Weathering the Tech Storm

How Michael Marks boosted efficiency at contract manufacturer Flextronics

During the tech boom, a class of manufacturers came to embody a nuts-and-bolts version of the Internet dream. By buying up thousands of underused factories around the world from the likes of Hewlett-Packard, Siemens, and IBM, these electronics- manufacturing services (EMS) providers brought ultralean efficiencies to the high-tech sector in the U.S. and Europe, where manufacturing operations had grown uncompetitive.

Yet proficient as they were at making digital gear, the EMS players' stellar expansion has been halted by the tech crash. After growing over 30% a year through the late 1990s, to $106 billion in 2000, the EMS market contracted by 13% to $92 billion in 2002, according to Technology Forecasters Inc. of Alameda, Calif. Sales shrank for all but one of the top four EMS players. The exception was Singapore's Flextronics International, which amassed revenues estimated at $13.1 billion in the fiscal year ending in April, 2003, producing everything from Nortel routers to Ericsson cell phones and Microsoft Xbox game consoles.

Today, even for such top-brand products, demand is weak. That has forced Flextronics' once-bullish CEO, Michael E. Marks, to tone down his optimism. Where he used to speak of racing to $50 billion in revenues in just a few more years, Marks now admits that the days of rapid-fire outsourcing megadeals are over. He does cite one encouraging development: Flextronics has used the lull in tech manufacturing to cut costs to about as low as they can go, leaving the outfit ready when the rebound comes. In an interview with Senior International Editor Pete Engardio, Marks explained his cautious outlook and the implications of SARS for electronics outsourcing.

Flextronics seems to have survived the tech crash better than its top rivals. What did you do right?

We were lucky because of all the companies in the EMS industry, we've been the most oriented toward consumer products, which have been relatively okay. Consumer products are half of our sales, compared with 30% in 2000.

It has been three years since the NASDAQ bust. Is there any visibility on when tech will recover?

There really are no end markets that are in good shape right now -- not telecom, not computers. Consumer electronics is pretty good, but not great.

In a shrinking market, how are you doing?

We're fine -- becoming profitable -- but we're in a kind of business that's not doing great.

You have a lot of production in China. How serious a threat does the SARS outbreak pose to electronics supplies?

If you are prepared, SARS won't bring work to its knees. In normal times, it's natural for supply chains to be disrupted. SARS is no different. There are port strikes, yield problems at semiconductor plants, labor shortages, and currency-exchange problems.

Will it slow the migration of factories to China?

The outbreak exposes the risk of companies putting all of their production in China. Some companies are moving stuff to China that really doesn't belong there. It makes sense to make cell phones in China because they are inexpensive to air-freight. But personal computers don't travel well. If you start air-freighting PCs because of a supply disruption, your cost-savings disappear instantly.

Does this mean some China operations are likely to be shifted elsewhere?

In recent weeks, several customers have told us they need a backup plan. We say fine -- go to Guadalajara. For some products, we could have production running in Mexico in two or three weeks. That's because all of the manufacturing information is in our IT system. For now, they're staying in China. But if [the SARS crisis] lasts for months, people will start to rethink.

Back to your performance. How have you cut expenses?

We cut overhead positions, such as in finance and human resources. We closed a lot of factories that we picked up in the days when we needed all the capacity we could get. We closed around 30 factories out of 150. And we cut our selling, general, and administration expenses [SG&A, an accounting measure of overhead] to about 3% of our sales, a record for us. During the bubble, SG&A was 4% to 4.5%.

How much more can you trim?

We're running out of stuff to cut. Eventually, you reach the point of diminishing returns. I worry about not being able to provide good service. And at some point there will be enough demand, and we'll want our remaining facilities. Still, we can do a lot to improve supply-chain efficiency, like upgrade our IT systems and increase staff training.

What about head count?

Our employment base is actually at a record high: about 100,000. We've done this by cutting the workforce in high-cost locations and adding in low-cost places. We put a lot of consumer products in Asia, where they belong in the first place. China has been the hottest location. We have 35,000 there. For example, we moved all of the production of Microsoft's Xbox consoles from Mexico to China.

What were your big mistakes during the boom years?

Looking in the rearview mirror, there are lots of things we shouldn't have done. We bought three optical-component companies for about $200 million, and today we have almost no revenue to show for it.

What lessons did you draw from some of the bad investments?

We'll be a little more cynical. When we enter a new area, we'll make sure the demand is real. We'll do more due diligence, and we'll turn down lots of deals we may not have turned down when things were great, especially in areas we don't understand and where demand isn't proven. We are more humble.

What proved to be some of your better moves?

Design and logistics are where the value proposition is -- not in manufacturing. We'll know better in three or four years whether vertical integration [handling everything from production to warehousing and shipping] is a good strategy. But I've staked my reputation on it. For example, in logistics, we bought Irish Cargo Express [in 2000] for $50 million, and it's now a billion-dollar business for us. A lot of the business is what you would think of as warehousing. We take orders from companies such as Hewlett-Packard or Xerox and ship their products directly to customers.

We also now have a lot of integrated circuit-design work. This has enabled us to develop a lot of products that we sell under other brand names. The only product we've admitted to is CDMA and GSM cell phones. In the next few years, we will be much more public about this business.

How will you grow in the future?

We won't do it with acquisition. One way to grow is to just take business from competitors, which is hard. The better way is to get more companies that manufacture internally to outsource -- and that will continue.

What's left to outsource?

Even in electronics, there's a lot left. The Japanese still make $250 billion in electronics but have outsourced relatively little. We also are starting to do business with Chinese and Indian companies. There is a slow-developing opportunity in automotive components, but it's a big one. Lear and Delphi -- those guys produce almost everything in-house and are increasingly looking to outsource.

Some would say companies went overboard outsourcing. Dell is doing well doing it all. Is a rethink under way?

If a company invests in manufacturing and is great at it, like Dell and Nokia, that's fine. These companies are very large, very good, very global. But there aren't many [like that]. The world has changed for new entrants. Ten years ago, you didn't have to be a behemoth to have good purchasing power for parts and materials. Now, you need $25 billion in revenue to match the purchasing power of a Dell. You also need flexibility with your production. Say you build a fabulous factory in China, but then it gets shut down by SARS, so you need to go to Mexico. If you aren't big enough, you can't play this game.

    Before it's here, it's on the Bloomberg Terminal.