What works in one industry may not always work in another.


Today’s volatile economy has forced most manufacturers to explore outsourcing. Many companies see it as an opportunity to reduce costs, improve flexibility and streamline production processes.

Outsourcing is similar to leasing a car; there are advantages and disadvantages to the strategy. Because there are numerous variables and risks involved, it can either be a cost-effective tool or a hair-raising experience.

To examine the pros and cons of outsourcing, Boothroyd Dewhurst Inc. (Wakefield, RI) hosted a panel discussion at its 18th annual International Forum on Design for Manufacture and Assembly (DFMA) in Newport, RI. Executives from three diverse manufacturers discussed the strategic role of outsourcing at their companies. The panelists were:

  • Robert Farra, P.E., director of engineering and new products introduction at MedSource Technologies Inc. (Newton, MA). The Minneapolis-based company is a contract manufacturer that specializes in medical devices. It operates 10 production facilities in the United States that assemble surgical instruments and devices used for cardiovascular, orthopedic, drug delivery, neurology and urology applications.
  • F. James McWilliams, project manager at Hewlett-Packard Co. (Marlborough, MA). The Palo Alto, CA-based company consumes 15 percent of the world’s dynamic memory chips. It boasts annual sales of 40 million printers, 18 million PCs and 1.2 million digital cameras.
  • Jay Mortensen, director of target costing and cost engineering at Maytag Corp. (Chicago). The Newton, IA-based company is a leading manufacturer of home and commercial appliances. Its brands include Amana, Hoover, Jenn-Air, Magic Chef and Maytag.

Here’s a brief look at how each company approaches outsourcing:

Like most companies in the electronics industry, Hewlett-Packard has embarked on an ambitious plan to outsource much of its manufacturing activity to electronics manufacturing service (EMS) providers. “The object of this strategy is to reduce total manufacturing cost and provide higher manufacturing responsiveness to changes in volume,” says McWilliams. “Our experience has been mixed, with both rewards and pitfalls.”

Hewlett-Packard is the world’s largest consumer of contract manufacturing. The company began outsourcing PC manufacturing in 1993. Today, it has contracts with many leading EMS companies, such as Flextronics International Ltd. (Singapore), Jabil Circuit Inc. (St. Petersburg, FL), Sanmina-SCI Corp. (San Jose, CA) and Solectron Corp. (Milpitas, CA). For instance, many of its best-selling printers are manufactured by Flextronics.

Maytag has been a vertically integrated organization for many years. However, the appliance manufacturer recently realized that this model isn’t always efficient. As a result, Maytag has developed a corporate strategy to look at each business unit and use outsourcing if it is cost-effective. For instance, the $4.7 billion company recently entered into an agreement under which Daewoo Electronics Corp. (Seoul, South Korea) will manufacture top-freezer refrigerators for Maytag.

“We’re trying to establish what we call our should-cost model very early in the process,” explains Mortensen. “We’re using our suppliers to…help us develop [cost] models. We are now introducing target costing tools—primarily the DFMA process—very early in the concept and project feasibility stage…long before parts are approved and projects are approved.”

Although MedSource is a contract manufacturer, it also conducts design work for many leading medical products companies. “We develop not only the devices, but components for our customers,” says Farra. “We do the development work and then we transfer it directly into our facilities to do the manufacturing. Our customers expect us to manage that entire supply chain for them. And we’re also in a very regulated industry, so design control is a key component of what we have to do.”

Outsourcing Realities

Outsourcing manufacturing is not the slam dunk that is portrayed in numerous business school case studies. “The theory of outsourcing is that the EMS operation can provide a more efficient use of capital and personnel, and absorb those costs over a larger volume, thus lowering the effective cost per unit,” says McWilliams. “One of the great disappointments of outsourcing is that costs may not actually go down—they may go up. If your product requires specialized production tooling or jigs, nonstandard processes, or specialized test and inspection gear, your savings will be diminished.”

Outsourcing can also be an emotionally charged issue. “You’re taking somebody’s baby and giving it away,” warns Mortensen. He says some people have put more time and effort into developing a machine or process than they have spent with their own children. So, if you take it away because of outsourcing, it’s like you’re taking away a “member of the family.”

“One of the tools that we’re using [to address this issue] is DFMA,” says Mortensen. “It provides us with a structured process that the engineering group and the factory group have an input to. They can specify what data goes into it, so they feel that they are involved in it. We eliminate some of that human factor, but it’s still emotional.

“We try to consider the human factor by involving people from the various disciplines up front and going through the product so the various questions can be answered,” explains Mortensen. “Many times, you get concerns, prejudice and open hostility to outsourcing. We try to walk people through the decisions and to address the issues as they arise in doing our analysis and comparisons.”

“Outsourcing, in general, results in downsizing the existing engineering staff,” adds McWilliams. “They either go away or they’re [hired by the contract manufacturer]. There’s nobody left at home to do the work, and we’re forced to trust the EMS vendor to provide it. It’s not a good situation. Often, the contract manufacturer tends to be light on engineering analysis, but heavy on engineering support.

“They are absolutely great at fixing the process,” McWilliams points out. “Most of them are very good at keeping control. But, optimizing the product is not yet a core competency out there.”

McWilliams says outsourcing challenges traditional rules of thumb. For instance, one unforeseen problem that he’s encountered is color blindness. “There are several populations in China that have a higher than normal incidence of color blindness,” explains McWilliams. “Interestingly enough, it’s not red-green, which is very common in Europe and North America. Instead, it’s blue-yellow. We had color-coded a bunch of assemblies in pale blues and pale greens, and [a lot of people employed by the contract manufacturer] couldn’t tell the difference. So, we ended up putting little stickers on them with numbers and they did very well.”

Panel Discussion

The following is an edited transcript of the panel discussion:

Q: The basic approach of DFMA is to bring predictive cost information on manufacturing and assembly into the early stages of product design. How does this affect strategies for outsourcing?

Mortensen: We’ve chosen to acknowledge the human factor and work with it and not around it. We use DFMA to sit down with the correct parties, such as people from product planning and marketing. We also [consider] cost structures for international locations and selected locations in the United States to help us determine where is the best [plant location].

In our concept feasibility stage, we use DFMA to help us identify potential [money-saving] opportunities. Outsourcing depends heavily on how you leverage your overhead. For instance, what products should we be [assembling] in our facilities to leverage that overhead more efficiently? And, what products should become candidates for outsourcing?

We try to define the scope…of outsourcing, identifying all the costs in the factory and in the organization that are impacted. We identify those costs…that we can drive out. With the DFMA modeling tool, we can track these things and develop different options as we move along. Each head count is identified. Each cost is specified.

McWilliams: Predictive cost modeling and process modeling tools are important [for evaluating contractors]. [You’re] dealing with a third party that has a profit incentive to keep things the way they are, whatever it has negotiated. Any process improvements [it comes up with, it] may pocket. You won’t know about it because there’s a loss of information flow. And the way I can make up those broken links is by [gathering] a lot more information on my side.

We need to optimize the design to fit into somebody else’s process. Before, our design engineering team knew where a product was going to be built. They designed it to fit into the machinery and the processes. Now, you’re dealing with somebody else. One of the things that we’ve had to do is expand the equipment libraries in a lot of the costing tools. We’ve had to customize material libraries because, as you know, a pound of cold-rolled steel does not cost the same throughout the world. As a matter of fact, in the Far East I’m still wondering how they get it for as cheap as they do. And there aren’t as many people we can go [to]. We use DFMA tools to support the contract negotiation team on the cost and the process capability issues.

Farra: The medical industry lags the appliance, automotive and high-tech industries a little bit in terms of outsourcing and its mind-set with regard to DFMA. If you listen to the manufacturing folks on the development teams, their focus quite often is on time-to-market and product cost. And [if it doesn’t work] their mind-set is, it’s R&D’s problem.

DFMA is a key component that we use very early during the planning stages. We have a lot of specialists at different facilities, from injection molding to stamping to machining to precision metal injection and coil fabrication. We utilize their input on DFM and DFA in the cost modeling. We also use analytical tools to predict whether a component will be adequate or safe. And we use a home-brewed product we call a “designer rationale tool,” which extracts all the requirements from a designer and translates them into manufacturing and assembly requirements.

Q: As we look at outsourcing, it seems like the focus is on cost. A different target is productivity. What things would be different between a productivity focus vs. a cost focus?

McWilliams: Cost is a quantifiable method of keeping score. And the problem with productivity is it’s a little harder to keep score by it. One of our problems was that we were decreasing the labor content of all of our products faster than we were selling more products. So we were [increasing] productivity. Normally, if we were growing quicker, we would just absorb those gains. The problem was, we weren’t. Now, can you sell that capacity? We tried to do that for a while and weren’t very successful at it. In the best of all possible worlds, [you should attempt to] push up productivity and redeploy those people to build more things for you, bringing the organization more revenue, more profit and more capability. Unfortunately, I’ve never figured out how to do that.

Mortensen: I [previously] worked at Toyota and was involved with the NUMMI plant (Fremont, CA). I agree, in a perfect world, I would love for Maytag to be a Toyota-esque company. We’re trying to achieve that same type of mentality, but right now our cost structure and our production strategy is such that we recognize we have certain operations in-house that just aren’t compatible with the strategy. So, as we reformulate a strategy, it is number-driven. [We only want to do] those things in-house that really make sense, that we can do well. [We want to] shift items to offshore or domestic suppliers so that we can avoid the capital investment. We want to get to the point where we’re using our factories properly and we can get to a productivity-based process like Toyota. But, in the meantime, we need to take advantage of other people’s capital to restructure ourselves.

Farra: Many of our customers end up redirecting their development and manufacturing folks on some of their core products. They’re able to focus better and identify next-generation products and really drive innovation within their company without having to worry about their more routine products.

Q: How much work actually is involved when you’re starting up with a new contract manufacturer? What percentage of the work is involved in auditing? How do you manage that when you’re developing a new product with the contractor?

Mortensen: We do a lot of up-front work to identify [contractors that can] maintain the quality levels and productivity levels that we’re after. We closed down our joint-venture operation in China, which was going to be our major vehicle for outsourcing to China, primarily because of some of the risks that we didn’t properly identify. We lost less money on it than other appliance manufacturers did on their joint ventures. So, having gone through that and learned the hard way, we’re now trying to spend more time developing a supplier before we give them the business. Hopefully, we don’t have to put in the tremendous resources of constant auditing, but there will be some.

McWilliams: One other thing that. . .we’ve done very successfully is find out who else [the EMS provider] ships products to and go interview them. Take them out in a less formal setting and start asking the hard questions. What do you see with quality? What do you see for responsiveness? But, other than that, we keep running totals of defective units. We keep cosmetic defects. We have upper and lower control limits, and [we ask for] corrective action.

Q: How do wildly fluctuating foreign exchange rates affect the outsourcing decision?

McWilliams: The magic word is “hedging.” We do it. All it does is delay the cost impact of a currency exchange. But, other than that, it can definitely tilt the tables. You’re seeing it right now with the dollar going down, and certain European currencies going up. Another issue is any profits put in various tax havens have a problem being repatriated.

Mortensen: We also use hedging. One of the other things that we do with DFMA is a sensitivity analysis. For instance, at what point does a currency change make this not a good thing to do? We have a fairly robust risk assessment [that identifies] everything from air freight delays to currency issues. Beyond that, we cross our fingers.

Farra: You may want to consider working with an outsource partner that has capabilities overseas, but gives you a price based on U.S. currency. So [it takes] that risk [instead of] you.

Mortensen: When we did that, we found [that contractors had a far greater] profit margin. So, we’ve chosen to take the risk because we feel in the long run we can manage it better.

McWilliams: In China, most contracts are dollar denominated. In Taiwan, they are not. It depends upon the locality. In some places, the dollar is more stable than the local currency.

Q: How does the process of selecting a contractor differ if you’re outsourcing product design rather than product assembly?

Farra: With regard to design outsourcing, a lot of our customers want to make sure that we have a product development process that is very compatible with their processes. So they want to make sure that we follow acceptable design controls. We’re going to use a phased development or milestone-based development. They want to make sure that there are checks and balances, whether you want to call them stage gates or design inputs and outputs of each phase. They also look at our technical capabilities. For instance, do we have design engineers who have experience with a product like this one?

We need to demonstrate competence in [mechanical and electrical engineering]. Do we use analytical tools? Do we use finite element analysis and mold flow analysis to guide the design process? [Our customers] want to make sure that there isn’t the throw-it-over-the-wall mentality that they sometimes experience in their own facility. They want to know that manufacturing is integrated into the development group.

Q: How long did it take before you were comfortable with what was coming out of those spreadsheets?

McWilliams: It’s a continuing development process. Our models keep on changing as the environment around us changes. One of the things that we always do at the end of a project is track the cost estimate. [Often,] we end up correcting the number because the product two-and-a-half years ago is not the same product [today]. It doesn’t have the same features. But, after all that correction, we generally find that our cost estimations are within 10 percent, which is really good. It’s the power of these DFMA tools.

Mortensen: We use ours as a scorecard to circle back after the project’s in place and identify how close we came on risk assessments. And these sheets can be quite lengthy. We use ours as a learning tool. For instance, where did we err and why did we err? As we do that, we’re getting a lot more acceptance by people to use them and to rely upon them.

Farra: We try to develop a cost baseline. And then we look at all the risks associated with the development activities or the manufacturing. We… then go quickly into a small manufacturing test or bread-boarding activity to make sure that we have a good understanding of those risks. Then, based on those, we’ll provide a cost range. For instance, the baseline is coming out to “x” dollars, but given the following activities or assumptions or complications, it could be impacted by whatever percentage. And then we work with our customers very carefully to continuously manage and update those activities.

Q: It is increasingly important to make sure product designs are right before they go out the door for outsourcing, particularly when the contract manufacturer is a long way from home. You can’t just get on an airplane and go to China because something’s gone wrong with a machine. Error-proofing is more essential. How do you handle mistake-proofing?

McWilliams: We treat outsourcing as a black box. Whatever you’re outsourcing, you need a set of specs and criteria for acceptance that are much more specific than you’d ever have if you were dealing with your own plant. All of a sudden, product specs at the sourcing level are much larger and much more comprehensive than they’ve ever been. We can claim we want to control mistakes, but the reality is that the only thing we can do is accept or reject products as they come in. However, our favorite type of error-proofing is color coding.

Q: How are you using DFMA tools to address outsourcing issues?

Mortensen: We use it to get manufacturing engineers and people from the shop floor involved in our cost analysis. We use it as a framework to bring these issues out.

Farra: We use DFA software in two ways. Our engineers on the development team utilize the software. As they. . .enter the information, they make recommendations back to the designers. In addition, we have our customers [meet with] the entire development team. They go through an exercise of doing a DFA activity and some of the DFM activities. It facilitates dialogue so that we can predict not only the manufacturability aspects [of a product], but what kind of changes we want to make to simplify the design so that it allows us to produce a quality product at the right cost. You have to treat it like you would any predictive tool, whether you’re predicting manufacturability or actual product performance. It all depends on the quality of the input and the type of dialogue that occurs.