Statistics don’t lie, but conventional thinking might lead to the wrong conclusions about customer ratings. You tell us.
At its Quarterly Forum in San Jose, California, last week, Technology Forecasters reported the results of the 2007 edition of its “EMS, ODM and Distributor Report Card: Customer Satisfaction Survey.” On five criteria OEMs gave their EMS and ODM partners mostly boring grades – a lot of C’s and C+’s, few A’s and no F’s. The results are consistent with OEM ratings from previous TFI studies.
And, perhaps adding insult to injury, most OEMs did not consider their primary EMS or ODM partner to be “best in class,” except on one of the criteria graded —responsiveness to change.
What shall we make of this?
The results caused a rousing discussion among TFI members at the Forum. One conclusion: OEMs are loath to give their supply chain partners anything higher than average marks because they want to keep the pressure on them to deliver better, faster, and cheaper. Some OEMs represented at the Forum agreed this could be the case.
The grades also prompted an interesting reaction from the EMS folks at the Forum: If an EMS company is getting all A’s, maybe it ought to be paid more by the customer. Think about it. If the EMS is getting C’s, then maybe the effort and expense it is expending is just right. Forget about “delighting the customer” when margins are razor thin. There’s some kind of weird logic to that.
We’re interesting in your comments. Be sure to identify whether you’re from the OEM or EMS side of the equation. Are C’s and C+’s the sweet spot from everyone’s perspective?
For the first time since I started measuring more than five years ago, and most likely for the first time ever, the labor cost for assembling printed circuit boards (PCBs) is lower in the U.S. than in Canada.
No, you did not misread that statement. By my recent calculations, the average U.S. cost for fully burdened direct labor for PCB assembly is $39.10 per hour. The comparable figure in Canada is $45.80 - 17 percent higher.
The weakened U.S. dollar has turned upside down many assumptions we’ve been using to calculate total cost of ownership (TCO) in electronic manufacturing sourcing decisions. Everyone - OEMs and contract manufacturers - has something to lose.
Canada is not a huge outsourcing destination for OEMs, but it has been a regular North American manufacturing site because it was less expensive than the Lower Forty-Eight. The strength of the Canadian dollar changes that.
The Euro is about 50 percent stronger against the dollar than it was when many current outsourcing contracts were written for European sites. China and the rest of Asia are also seeing some, though less, impact from the falling U.S. dollar.
If your company’s CFO doesn’t have a corporate currency risk mitigation strategy, then shame on him or her. Contingencies for currency fluctuations are not typically written into outsourcing contracts.
U.S. OEMs and their CMs usually set contracts in U.S. dollars; if the dollar weakens someone takes a hit. Right now, that someone is the CM that must pay a labor force in Euros or Canadian dollars.
OEMs now negotiating contracts for future sourcing can expect U.S. dollar costs to rise accordingly, especially if they’re determined to outsource to Canada or any country using the Euro, or that has a local currency closely tied to the Euro, as many Eastern European countries do. (Most of them intend to move to the Euro but not for a couple of years yet.)
As ever, TFI’s Outsourcing Navigator can help you sort this out; in December I’ll issue a revised version that takes the weakened dollar and other changes into account.
Stanford University researchers recently completed a study on the return on investment from outsourcing the technology infrastructure for supply chains. The sample size was small, but the results worth a look.
Barchi Gillai, research director at Stanford’s Global Supply Chain Management Forum, found that when companies outsource some or all business-to-business e-commerce technology and processes, the benefits on average amount to more than twice the cost.
The study included a survey and was followed by in-depth interviews with most of the respondents. The sample size was just 25, and in various industries, not all in electronics. But the in-depth interviews rounded out the findings and provide some insight into outsourcing B2B infrastructure. At any rate, the findings are probably less interesting than some of the issues raised in the report, “Driving Business Value Through B2B Outsourcing: Improving Business Performance, Trading Partner Satisfaction, and B2B Capabilities.”
According to Gillai, benefits on average amounted to 2.5 times the costs of outsourcing B2B infrastructure. All but one survey respondent reported that their expectations for B2B outsourcing were met or surpassed, according to Gillai. The supply chains studied also reduced the number of employees needed to run complicated supply chain networks.
In the survey, 75 percent of the respondents reported improvement in their B2B technical capabilities; 55 percent reported an increase in predictability of B2B IT costs; 62 percent reported an increase in customer satisfaction; 57 percent reported higher productivity of internal IT resources.
We’re interested in the experiences Technology Forecasters members have had in outsourcing part, or all, of their B2B supply chain infrastructure. Please write a post.
A series of recent encounters suggest a software development with substantial impact on the hardware business.
First, a professor of computer science at Georgia Tech called me to talk about virtual machines. He runs three operating systems — OS X, Windows and Linux — on an Apple MacPro, and uses his favorite tools from each environment, cutting and pasting seamlessly from Word 2000 running on XP to iPhoto running on OS X and so on.
Then I asked a friend who runs a department at a high tech marketing company whether virtual machines make a difference. “Are you kidding?” he said. “I used to buy ten PCs and ten Macs for a new team. Now, I get fifteen Macs and run them as PCs when necessary and I need fewer administrative resources.”
The success of VMWare proves that he’s not alone.
Next a consultant for 3Tera called. He had a hard time explaining what this early stage virtualization company does. The easiest part to grasp is “clustering.” They can link any number of servers together and treat them as a single virtual machine. Drain on CPU resources is minimal and managing the system is pretty easy.
The dramatic part is this: Their software can virtualize all the functions of a client’s network — take the firewall, the storage arrays, and other hardware and reproduce them in software. The client can tweak the system, then press a button and the virtual network propagates to the whole network. One of their clients is a large hosting company. The hosting company used to add one system administrator for every ten new servers; where 3Terra is deployed, it adds one for every hundred new blades.
Now consider this. Vanu makes a radio that runs in software for cell phone base stations. Bug fixes and upgrades can be deployed remotely to a carrier’s base stations.
What does it all mean? Clearly software is going to give more power to low-cost, generic machines. A complex device like a firewall currently needs much more than a CPU, but if it can be run in software on generic machines, then many markets now served only by high mix, low volume specialists might fall prey to the high volume ODM.
One limitation of virtual machines: The extra layer of software reduces performance — a hit of 5 percent to 7 percent on a PC. Manufacturers might respond to the competition by focusing on very fast throughput devices that cannot accommodate any reduction in performance. Other manufacturers might incorporate virtual machines into design for manufacturing programs, using virtualization as a way to reduce costs.
The trend to virtual machines, however, is likely to lead the hardware industry further toward thin margin, commodity business, especially in the computer and telecom infrastructure segments. It looks like software is hot on the hardware manufacturer’s trail, and they’ll have to sprint to stay ahead.
Last year, TFI launched the EMS-ODM Report Card program. Our goal was to help OEMs get objective information about leading EMS providers and provide EMS companies with insight into areas for improving customer satisfaction. This year, after receiving valuable feedback from last year’s participants, we’ve expanded the survey to include electronics distributors, fine-tuned the survey, and expanded the base of respondents to get even more useful, targeted information in five critical areas:
As the outsourcing trend continues, OEMs in market segments such as medical and automotive electronics are outsourcing for the first time. At the same time, OEMs with experience in outsourcing are refining their strategies. This research will benefit both sides of the outsourcing equation by uncovering strengths and weaknesses in their relationships. And it will provide TFI will valuable insights to assist these OEMs select and manage relationships with their EMS providers, ODMs, and electronics distributors.
All participants will receive a copy of the executive summary of the results and all TFI Quarterly Forum members will get access to the complete report.
If you work for an OEM, please take some time now to complete the survey. If you work for an EMS provider, ODM or distributor, please forward this blog to your OEM colleagues. Your participation is greatly appreciated. Thank you.
Some things are just so much common sense that we tend to overlook them. Here’s one piece of such wisdom.
Sparton Corp., a member of Technology Forecasters Quarterly Forum, has been pursuing a quality program it calls Performance Excellence — a combination of Lean, Six Sigma and common sense. For more details, listen to an interview with Spartan CEO David Hockenbrocht on our website.
Among other things, Hockenbrocht says: “When we go after excess costs and drive them down to what we believe is the lowest cost solution in almost anything, several other things happen. First, the quality of the output goes up — and that is very quickly measurable. Secondarily, the speed of delivery also goes up. We get two not so much thought about outcomes when we go after costs: Improved quality and higher speed.”
If you stop to think about it, that equation makes sense, but relatively simple concepts can get lost amid the complexities of the EMS business.
Hockenbrocht also reminds us that successful EMS companies are striving for more collaborative partnerships. Sparton, for example, is extending its Performance Excellence effort to at least one of its OEM customers. Based on a survey conducted earlier this year, Technology Forecasters concluded that adoption of Lean among partners will be the next crucial phase in driving out inefficiencies in supply chains.
Hockenbrocht also sees the sharing of goals as a key to partnerships. He tells how one customer recently invited about three dozen of its most important suppliers, including Sparton, to an event where they got to look at the OEM’s three year business plan in great detail. “I was so impressed with the openness,” Hockenbrocht says. “This is where a business partnership needs to be.”
Collaboration will be crucial to electronics manufacturing success in the years ahead. Kinaxis, a strategic partner of Technology Forecasters, always has interesting thoughts about supply chain collaboration. Its blog posting of October 4 is worth a look.
For obvious reasons, people with waterfront property might care more about global warming than people who live on mountaintops. Likewise, you’d think U.S.-based EMS companies with razor-thin margins might care more about curbing health care costs than industries with profit margins wide enough to absorb this growing expense.
Oddly, that does not seem to be the case. U.S. electronics industry thought leaders who address this problem are the exception, not the rule.
The continued health of the U.S. manufacturing base depends to a great extent on the productivity of American workers. Productivity is a function of input versus output - the cost of workers in wages and benefits versus the value they create. That ratio is heavily impacted by rising health care costs. Reining in these costs through government policy is the key to keeping the productivity of American workers high enough to justify manufacturing domestically.
In its most recent annual study of health care costs, the Kaiser Foundation reported that the average cost of employer-provided family health insurance in the U.S. increased to $12,106 per employee family in 2007. The employee pays about one fourth of the cost and the employer pays the remaining $8,825. To put that number in perspective, it’s enough to pay the fully burdened costs of a semi-skilled worker on an SMT line in China for six months.
According to Kaiser, these insurance costs have risen 78 percent since 2001. This phenomenon is widely remarked on in the automotive industry, where obligations to current and retired workers add nearly $1,500 to the cost of each American-made car.
EMS companies suffer less than automakers from the legacy of large numbers of retired, unionized workers, but in an industry with average after-tax profit margins hovering around 2 percent, an extra $8,825 per employee could make the difference between a production line staying in the U.S. or going overseas, or the difference between winning a contract and making a dime.
Several of the 2008 presidential candidates have now weighed in with health care proposals. The U.S. electronics industry has a major stake in the evolving debate over how to resolve rising health care costs. It is time to pay attention.
On a tip from a Technology Forecasters member, we’ve discovered a Catch 22 for some CMs that want to do certain kinds of business in medical electronics. It stems from a well-intentioned government effort to reduce paperwork. Here’s the scoop.
Medical electronics is highly regulated by the Food & Drug Administration. There are several hurdles a medical OEM must get over - and these often apply to CMs, too. As spelled out in the Code of Federal Regulations (insert “Section 807.20″ and click “search”), OEMs and CMs are required to register with the FDA and list specific devices they are bringing to market.
In June 2005, the FDA eased the rule for domestic CMs, declaring they only need to register if they ship devices directly to end customers or into commercial distribution. Domestic CMs building PCBs, sub-assemblies or even finished devices that they send to the OEM need not register. The FDA further elaborated a month later.
Clearly, this reduces paperwork for some CMs. However, the Law of Unintended Consequences got in the way on two fronts. The first is your basic chicken-and-egg problem.
As TFI concluded from research earlier this year, many CMs want to offer medical OEMs end-to-end services, including shipping completed products directly to the OEM’s customers or into commercial distribution.
But now they can’t register with the FDA until they have secured business that includes distributing completed devices. Many medical OEMs - new to outsourcing - are not up to date on the amended rule, and still expect CMs to be registered. CMs will have to educate OEMs.
The second issue is competitive advantage. CMs already registered with the FDA, even if they were not shipping products directly to end customers, remain in the registration database. If they now compete for business that involves placing devices into commercial distribution, they have an advantage over CMs who were never listed. From the OEM’s perspective, one CM is listed and one is not.
The FDA may soon address this matter. A bill before Congress (Item 4 on Page 2), would authorize the FDA to remove all CMs listed in the database that are not shipping products directly to customers or into commercial distribution.
This issue most likely impacts medium-sized and smaller CMs that are just starting to enter or expand their medical electronics business. We’d like to hear your experiences.
When Flextronics International and Solectron announced their merger in June, many of us wondered whether this signaled a new round of consolidation and restructuring for contract manufacturers. It would appear that it does: Restructuring through internal downsizing is already taking place.
In a presentation at Technology Forecasters Inc.’s Quarterly Forum last week, Louis Miscioscia, managing director of Cowan and Co., an equity research firm, offered evidence that CM plants outside Asia generate less revenue per square foot than those in Asia. He estimates geographies outside Asia are getting $500 annually per square foot of manufacturing space; most Asian locations, at fuller capacity, are getting as much as $1,000 per square foot annually.
“This would suggest that there is still too much square foot capacity and it needs to come down more,” Miscioscia said in an email exchange with me.
He says how much restructuring is underway is less clear: “It is hard to tell since we only get square foot on an annual basis and almost every EMS company is restructuring now.”
He offered several estimates for the year to date: Celestica has closed 306,000 square feet in Europe; Jabil Circuits, 656,000 square feet in Europe; Sanmina-SCI, 260,000 square feet in Europe, and Solectron, 400,000 square feet in North America.
Flextronics and Solectron hope to complete their merger by the end of the year, and Flextronics has estimated the combined companies will cut costs by $200 million within two years. That certainly involves some restructuring of plant space.
Miscioscia pointed out that restructuring can lead to disruptive costs and excess inventory issues while management focuses on downsizing costs rather than winning new business or improving operations.
We’d like to hear from OEMs and from EMS companies on this. Let us know your thoughts.
We can expect reasonable growth over the next five years in the market for aerospace, defense, and homeland security electronics, according to new research from Technology Forecasters Inc.
“Electronics in Aerospace, Defense and Homeland Security: Repositioning for the Future,” a Quarterly Forum Report released this week, also concludes that significant opportunities exist for suppliers to serve customers in these segments as they increase their reliance on outsourcing.
TFI estimates the total available market for this sector will grow from $112 billion in 2006 to $152 billion by 2011, a 6.3 percent compound annual growth rate (CAGR). EMS revenue from the sector will increase at a 9.3 percent CAGR, from $2.8 billion in 2006 to $4.4 billion in 2011. EMS penetration will remain relatively low in the 3 percent range.
The research conclusions are based on in-depth telephone interviews and an online survey directed at management and senior technical personnel. We also conducted secondary research in the appropriate markets.
In choosing OEMs, these customers focus primarily on quality, dependable delivery, a competitive price, technical capability, and adherence to regulatory standards. Flexibility, responsiveness, and customer service also are critical issues for customers.
The aerospace/defense electronics industry is attractive to suppliers because of growth potential and portfolio diversification. However, the report flags several cautions:
-Contractual requirements can prove to be challenging.
-Demonstrated experience in the government and aerospace contracting arena are critical for selection as a partner.
-Government contracting moves at a much slower pace than the commercial world. Up-front investment may take years to recover. Project cancellations can occur.
-Margin pressure in aerospace/defense business is not as critical as the commercial world, but cost management is still an issue.
How do these match your company’s experience in aerospace, defense and homeland security? We’d like to hear from you.