designers visited computer retail stores and brainstormed the product’s external design. In two months Frog Design developed six foam model prototypes. In January 1995, the AAC team chose the final design: a sleek low-profile shape with rounded edges presented in a choice of colors—charcoal grey or emerald green. (See Exhibit 6.) Meanwhile, the team had been defining the multimedia capabilities built into the computer. One significant innovation was to be voice-recognition software that would enable users to manipulate programs by voice commands. But the innovations presented significant design challenges. In addition to the microphone and speakers built into the monitor, the PC would also have fax and telephone capabilities with a built-in modem and answering machine. Finally, the Aspire was designed for ease of use from set-up to start-up to operation. Color coded cables, graphic icons, simple exploration and navigation software were all part of the design.

Global Aspirations: Developing the Product
In the early stages, the design of the product had been top secret, and even Shih viewed it for the first time only in February 1995. His immediate reaction was positive and enthusiastic: ‘’This product will make Acer a household consumer electronics brand.” He immediately asked Culver to present the concept at the next meeting of RBU heads. By May a consensus had emerged to aim for a global launch of Aspire in September. With RBU heads’ support and Shih’s commitment to provide significant funds for “global” advertising, Culver and Chwang set out to present Aspire to SBU executives in Taiwan. Although the SBU heads liked the idea, their engineers felt they could be of only limited help at this stage of the development process. As one SBU executive recalled: Because the project was owned by AAC, our engineers did not have much influence during the integration phase. Most of the product design was dictated and decided by AAC and Frog Design, so when the SBU engineers came in it was a little too late. This was the first time that the product had been designed from the outside in and the radical housing created many challenges in getting our standard components to fit. When we designed products in Taiwan we worked from the inside out, so we had never run into this kind of problem before. Because AAC lacked product development experience, four Taiwanese engineers from the Information Products Group (IPG) and Acer Peripherals (API) were sent to the United States in March to aid in the mechanical design. As the project team continued to develop the cosmetic design behind closed doors, the SBU engineers working in parallel on component design realized they had no control over product design. “At first there was a little of the ‘not invented here’ syndrome,” Culver recalled, “and we had to work to make the SBU engineers feel a part of the process.” In May, the top-secret development room was finally opened and the design and integration of key components proceeded more easily. The computer’s complex multimedia system was broken down into sub-assembly systems, each of which was assigned to a manager and a design team. But the greatest challenge came in integrating the components and subassemblies into the final product. Culver explained: Different product managers across the globe were responsible for different sub-assemblies. One might be in charge of the speakers, another in charge of the voice-recognition software. I was in charge of overseeing the entire product development and had to coordinate over 70 different contacts. The biggest problems occurred when we ran into a delay with one of the sub-assemblies. For example, after we had selected a microphone for the monitor we found that the software didn’t recognize the input. We then had to isolate whether it was a problem with the
4

Acer America: Development of the Aspire

399-011

microphone, which would mean resourcing, or the software, which meant redeveloping the application. When a delay threatened to throw off the schedule for the next part of the system we had to make snap decisions and go with them. Similar problems occurred with many other components. With the answering machine, for example, the unit received the audio input signal but did not play back the message satisfactorily. Engineers had to decide whether the problem was in the software, the microphone input or the speakers, each of which was the responsibility of a different person. Aspire’s innovative design also led to complications. For instance, the unique shape of the computer casing created a problem with the opening and closing of the CD-ROM tray, while the challenge of matching designer colors on components caused additional headaches. In the end, no standard Acer parts were used in Aspire, and Culver estimated that specialized components added 10%-15% to the cost of items such as the CD-ROM units and monitor. The RBU-SBU coordination challenges were further complicated by the fact that most of Acer’s qualified suppliers were in Taiwan. As a result, the Taiwan-based SBU product managers had to source components and send them to AAC for approval and integration into the prototype. One engineer responsible for much of the sourcing described some of the challenges in Taiwan: AAC did not have the engineering capabilities that we had in the SBUs, but they didn’t trust us to do any of the integration. And it didn’t matter if we had an opinion on the quality because Acer America insisted on doing their own review. For example, somebody from the SBU had flagged the CD-ROM tray problem early on, but he wasn’t listened to. If AAC encountered a problem with a component, it took almost a week before I heard back from them. Several times an engineer from Taiwan had to get on a plane and hand-carry components to AAC to meet the subassembly deadline. In the end, we gave up questioning and just implemented what they told us. As a result, production dates kept getting pushed back and costs kept escalating. The large projected U.S. demand and the constrained supply forced the SBUs in Taiwan to allocate all component units to AAC, abandoning the original plan for a global launch in order to keep the September 1995 U.S. introduction on target to catch the Christmas season. Culver explained: The delays in integration caused us to delay manufacturing. Because the monitors were the last to be finalized, we had to ship them by air to get them to the United States in time. This cost almost $70 a monitor instead of the usual $10 by sea and added another $3 million to $5 million to our costs. Despite the difficulties, the product was designed and manufactured in record time, and th shipping began on September 5 —just nine months after concept definition. As Shih commented: I believe that if the origins of the Aspire were to have gone through the traditional communications channels—to file reports at headquarters with SBUs and to go through the process of argument, revision, and approval—it would have taken at least one-and-a-half to two years to see the final product. By that time, the daring creativity of the product may have already been diminished.

Planning Implementation: Creating the Marketing
Parallel to the product development activities, Culver was working on a marketing program for the product launch. Having decided that many of the targeted home users would be first time buyers, the team created a marketing concept they described as “OOBE”—the “out of box experience” of being able to set up and use the PC immediately. Priced from $1,199 for the basic
5

399-011

Acer America: Development of the Aspire

product to $2,999 for the high end system with monitor, the Aspire was positioned in the middle of the 15 to 20 percent price gap between the top-tier PC brands like IBM and Compaq and the low-end products like Packard Bell. Distribution would be through specialist computer electronics stores that had previously carried Acer’s Acros PC. They would take their standard retail markup of 10-12% plus an additional 5% marketing allowance. To build confidence in this largely unknown brand, the one year warranty Acros had offered would become three years for Aspire. Keen to build Aspire as Acer’s flagship brand in the U.S., Culver budgeted $25 million in advertising for four months following the launch. In addition, an intensive public relations program targeted computer trade publications like PC Magazine, the general business press like Business Week, and even lifestyle magazines like Architectural Digest. He also planned to support the launch with extensive merchandising and point of purchase materials. In all, this would be the most ambitious new product launch in AAC’s history—and even for the company as a whole.

The Launch
In September 1995, the first month of its introduction, Aspire sold 40,000 units at retail, double AAC’s normal monthly sales of the Acer Acros after 3 years on the market. When October sales leaped to 80,000 units, Shih, Chwang, and Culver were ecstatic. Early reviews in the technical magazines were also positive, reinforcing the euphoria in AAC. Caught up in the excitement, Chwang and Culver forecast retail sales into 1996 at around 100,000 units per month, triggering orders on Taiwan for components and subassemblies to meet this expected demand.

Supernova Burnout?
Then the unexpected happened: sales fell to less than 60,000 units in November and 35,000 in December. The first clue to the reason for the decline came from customer service, where telephone representatives were being swamped. As product return rates approaching 15% confirmed, the Aspire still had several minor technical problems that had not been completely resolved before the September deadline—the problem opening and closing the CD-ROM tray was typical. This situation was exacerbated by the fact that Aspire had been targeted to first-time computer buyers and many of the calls to customer service reflected their inexperience. “Customer service was getting killed,” explained Culver. “We couldn’t keep up with the calls. Some lasted as long as 30 minutes!” By the end of December, the AAC team finally acknowledged the need to cut dramatically their forecasts. By that stage, however, the company had already accumulated two to three months of excess inventory of a product that needed design changes. Although Culver and his team paused briefly to celebrate when Aspire was featured in Business Week’s annual awards for new product design in 1995, they were more concerned that sales had showed no rebound in early 1996. (See Exhibit 7 for AAC’s next month forecast and actual sales into channels.) While other RBU heads had all committed to launch of the Aspire, nobody was clear how this global rollout would be coordinated. Shih was very active behind-the-scenes urging the organization to support Aspire, but also encouraging Culver to take on an informal leadership role—presenting the product plans to top management meetings, hosting visits of other RBU teams to AAC, and updating all RBUs on progress and performance. Despite early tension over the decision that only AAC would be able to launch in September, the other RBU heads mostly deferred to Culver and his team as the product experts. Yet they were looking for advice rather than directives and remained highly protective of their right to adapt the Aspire strategy to local markets—particularly as they watched the problems emerging with the U.S. launch.

6

Acer America: Development of the Aspire

399-011

By early 1996, Taiwan had developed sufficient production capacity to begin a worldwide phased rollout of Aspire during the spring. However, each RBU argued that the U.S. product would have to be adapted to suit its local markets. Typical was Michael Mak, Managing Director of Acer Hong Kong, who described the process he adopted: From the first time I saw Aspire, I knew we had to have this product. Originally, we brought in the unaltered U.S. product for a “soft-launch.” Although the unadapted product generated some sales, most consumers were concerned by the $3,000 price tag we had to charge. At this stage, we realized we had to make some local modifications. We simplified the user interface and changed the language from English to Chinese, we included a card that allowed the use of a video-CD, we adjusted the phone software for the local telephone system, and we changed the modem. We also created a poster and videotape in Chinese and English showing the user how to get started. The Hong Kong office had few local technical staff to make these changes. Of 200 people in the Hong Kong organization, only three or four were in product engineering and these individuals coordinated most of the software modifications. The hardware changes were largely handled by ACI, Acer’s Asian RBU headquarters in Singapore. ACI negotiated directly with the SBUs, but some components (e.g., the video CD card) were directly sourced locally in Hong Kong. By the time the Aspire had been launched in more than thirty countries, local adaptations had created over 100 different configurations. As one marketing manager pointed out, “Sometimes even a product of the same model number would have a completely different configuration in another market. For example, what sold in Singapore under one model number was completely different from the same model number sold across the bridge in Malaysia.” While local designs differed, most of the configuration problems—the CD-ROM tray jam, for example—remained the same from region to region. As local sales companies with newly created Uniload lines began assembling components and subassemblies, they experienced many of the same problems AAC had confronted. When the RBU contacted Taiwan, the SBUs often were unaware of the problem. Typically, they explained that they had simply supplied components and that the integration expertise was in AAC. But because AAC’s earlier experience had not been documented or disseminated, each unit had to find its own solutions.

Marketing Coordination: A Global Brand?
Similar coordination difficulties hounded the plans to make Aspire the company’s first global brand. Acting in his informal global champion role, in September 1995, Culver invited marketing teams form RBUs worldwide to the United States to review their marketing program. “They were all very interested,” said Culver, “but when they got back home they pretty much developed their own programs. As a result, Aspire had a very different look and feel from country to country.” The different look began with the cost-driven decision to sell only the emerald green version outside the United States. The local differentiation of product features—Hong Kong’s decision to add a video CD card, for example—further emphasized the divergence. But it was in product positioning that the marketing differences were most evident. While Australia largely followed the U.S. positioning of an innovative multimedia PC for home, Singapore positioned it as a “fashion” product emphasizing its color and design, and Taiwan chose to sell it in a stripped down version (e.g., no voice recognition or user interface) as a basic entry-level computer. Pricing strategies also varied widely. In contrast to AAC’s mid-range, value-pricing, Taiwan offered its stripped-down version at a low end price. Meanwhile, the Europeans began by listing

7

399-011

Acer America: Development of the Aspire

Aspire at the top end of the market, then lowered price to build volume. And in Asia, ACI listed Aspire at a $200 premium to mainstream home PCs. ACI’s premium pricing strategy allowed companies in Asia to support a heavy advertising campaign, albeit one very different from Aspire’s U.S. approach. Culver explained, “ACI worked with an advertising agency in New Zealand and came up with some pretty interesting ads they ran in Asia. I say ‘interesting’ because they had a slightly sexual overtones that I don’t think would have worked in the States.” In contrast, Acer Europe invested comparatively little in advertising, arguing that their PC retail channels (e.g, Dixons in the United Kingdom and Carrefour in France) were still fairly small in relation to Acer’s other markets. Promotions were also very localized. In Hong Kong, for example, Mak and his team developed a special promotion and marketing program for their adapted model. In an unusual move, they teamed with Chase Manhattan Bank to allow the computer to be purchased on a credit card with payments made over 18 months interest free. With the price set around $2,000, the Aspire was finally launched in the fourth quarter of 1996 with locally produced advertisements featuring a popular Hong Kong singer. The credit card-linked program proved to be one of the most successful PC sales campaigns in Hong Kong. Within the first six weeks, the company received 20,000 orders, making it the number one PC product in a market of 300,000 units per year. The success of these various approaches was mixed. While Hong Kong, Singapore and Australia had considerable success, the Aspire struggled in Europe. Furthermore, the diverse design and marketing approaches created problems for Acer’s first globally advertised product. Following his commitment to fund a global branding campaign, Shih allocated significant corporate funds to advertising in international periodicals and in-flight magazines, a first for Acer. But because local subsidiaries had modified the design and product positioning, the global ads were not very effective. For example, a Taiwanese businessman might see an advertisement for a sophisticated prestige product in an international airline’s in-flight magazine only to find that the Aspire available in his home country was a stripped down version that did not have the features he sought.

RBU/SBU Negotiations: Growing Tensions
Local independence also presented problems during the internal price negotiation process. After developing its market-tailored configurations, each RBU headquarters generally took tender bids for component parts in their local markets, using that data to negotiate prices with the SBUs in Taiwan. According to client-server organization policies, RBUs were not required to purchase components from an SBU if its quote was more expensive, and SBUs were not required to sell to RBUs if they could obtain better local distribution by selling directly to other customers in that market. In practice, however, the threat of either the RBU or SBU sourcing or selling outside of the company—the “nuclear option” in Acer parlance—was usually enough to keep the two units in line. Nonetheless, the principle that “every manager was lord of his castle” could lead to unending negotiations. Instead of exercising the nuclear option, price negotiations often ended in stalemates: rather responding to a renegotiation request, the business units would simply sit on the request for weeks. As Simon Lin, CEO of IPG, pointed out, “In a time-sensitive industry with rapid price fluctuations you can’t afford to do this; it delays production and reduces profit margins.” Pricing was not the only area in which negotiations became bogged down. Similar problems surfaced in inventory control, a problem area that had begun with Aspire’s unique components and colored subassembly and had ballooned with the proliferation of models and component sources. The RBUs’ unique local configurations made it difficult to re-distribute excess inventory of products or parts to markets where demand was outstripping supply. As a result, there was constant tension between RBUs and SBUs regarding the need for new models that involved only minor changes from an existing model.
8

Acer America: Development of the Aspire

399-011

Next Generations’ Adjustments
From the September 1995 launch on, competition in the United States was intense since other well-known brand companies like Hewlett Packard had also introduced their own multimedia home PCs. At AAC, design teams continued to work on the various problems—component integration, monitor problems, color matching—that continued to plague the manufacturing teams. With the release of each new generation, the sales organization’s optimism surged, only to be deflated when it recognized that the product’s quality problems continued.

New Product Generations
The design for the second generation, released in February 1996, was reviewed by SBU managers who gave their input, but Frog Design didn’t want to compromise the styling to make the product more manufacturable. “After the first generation,” remarked one manager, “Mike [Culver] recognized that AAC had limited experience in development and manufacturing, and tried to share more decision making and coordination with the SBUs. Unfortunately, we were not very successful.” As a result, technical problems such as the CD tray fit continued. This led to key managers in Taiwan and the United States deciding that direct coordination between AAC and the SBUs had to become an immediate priority. They named Arthur Pai, a senior engineering manager from Information Products Group (IPG was the key supplying SBU in Taiwan) the project coordinator with responsibility to gain internal agreement and implementation of their urgent changes. The third generation, launched in August 1996, sought to resolve continuing technical problems and to capture scale economies by increasing the number of common parts in the product. Under Pai’s leadership, the CD-ROM tray problem was fixed, the ambient noise of the machine was dampened, and a quick start-up feature was added. Simultaneously, Culver changed the marketing strategy and segmented the product into three different categories: business, family, and gameenthusiasts. Each was sold through its own channel—the business version through office superstores, the family version through mass merchants, and the gaming-enthusiast version through consumer electronics superstores. Expansion and reengineering customer service eventually reduced the 20minute average wait time to less than two minutes. The customer satisfaction index doubled, but service costs rose to 8% of sales. Then, in October 1996, Compaq again cut its PC prices. Culver recalled: They narrowed the price gap between top brand PCs and the secondary brands to about 5%. Basically they priced right on top of us. We didn’t create enough perceived value to get a premium like Compaq, but our cost structure for our additional features was too high to match Packard Bell [in the lower tier]. So it took us about 12 months where we lost market share and money, until we got our fifth generation Aspire to market. It was completely redesigned for quality and cost. For the fifth generation Arthur Pai asked Frog Design to design a new housing permitting Aspire to be made entirely of standard rather than custom parts. To reduce inventory problems and respond to dealers’ requests, he replaced Aspire’s emerald green version with the industry standard off-white that retailers could more easily bundle with other peripherals. In August 1997, an entirely revitalized Aspire product line was introduced, including two models priced below the important new $1,000 price point. Simultaneously, the company announced new distribution partnership with Sears and WalMart to capture consumers at the low-level entry point and provide them with an ability to upgrade easily within Aspire’s wider product line.

9

399-011

Acer America: Development of the Aspire

Shifting Economics, Evolving Controls
As the substantial losses recorded on Aspire’s 1996 sales continued into 1997, Culver and his team came under growing pressure. (See Exhibit 8.) Culver explained the problem: The industry rule of thumb is that PC prices come down 20% a year, but in reality they go through more dramatic, less orderly drops driven by price wars. We were not getting our costs down fast enough so our gross margins dropped from more than 20% in 1995 to about 15% in 1996 and headed for 12% by 1997. That’s not much in a consumer business. On top of continuing SG&A expenses of 2-3% and our ongoing advertising and marketing cost of 1% of sales, customer service costs had risen from 2% of sales to 8% by 1997, and product returns were taking another 5% of our margin. Worst of all, retail inventory write-offs were running at 10% of sales in 1996 and 1997. There was no way we could be profitable. The problem was in part created by accounting practices that made no provisions for future service costs, returns and inventory write-offs but instead charged them against same period sales. With a three-year warranty, this resulted in all the first and second generations’ problems pulling down subsequent years’ profitability and disguising real current performance. Still, Stan Shih was concerned enough about AAC’s deteriorating profit situation that he initiated a series of quarterly meetings between himself, AAC’s Chwang and Simon Lin, head of IPG, the SBU supplying the Aspire. In July 1997, they dispatched Arthur Pai and another engineer from IPG in Taiwan to the United States to coordinate Aspire sourcing and logistics in AAC, both of which had become expensive problems. Pai commented, “AAC didn’t understand Taiwan operations very well so we had to send over some individuals who could keep in contact with Taiwan and help with the communication with the component sources.” IPG engineers were now tightly integrated into the design process and again felt responsible for the success of the product rather than simply for the supply of component parts. As AAC’s losses continued through 1997, more support was sent. A three person advisory committee of experienced Taiwan-based managers was assigned to work U.S. top management to develop new strategies and tighten operations. One member was assigned to work on AAC’s customer service programs, a second was to concentrate on manufacturing improvements, while the third was to team up with Culver to strengthen the U.S. consumer business. As Pai commented, “Acer America learned that it couldn’t work without the SBUs, and we learned that Acer America was better off not being so totally independent. It was like bringing a runaway teenager back home.” By late 1997, inventories were being cleaned up, customer service reengineered, and logistics streamlined. Aspire’s sales volume had returned to the levels reached in the second half of 1995, and its monthly losses had shrunk from $6 million to $2 million. Culver, now promoted to vice president of AAC’s consumer products division, seemed relieved. “Now we have a second chance,” he said.

Rethinking the Model
While pleased with the turnaround following the mid-year shakeup, Stan Shih was less sanguine about Aspire’s long-term prospects. Over the previous two and a half years, the oncepromising new product had generated losses of almost $100 million in AAC while its share of the home PC market had fallen from almost 14% at its introduction to less than 5% by the end of 1997. (See Exhibit 9 for U.S. market share figures.) Primarily due to Aspire’s problems, AAC was expected to report losses of $80 million on sales of about $1 billion in 1997.

10

Acer America: Development of the Aspire

399-011

Once again, many were urging Shih to give up on his goal of building a global brand, arguing that the company could make much better profits on its component business selling CD-ROM drives, monitors, etc., and on OEM contracts to supply PCs to major players like IBM. (In 1997, the company manufactured 6.2 million computers. It was the number three PC maker in the world, with about half its output being branded and the other half sold under OEM contracts.) But the determined CEO remained firm in his commitment to elevate Acer’s own branded PCs from their 1997 ranking as number eight to one of the top five brands in the world. At the same time, however, he felt there were some important lessons to be learned from the Aspire experience and he began to think through some key strategic and organizational issues.

Strategic Questions
Although the development of the Aspire was exciting because it had emerged in response to opportunities detected in a leading market, its global implementation had been hampered by the subsequent need to adapt the product to different market requirements around the world. Always trying to think ahead, Stan Shih began to look for opportunities not yet defined by the market but with global potential. Shih began to describe a vision for low-cost, limited task devices—he used the term “machines” or “appliances” rather than “computers”—that Acer would sell as “XCs.” (“X” represented the individual applications these appliances could support; “C” stood for computer.) Costing less than $200, these devices would be designed for focused tasks such as playing computer games, supporting home-use applications such as banking or stock tracking, sending and receiving email, or surfing the Internet. Shih believed his new concept had an annual potential of a billion units by 2010 (ten times the current level of PC sales), expanding computer use in less affluent markets worldwide. He hoped Acer would supply 10%. Should this become the company’s primary thrust in the consumer market? What continuing role, if any, should Aspire have in Acer’s product line? And, most important, how should these issues be decided?

Organizational Options
While AAC had provided a textbook example of how an RBU could become a valuable source of entrepreneurial initiative, it had done so at considerable cost to Acer’s traditionally efficient Taiwan-based operations. Was it time to rethink the networked organization model Shih had rolled out as his innovative “client server” concept? If so, how should he redefined the roles, responsibilities and operating relationships of Acer’s SBUs and RBUs? What implications did any such changes have for the ”Global Brand, Local Touch” philosophy of locally tailored products? And how could any proposal for greater centralization be implemented through the numerous independent companies spun out as independently listed public companies as part of the “21 in 21” program? (See Exhibits 10a and 10b for subsidiary ownership information.) As Shih reflected: “In the mid- and late-1980s, we experienced the difficulties that come with heavily centralized systems and we don’t want this to occur again. We have always striven to empower our employees at every level, but it is difficult to strike a balance between our principles in this arena and the need for greater cost efficiency.”

11

399-011 -12-

Exhibit 1

Acer Selected Financials, 1993-1997
1993

a

December 16, 1998For the Year Combined Total Revenue Revenue Growth (%) Net Earnings Net Earnings (%) Total Equity Return on Equity Total Assets Return on Assets Net investment in property, plant, and equipment Working Capital Number of Stockholders Number of Employees 1,833 49.4% 86 4.6% 497 18.5% 1,584 5.7% 497 173 70,000 7,200

1994 Excluding a TI-Acer 2,901 75.7% 103 3.6% 420 28.1% 1,520 7.8% 197 280 69,000 8,612

1995 Excluding a TI-Acer 5,262 81.4% 163 3.1% 939 23.9% 2,340 8.4% 284 758 89,000 13,942

1996 Excluding a TI-Acer 5,346 1.6% 150 2.8% 1,321 13.3% 3,156 5.5% 418 995 122,000 15,272

1997 Excluding a TI-Acer 6,132 14.7% 262 4.3% 1,638 17.7% 3,608 7.7% 616 974 154,000 21,307

Excluding a TI-Acer 1,651 38.4% 22 1.3% 316 7.0% 1,143 2.0% 181 149 44,000 6,348

Combined 3,220 71.0% 205 6.4% 703 34.2% 2,082 11.2% 538 288 70,000 9,700

Combined 5,825 80.9% 413 7.1% 1,450 38.4% 3,645 14.4% 963 767 90,000 15,352

Combined 5,893 1.2% 188 3.2% 2,008 10.9% 4,192 4.8% 1,347 996 123,000 16,778

Combined 6,509 10.5% 89 1.4% 2,065 4.4% 4,758 2.0% 1,470 875 155,000 22,948

a

Due to the drastic drop in the market price of DRAM during 1996-97, the Acer Group reported its financial results excluding TI-Acer operations to allow evaluations of non-DRAM Acer Group operations.

399-011 -13-

Exhibit 2a

Acer, Inc. Organization Structure, c. 1997

Acer America (AAC)

Acer Europe

Acer Sertek* & Acer Marketing Services (China & Taiwan)

Acer-Computec Latin America (ACLA)*

Acer Computer Int’l (ACI)* (Rest of World)

Acer Group

Fora Inc. Acer Property Devpt.

Info. Products Group* Servers, Desktops, Notebooks, Motherboards, Multimedia, Video Conferencing SIMM

Acer Peripherals* Monitors, Keyboards, Scanners, CD-ROM, Mobile Phones

TI-Acer DRAMs Acer Laboratories ASICs Ambit Hybrid Designs Acer Weblink

Acer Netxus Networking & Communications Products Acer Internet Svcs. Internet Services Acer Third Wave Publ. Periodicals & Software

Acer NeWeb Antenna, Wireless Communications Advance Info. Motherboard CD-ROM Backbone Systems Add-on Cards

Names in plain text are RBUs, names in bold are SBUs, and names in italics are classified as “Other.” * Indicates a publicly listed company

399-011 -14-

Exhibit 2b

Acer, Inc. Regional Locations, c. 1997

San Jose: Acer America

Mexico City: Acer-Computec Latin America Subsidiaries: Argentina Chile Colombia Mexico Newtec (Mexico) Peru Venezuela

Amsterdam: Acer Europe Subsidiaries: Belgium Holland Finland France Germany Austria Norway Poland Hungary Spain Italy Sweden Denmark United Kingdom

Beijing: Acer Market Services, Ltd.

Taiwan: Acer Group Information Products Group Acer Peripherals Inc. TI-Acer Acer Laboratories Fora Inc Acer Property Development Acer Sertek

Singapore: Acer Computers Int’l (ACI) Subsidiaries: Hong Kong Middle East, UAE Turkey Japan Korea Malaysia Thailand India Australia New Zealand CIS Acer Africa (South Africa)

Acer America: Development of the Aspire

399-011

Exhibit 3

Stan Shih's PC Industry Conceptualization

Source: Company document

15

399-011

-16-

Exhibit 4a

1997 Acer Brand Ranking and Market Share by Product

Product

Worldwide Market Rank Share 8 8 6 6 3.6% 3.2% 5.1% 3.0%

United States Market Rank Share 9 8 5 7 3.3% 3.0% 5.6% 2.8%

Latin America Market Rank Share 3 3 2 4 9.1% 8.6% 21.5% 6.8%

Asia/Pacific Market Rank Share 5 6 3 4 5.8% 5.3% 9.6% 8.9%

Europe Market Rank Share 10 10 4 7 2.8% 2.1% 7.8% 3.1%

Overall Desktop PCs Portables Servers

Source:

Company documents

Exhibit 4b

Acer Brand Ranking by Country, 1997

Rank #1

Country Taiwan, Malaysia, Indonesia, Philippines, Bangladesh, Mexico, Chile, Panama, Uruguay, South Africa, Oman, Morocco. Bolivia, Venezuela, Colombia, Peru, Ecuador, Brazil, Singapore, Thailand, Bahrain, Jordan, Syria, Cyprus, Sri Lanka, United Arab Emirates, Tunisia Hong Kong, Israel, Turkey, Greece, Norway, Finland, Saudi Arabia Australia (6), Holland (7), Italy (7), Germany (7), Austria (9), USA (9), China (9), France (10)
Company documents

Top 3

Top 5 Top 10
Source:

Acer America: Development of the Aspire

399-011

Exhibit 5

AAC Simplified Financials: 1990-1997
1990 161 133 27 20 5 (24) (1) (25) 1 (26)

a

(Million US$) Revenue Cost of Sales Selling and Marketing General Administration Research and Development Operating Profit/(Loss) Non-operating Profit/(Loss) Profit/(Loss) Before Tax Tax Net Income/(Loss)

1991 235 190 61 16 8 (40) (7) (47) (2) (45)

1992 304 283 25 17 6 (26) (3) (29) 0 (29)

1993 434 399 23 19 4 (11) (5) (16) 0 (16)

1994 858 764 55 20 4 15 (3) 12 1 11

1995 1437 1303 103 22 4 6 (4) 2 1 1

1996 1268 1225 84 26 6 (74) (7) (81) 0 (81)

1997 1141 1125 72 29 4 (89) 3 (86) 0 (86)

Current Assets Fixed Assets (net) Other Assets (net) TOTAL Assets

155 39 37 231

153 43 37 233

123 28 31 182

144 25 19 188

242 25 11 278

449 26 9 484

236 32 9 276

304 33 8 345

Current Liabilities Long-term debt Stockholder Equity (including additional capital) Total liabilities

155 17 58

169 15 50

154 18 10

136 58 (6)

218 47 12

423 10 51

243 14 19

365 10 (30)

231

233

182

188

278

484

276

345

a

Totals may not add due to rounding

17

399-011

-18-

Exhibit 6

Aspire Design and Characteristics

Acer Aspire Key Features

Source: Company documents

Acer America: Development of the Aspire

399-011

Exhibit 7 x AAC’s Forecast of Next Month’s Sales into Channels vs. Actual Sales into Channels

1995

1996

Month

Forecast (000 units)

Actual

Forecast (000 units) 15 50 35 25 25 30 25

Actual

January February March April May June July August September October November December 30 50 80 100 40 10 55 70 70 15

0 20 50 5 15 25 20 20 45 20 15 5

50 65 80 80 55

Source: Company documents

Exhibit 8

Aspire Profit Performance in AAC: Actual vs. Budget, 1995-1997

Millions US$

1995 (part year) Budget Actual 446 (2) Budget 1015 9

1996 Actual 454 (51) Budget 540 (19)

1997 Actual 268 (38)

Sales Profit

570 17

Source: Company documents

19

399-011

Acer America: Development of the Aspire

Exhibit 9

Acer Market Share

Acer's Home-PC
15%

10% Percent

U.S. Retail Market Share

5%

0% IV'95
Source:

I'96

II

III

IV

I'97

II

III

IV Est.

“A New Attack Plan for Acer America,” Business Week, December 8, 1997, p. 82.

Exhibit 10a
Subsidiary

Acer, Inc. % Stake in Acer Subsidiaries, 1996
% Stake 100.0 100.0 65.7 63.4 50.0 48.8 40.9 36.2 100.0 23.4

Acer America (AAC) Acer Europe Acer Laboratories Acer Computer International AMBIT Microsystems TI-Acer Acer Peripherals Acer Sertek Information Products Group (IPG) Acer Computec Latino America (ACLA)

Source: Morgan Stanley Analyst Report, 1996

Exhibit 10b
NT$ millions

Estimated Subsidiary Contributions to Acer, 1996-1997
1996 229 109 525 -1,365 406 22 12 -175 1997 -217 134 593 41 442 20 16 17

TI-Acer Acer Sertek Acer Peripherals Acer America Acer Computer International Acer Laboratories AMBIT Acer Europe

Source:

Morgan Stanley Analyst Report, 1996

20