The ups and downs of contract manufacturing

Every industry has its ups and downs. Flextronics strategically managed its growth not only during upturn but even during the downturn to edge out Solectron from no.1 position. The article shows how. With the industry on the upswing again after the most serious downturn in its history, Flextronics continues to strategically manage its growth.

Solectron Corp. was riding high in 1999, the year it became the largest contract manufacturer (CM) in the world. All CMs grew during the 1990s but none faster than Solectron, whose revenues rose from $300 million in 1989 to $8.4 billion a decade later. In December 1999, EB named Koichi Nishimura, who had engineered Milpitas, CA-based Solectron's growth through a flurry of acquisitions, CEO of the year. A few months later, Solectron topped our annual list of CMs for the first time.

Further down on that same 2000 list, at No. 6 (with rankings based on previous calendar-year revenue), was Flextronics International Ltd, with $1.8 billion of revenue. With corporate offices in Singapore and San Jose, Flextronics survived severe difficulties in the early 1990s. It has been led by CEO Michael Marks since 1993, when revenue was a mere $100 million. It climbed to the No. 2 spot in 2001, and in 2003, moved to the top of the heap, with revenue of about $13.4 billion, up from $13.1 billion in 2002. Solectron fell, with revenue of $11.7 billion in 2003, a plunge from $12.3 billion in 2002 (Table 1).



What happened? Solectron pioneered the electronics manu-facturing services (EMS) business (essentially what CM has evolved into), setting standards that everyone followed and winning two Malcolm Baldrige National Quality Awards for manufacturing. How did Flextronics manage to overtake it? Manage is the key word. Consultants, equity analysts, market researchers and executives at the two companies concur: Flextronics outmanaged Solectron before and during the downturn.

"Flextronics stumbled and lost money but not momentum during the downturn," says Randall Sherman, principal of New Venture Research, Nevada City, CA. "Solectron lost momentum and is still trying to stem the bleeding."

Before the downturn, both companies aggressively pursued acquisitions. They also won accounts with big OEMs (sometimes stealing them from each other), diversified geographically and expanded their services. In the downturn, they both undertook severe restructuring, but Flextronics moved faster to cut capacity, reduce workforce and move manufacturing to lower-cost sites. Both depended too much on telecom, but Flextronics branched out into consumer electronics.

Solectron mishandled its new assets, lost sight of customer service and became a bloated bureaucracy in which decisions passed through many layers of management. Flextronics integrated acquisitions smoothly; steadily improved customer service; and kept a lean, flat structure in which the executives closest to the customers made the key decisions.

What Flextronics did well and Solectron did poorly offer lessons for the entire industry, says Pamela Gordon, president of Technology Forecasters Inc., an EMS consulting firm in Alameda, CA. She says CMs must manage acquisitions well, balance their product portfolios, pay attention to customers and create cultures that can act decisively. "The message is: Pay attention to the fundamentals."

Acquisition troubles
All large CMs grew by acquiring competitors, suppliers and OEM factories, but Solectron and Flextronics were the most aggressive. Flextronics tended to buy companies that were culturally compatible and geographically diverse and that broadened its competencies for vertical integration. Solectron tended to acquire raw capacity and revenue. From 1997 through 2002, according to Sherman, the companies made a combined 75 acquisitions. Twenty-seven of Flextronics' 43 acquisitions were other CMs or OEM plants. The rest were a mix of logistics providers, materials suppliers, software developers and design firms. Twenty-four of Solectron's 32 acquisitions were OEMs or CMs. Of the other eight, six were in repair, including a customer call center company. "Solectron bought revenue," says Kevin Kane, an analyst at International Data Corp., Framingham, MA. "Flextronics bought services and capabilities. It was able to leverage the services and capabilities during the downturn, because there was a demand for them. Solectron ended up with a lot of excess capacity."
Solectron's big mistake was to acquire two large competitors in the same year the downturn began. In early 2001, it bought NatSteel Electronics Ltd., Singapore, for $2.4 billion, and later that year, acquired C-MAC Electronic Systems Inc., Montreal, in a $2.7-billion stock swap. NatSteel had more than 2.3 million square feet of factory space at 11 sites, mostly in Asia, and 12,000 employees. C-MAC had 52 manufacturing facilities around the globe and 9,000 employees (its comparative square-footage statistics were not available). Both were among the top 10 CMs, based on revenue, when they were acquired.

In 2001, Solectron also took the unusual step of acquiring a call center company, Stream International, Canton, MA. The terms were not disclosed, but Stream didn't affect Solectron finances like the two CM acquisitions. It did add 10,000 employees in 22 call centers and a business model foreign to Solectron. The acquisition's goal was to service a budding repair business.

Mergers such as Solectron's with NatSteel and C-MAC or that of Sanmina Corp., San Jose, with SCI Systems Inc., Huntsville, AL, to form Sanmina-SCI Corp., San Jose (No. 3 in 2003), create huge challenges, says Bob McNamara, managing director of Broadview International LLC, Fort Lee, NJ, a mergers and acquisitions investment bank. None of these mergers have gone smoothly, he says. "Doing large acquisitions requires a carefully thought-out integration strategy and flawless execution. The bigger the acquisitions, the bigger the distractions."

Gordon agrees. "C-MAC alone had about 35 different enterprise resource planning (ERP) systems. Just integrating C-MAC's IT into Solectron must have been a CIO's nightmare. And that's just one example of the difficulties."


Some critics are harsh. "Any culture Solectron might have had never got transferred to new companies, and Solectron did not integrate IT or business processes," says Louis Miscioscia, an equity analyst at Lehman Brothers Inc., New York. NatSteel C-MAC were "horrible acquisitions," he says. It essentially spent too much money and wound up with excess capacity.

Problems with economies of scale
Solectron's two big acquisitions were so costly that it ended up with more debt than shareholder equity. In 2003 ended May 31, its long-term debt was just under $1.9 billion and shareholder equity was just over $1.6 billion. In the fiscal quarter of 2003 ended June 30, Flextronics had long-term debt of $1.4 billion and shareholder equity of $4.3 billion (Table 2).



Acquisitions should also lead to lower selling, general and administrative (SGA) costs, McNamara says. At Solectron they did not. Solectron's SGA costs were $172 million in its 2003 May quarter, and Flextronics' were $116 million in its 2003 June quarter. McNamara says Solectron may never achieve SGA savings with the call center, because it is so different from its core business.

Flextronics' only big acquisition was DII Group Inc., Niwot, CO, for $2.4 billion in stock in 2002. DII was among the top 10 CMs, with 2.9 million square feet at 22 sites and 12,000 workers. "Flextronics' acquisitions have typically been smaller and more strategic, easier to integrate," says McNamara. "This may have cost it revenue, but it has benefited the company in overall operating performance."

CEO Marks says that Flextronics has not been immune to integration problems, especially in IT, but that it does well with culture. "We always buy a culture that is similar to ours, with an executive who thinks the way we do," he says. "Our primary criterion is that it have a similar culture. We have walked away from acquisitions when that wasn't the case."

Marks says Flextronics' strategy was to expand evenly in Asia, Europe and North America, especially to low-cost places such as China. "We were much more global than Solectron. We started in Asia, and it started in North America. We had a strategy of being a player in all three markets. Solectron was slow to recognize the other two."

In January 2003, CEO Nishimura retired from Solectron and was replaced by Michael Cannon, the former CEO of Maxtor Corp., Milpitas, CA. Cannon declined to be interviewed but his view of Solectron's problems is clear from a presentation to equity analysts last year. Cannon vowed to get "back to basics" and identified four priorities: business growth, return to sustained profitability, better use of assets and achieving execution excellence. Cannon called Solectron's past acquisition strategy part of an industry-wide "land grab mentality" and admits that all CMs paid too much for plants. He said, "I hope the industry never repeats this, because it was a burden for the industry as a whole."

Besides integration, the challenge for Cannon was to decide which assets Solectron needed and which ones it didn't, says McNamara. "The prior management team was so busy looking at the next deal that it didn't take time to look at prior deals and peel off what wasn't working."

Cannon told the equity analysts something similar-Solectron must reevaluate all acquisitions. "If we find that certain businesses and activities are, for whatever reasons, no longer core or may not be strategic, then those are businesses we are going to look to potentially divest."

Unbalanced portfolios
Solectron's acquisition problems overshadow everything, but it also misfired elsewhere. When the downturn hit, Flextronics and Solectron were too heavily dependent on networking and telecom, which accounted for half or more of revenues. Gordon says those segments were just too tempting to pass up in the late 1990s and 2000.

Here, too, Flextronics suffered less. Before the downturn, Flextronics won contracts in consumer electronics, the first large CM to do so on a large scale. Even much of its telecom business was on the consumer end-in mobile handsets-which was not hurt as much as telecom equipment. Consumer spending was the one bright spot in the downturn, and Flextronics benefited.



Contract manufacturing market share (2004E).
Source : Company information and Deutsche Bank estimates

Aside from mobile handsets, Flextronics' first consumer contract was the Xbox game console for Microsoft Corp., Redmond, WA. "Xbox was a high-profile project. The volume and revenue weren't that high, but it was the visibility that mattered," Kane says. The Xbox had a domino effect, bringing Flextronics new consumer business, including the 2002 acquisition of some plants of Casio Computer Co. Ltd., Tokyo, a consumer electronics OEM.

Analysts say Flextronics' move into consumer electronics was indicative of Marks' visionary leadership. "Flextronics got on the consumer electronics bandwagon when it was extremely unpopular, and everyone thought the company was crazy," says McNamara. "This strategy has sustained it through the downturn in telecom, computers and capital equipment."

Consumer electronics are not popular at CMs, because the margins are the lowest in a low-margin business. "When telecom was on fire, Wall Street was naysaying our efforts on the consumer side," says Marks. But the company persisted and cell phones, the Xbox and HP printers have been stable through the downturn, he says. Marks acknowledges that Flextronics has recently swung too far in the consumer direction and is trying to rectify that.

In one of his presentations, Cannon said there's great potential in consumer electronics. He also suggested that automotive, industrial, defense and medical are segments Solectron could pursue. He was banking on repairs becoming a significant business. Solectron won the repair contract for the Xbox, although that contract is much smaller than Flextronics' Xbox manufacturing deal. Analysts say that it would be a few years before the success of the repair business can be fully judged.

Customer service follies
Despite its troubles, Solectron has a lot going for it. "Its name is still good in the industry," Kane says. CM customers still consider Solectron's process technology the best, but many believe that Flextronics is better at hand-holding, Sherman says. Solectron set a high bar for quality in an industry that once notoriously lacked it. Just to compete, everyone else had to catch up and they have, says Sherman. Customer satisfaction, a more ambiguous metric, is now as important.

In its acquisition blitz, Solectron may have lost sight of customer service. "Maybe it rested on its laurels. Maybe those Baldrige awards went to its head," says McNamara. "Maybe it thought the customers needed it so badly for its quality and technology that they would never leave it." Solectron's well-publicized acquisition problems and its financial difficulties didn't help. By 2000, a lot of customers recognized that Solectron was not doing well from an operational standpoint and started to pull work from it, says Miscioscia. For example, he says, Ericsson moved much of its cell handset business to Flextronics. In one of his recent presentations, Cannon talked about improving operational execution and bringing in new leadership to help but did not directly address customer satisfaction.

One key to customer satisfaction is lower costs. Flextronics won fans when it led the industry in moving to low-cost locations, especially China. It also established industrial parks to bring together suppliers and manufacturing at one site to keep costs down. Its geographic diversification has helped it woo and keep customers.

Flextronics also strives to make every plant look the same and meet the same standards. Mark Zetter, a former Flextronics manager and now president of Venture Outsource Group, San Jose, a consulting firm, says its factories are the best organized, cleanest and most consistent of any he has seen. Zetter has been in Solectron plants and those of other CMs, because his firm helps OEMs develop and execute outsourcing strategies, including choosing CM partners. "Flextronics created an organization with a sense of urgency and accountability across all levels," he says. "That's a key difference between Flextronics and everyone else."

A question of culture
The sense of urgency seems to apply to most things at Flextronics. There's a dramatic difference between it and Solectron in culture. Flextronics is a flat organization that has proven that it can quickly adapt. Solectron became a lumbering bureaucracy in which most decisions were made or signed off by Nishimura.
"Marks' philosophy is to empower small teams of competent people to make their own decisions to respond to customers' needs," says Technology Forecasters' Gordon. "Ko Nishimura had a more hierarchical style, in which decisions were passed upward along the chain and tended to take longer as a result."

Gordon says business unit leaders at Flextronics have a high level of spending authority - to buy capital equipment, for example - so they can move quickly to meet customer needs. When Zetter worked there in the mid-1990s, he says, unit executives could spend up to $2 million on their own signature. "Marks used to say it's not the big fish that eats the small - it's the fast fish that eats the slow."

"I don't need to make all the decisions," says Marks, "and I hire people like that, too." He says Flextronics could have moved faster in the last downturn "if we knew how long it was going to last." It shuttered 40 plants and laid off 15,000 workers in that downturn and shifted much of its business to lower-cost sites. The only reason it kept manufacturing in North America, says Marks, is because some customers wanted it.
McNamara says Flextronics is more decisive than most other CMs, especially Solectron. "It was more disciplined at quickly shutting down excess capacity in North America," he says. Because it had a healthier balance sheet than Solectron, it could more easily afford the approximately $75-million write-down per major manufacturing facility, McNamara notes. "Flextronics has led the way in shrinking excess capacity. It is still the leader in percentage of capacity in a low-cost environment. The Solectrons and Sanminas are just now trying to catch up."

Miscioscia says Flextronics' mix of low-cost/high-cost sites was 50-50 at the start of the last downturn and was around 30 percent high and 70 percent low in 2003. Solectron was 70 percent high when the bust hit. Cannon has said its goal is 70 percent low-cost, 30 percent high-cost. When the downturn hit, Solectron had more excess capacity than any other CM, much of it high-cost, due in part to the acquisition of C-MAC, which had many North American sites.

There's wide agreement that one of Solectron's big problems was too many layers of management and the hierarchical decision-making it caused, problems Cannon seems to recognize. "You'll see us simplify the organizational structure, and in parallel with that, you will see us continue to take cost out of overheard while making the company more efficient," he told analysts.

Cannon admitted that customer satisfaction problems grew in part from Solectron's inability to deal quickly with customers. He commented on a customer who complained that doing business with a Solectron competitor required one meeting with one person but that with Solectron he needed four meetings with four people or one meeting with four people. "Going forward, I don't think this is the optimal model for us. We need to be easy to do business with."

Easier said than done. Two former Solectron managers, who spoke on the condition of anonymity, because they still work in the CM industry, described a company not only laden with bureaucracy - eight layers from factory worker to CEO, according to one source - but also rife with turf battles and decision gridlock. "No one understands Solectron until they get on the inside," says one of these former managers. "They think it is like any other company, and it is not."

Blame could be heaped on former CEO Nishimura, but most analysts pull their punches. "Nishimura did a great job over time and was a great pioneer of the industry," says Kane. "But his time had come and gone." Cannon expressed similar feelings to the analysts: "I'm not spending my time worrying about what happened in the past."

The question for Cannon is whether he can turn the ship around. Solectron still has some restructuring to do, cleaning up the mess, whereas Flextronics can concentrate on acquiring new business and taking advantage of its competitors, McNamara says.

Electronic Business, a sister publication of EM Asia (published in EB in September 2003).


Figure 1 : 2003 to 2008 electronic products EMS forecast ($mn)
Source : Gartner


Figure 2 : 2003 to 2008 electronic products ODM forecast ($mn)
Source : Gartner


Flextronics' recent acquisitions
Flextronics has purchased India's Hughes Software Systems (HSS) - a provider of convergent software solutions for fixed and mobile networks -from Hughes Network Systems for $226 million. Reporting a healthy 63 percent revenue growth in fiscal 2004, about $80 million, a 25 percent revenue growth is anticipated next year for HSS. With this acquisition, Flextronics claims it is now the first EMS provider to offer embedded and application software development of telecom infrastructure products.

Flextronics has also concluded a deal to take over Nortel Network's manufacturing, repair and optical design operations for between $675 and $725 million. This acquisition is unlikely to adversely affect Nortel's net earnings, which last year hovered around $732 million. Instead, Flextronics expects the deal to be accretive by approximately 10 cents per diluted share in fiscal 2006 and 15 cents per share in fiscal 2007.



These recent acquisitions enable Flextronics to boast robust hardware and software telecom and datacom design capabilities in an industry that sees stiff competition. According to a company spokesman, the convergence of capabilities through acquisitions blends well with Flextronics long-term strategy and bolsters its resilience to future economic uncertainties.

Based on a Gartner report, the Asia-Pacific EMS production growth from 2003 to 2008 will be at CAGR of 12.6 percent, with Asia-Pacific taking the no.1 production base from 2006 (Figure 1). EMS providers expect to see slower growth due to stronger competition from ODM especially in Asia-Pacific (Figure 2); and the number of EMS companies is expected to reduce through mergers and acquisitions.
Flextronics' end-to-end solutions business model (Figure 3) extends from OEM/ODM product and design, to OEM sales and marketing. Differentiating itself from the traditional EMS model that usually starts from printed circuits board assembly (PCBA) and ends at final testing, Flextronics expects to remain strong at tier-one this year.

by Tay Hui Leng



"EMS is becoming redundant; supply chain is replacing it"

Peter Tan, executive vice president and managing director, Asia, Flextronics, calls for eliminating the concept of EMS altogether in view of the widely changing and expanding role of the so-called EMS companies.

Excerpts:

EM Asia: How do you look at Singapore as a production hub, vis-ˆ-vis the emergence of China and India as global manufacturing and service hubs?

Tan: The unique advantages of Singapore are stability and predictability, and no surprises, unlike China and India, where there are lots of surprises. Everything works in Singapore. Our customers who are in the high value product range prefer Singapore for high-speed and high-value commercial products, as distinguished from consumer products. In China high-skilled technical and managerial manpower is not easily available. When such expertise is available, local expertise may cost even more than expatriates. Power issue is another problem. Inflation is yet another. The government of China is trying to slow down growth. India is ideal for software development and support, but is way behind in manufacturing. We make the best out of the situation by making acquisitions abroad, while continuing to strengthen our base. We have recently acquired Hughes Software in India, and have more acquisitions in mind, which we can't reveal at this stage.

EM Asia: With Flextronics involved in design in a big way, does it prefer to be called EMS or ODM?

Tan: I think the concept of EMS is becoming redundant, and supply chain is replacing it. Flextronics is a supply chain, and provides full range of services from design and manufacturing to logistics. The life cycle of a product begins with the product concept. Our services begin with the concept, and we take the product through design, software design, IP design, tooling, mechanical design, bill of materials, sourcing, manufacturing, and so on. One more concept that needs to be examined is CDM - Co-design and Manufacture. Many companies are asking us to partner in design, but they want to own the IP. In ODM approach we own the IP. In CDM approach they own the IP. We have both ODM and CDM capabilities, while essentially being a supply chain.

EM Asia: Would you also be an OBM -Original Brand Manufacturer? Some EMS companies have started selling products using their own brand names.

Tan: Flextronics will never do so. We shall take a product from concept through design through logistics, but never use our own brand name.

EM Asia: How far is it true that as a design hub Singapore is falling behind Taiwan, Korea, India and China?

Tan: Design houses in Singapore exclusively devoted to design may have some trouble. But design as a part of the supply chain is thriving in Singapore and will continue to grow.

EM Asia: What is the projected revenue turnover of Flextronics from Asia?

Tan: In 2004, 50 percent of our revenue will come from Asia, and in coming years this percentage will grow.

by Kirtimaya Varma