Since the beginning of the modern electronics era, between World War I and World War II, numerous facets of engineering, design and manufacturing have needed to work in harmony in order to manage an electronic product from concept through manufacture. As industry and technology have driven massive changes in the way we engineer and build the devices that power every day life, many critical elements to this process remain relatively unchanged in their importance and process.
The advent of computers in industry, particularly in CAD, CAM and robotic manufacturing processes has removed much of the "human touch" from the mechanical process of physical design, manufacture and assembly of items we take for granted. Amidst this sea of change, one of the most important areas, the graphical translation of engineering schematic to manufacturable database still remains an area dominated by human expertise and skill.
This is the realm of the service bureau.
The term "service bureau" refers specifically to the agency-usually a stand-alone-entity-that is tasked with the competent translation of an engineer's wishes tempered by the printed wiring board manufacturer's requirements. There are three major types of service bureaus. This article will not discuss the captive service bureau-typically a division-within-a-division of an OEM-though many large OEMs have extraordinarily competent and large design and layout groups.
The largest service bureaus are almost exclusively owned and run by a handful of large contract electronics manufacturers (CEMs). Although they may contract out and be very competent, their parent corporation will guide their final mandate and that is usually to act as a "value-added feed" to the CEM's volume-based profit areas.
All three sizes of service bureaus share common issues, concerns and requirements. How each elects to handle those issues defines their viability and growth potential in the future.
A brief history
Many service bureaus came into being, grew and flourished during the 1980s. This period, marked by heavy military spending and a shortage of truly "new' electronic technologies, created the perfect breeding ground for younger entrepreneurial PCB designers. The availability of large, regional accounts (many focused on DOD work) gave a new generation of skilled "rule and tape" designers huge captive workloads where they could hone their skills as designers and business people. Concurrently, the early consolidation of numerous defense contractors fed this frenzy by creating larger and larger contract-based engineering groups, constantly short of the support staff required to see their contracts through to fulfillment. Boeing on the West coast, Northrop on the East and scores of accounts in-between brought the service bureau into its own by the late 1980s.
At this juncture and with the slowdown of military spending in the early 1990s, two distinct entities began forming. The more aggressive service bureaus-those lucky enough to have "tradesmen" with business acumen to see the future-sought commercial clients, particularly in the burgeoning and profitable telecom sector. The smaller agencies, as well as those not gaining tool knowledge or additional staff as quickly, locked into a smaller set of clients.
The evolution of those two groups today is clearly evident in the 7-10 "super-regional" service bureaus who perform probably more than 60% of the design work nationwide for a huge client base covering the entire technology spectrum-from defense through optical networking. Conversely, this group competes with a much larger pool, probably 200+ "local" agencies, maybe each having four or five primary clients.
The expense of modern design tools, on-going training and hiring issues, and an ultimately shrinking client pool threatens to toll the death knell for many "local" agencies. Super-regional groups can manipulate tool selection, perform 10-12 multiple simultaneous design efforts (for the same client) and grow into additional services-all at the expense of the "garage" shops. Though this group of smaller agencies will always exist in some fashion, they continue to go the way of many "mom and pop" businesses in a modern economy based on scale, value and growth.
The mid-1990s saw the emergence of the third primary type of service bureau-those that are attached to their contract manufacturer parents. Many CEMs have seen the light in this regard as their blue-chip client base drives them to perpetually lower the costs associated with bringing a product to market. They have rightly calculated that by bringing one of the most expensive out-sourced services inside, they stand to earn more and exercise a greater level of control over their assembly plant throughput and clients' schedules.
Paradoxically, in instances when the parent firm does not maintain the management expertise required to run this complex business sector, the result is additional business for the "super-regionals." The client base best served by large-volume CEMs is often worst served by their design divisions. Inept planning, poor design skills and lock-step operation theory condemn many elegant and critical designs to the trash heap of blown fabrication efforts and unbuildable boards.
Today's business growth in PCB design is fueled explicitly by the expertise and investment of the mid-tier "super-regionals." The marketing efforts, tool investments, additional offerings and additional staff that this group pursues will define excellence in design and American competitiveness (on a global scale) into the far future.
Regardless of the type or scale of service bureau being discussed, future growth will consistently be determined by two primary factors: 1) broad and continuous client acquisition, and 2) broad and continuous skill acquisition (defined loosely by the addition of tools and services at the request of newer clients).
By constantly adding to its client base, particularly outside of its "normal" customer-i.e., those who require new and different design skills, and those outside of a service bureau's "comfort" zone of technical skill-a modern agency virtually assures itself growth. If one sector dries up, there will always be something else they are good at or another client type to pursue.
Pursuit of growth in this style requires the addition of skills and techniques required by an ever-expanding customer base; consequently, the modern growth-oriented agency may find itself faced with mastering a tool or technique initially and exclusively utilized on behalf of an older client.
This is the perpetual dilemma of the smallest of agencies: Buy the tool to satisfy the client, learn the skill to satisfy the client or say good-bye to the client. Then what? Say an agency invests and learns, but if it isn't capable of applying what it has learned to additional new business, all is for naught. This is the chicken or the egg theory taken to new heights but it plays itself out every day, every month in every major metropolitan area in this country and usually to the advantage of the "super-regionals."
An agency determined to grow must be prepared to face all the sacrifices of any growing company. The typical service bureau will go from $250k-$500k billing in a year (or for a period of years) to a doubling and doubling again as new tools feed new clients and new clients demand new tools. Only the strong survive this process without sacrificing the "human touch" that made them important to start with.
The best agencies will constantly pursue relationships and challenges that foster their growth. Networking within an existing client base...pursuit of additional client locations as customers...and merciless acceptance of temporary licenses (no matter how strange or worthless the tool may appear). Throw these things in with competent sales and marketing efforts, and growth will occur-perhaps more quickly than an agency is prepared for.
The reality check for many service bureau owners is this: A good designer is able to earn what a good engineer does. Inherently, this appears out of wack for the small service bureau owner who has spent a career building his /her own business. All day long the small agency is visited by client engineers sporting multiple higher education degrees. These same engineers may sit down to discuss their product requirements with a 25 year-old designer who decided not to attend college, yet may earn more than they do!
Reality beckons. Rough math says that a very busy original equipment manufacturer (OEM) may require, but not actually have, a maximum of one full-time designer for each group of four engineers-assuming an engineer is only working on one board at a time. Would you, as a service bureau owner, rather lose the work of these OEM engineers or pay the going rate for a talented designer? A service bureau may not have much control over the pay scale, but the ROI of a designer who works for four (or more) client engineers is well worth the pay-even if the pay scale is dictated by the going rate at the typical OEM.
If growth is the goal of service bureau ownership, then pay scale equality (with the client base) and the tacit understanding of its importance is paramount in this effort. Growth occurs most easily in two distinct areas: additional work from existing clients and taking over all work from existing clients. In both cases, an agency's design staff is accomplishing for it what an OEM designer did until the agency stepped in. At the same time, new opportunities arise for hiring as an agency increases its workload from the existing client base.
Currently, there is a very real, very valuable move on the part of the big "IP" or "Patent house" OEMs to outsource more and more "hard" engineering skills. Many OEMs find it tough to stomach the cyclical nature of the design side. They can keep an engineer busy all year long for $70k, but a designer, earning approximately the same amount, may only be busy for a month out of any given quarter. This is a fantastic opportunity for the service bureau: skilled designers, familiar with their client's design requirements. The obvious downside to the service bureau is the pay scale equality requirement.
Above and beyond base pay, the technical personnel within the service bureau industry have absolutely the same requirements as many tech professionals within the "soft" engineering industries. More family time, flexible working hours, additional income opportunities, training and self-improvement opportunities, and company benefits (such as stock purchase plans) are all things that drive a successful, stable PCB design force. This applies to all types of staff in the bureau-from experienced designers to CAD specialists, data prep, pre-engineering specialists and the admin staff that supports this group.
Another of the most serious concerns for today's service bureaus is employee poaching. Many of an agency's best potential clients are its worst potential employee thieves. A prime example is an optical start-up. This type of client, requiring extreme technical design and layout skills in high volumes for multiple boards at an unrelenting pace, appears as a huge earning opportunity for the competent and focused service bureau. However, service bureau ownership is well advised to closely watch this type of client. Many service bureaus choose to have newer clients sign employee "anti-theft" documents specifying that the client will be liable for the agency's lost income should the client poach an employee. Though a document like this may not dissuade some clients from trying to poach an agency's employees, it will take the fiscal edge off. Of course, agency owners should closely evaluate the long-term value of any employee so quickly lured away. (Perhaps the client has done the agency a favor.)
Choosing a service bureau
Today, the service bureau represents a critical component in the product design and manufacturing chain. Without competent PCB designers, CAD engineers and parts librarians, many electronics companies would simply grind to a halt. How a company selects, interacts and communicates with a selected partner in this endeavor is key to their success. The rest of this article will focus on a few of the challenges potential clients (like you) and service bureaus face during this process.
Every day service bureaus are "pinged" about their availability to perform various types work for potential new clients. As in all service industries, understanding a potential client's expectations and requirements, as well as quirks and concerns, is the only way to interact in a manner that is profitable to both parties.
When deciding on the right service bureau to fit your needs, you need to identify your primary requirement for needing outsourced design and layout assistance in the first place. You may approach a service bureau for the obvious reason-that you don't have this skill set internally. This has become the "vogue" reason for outsourcing design. Many OEMs start out and later will "end" with an IPO or purchase, complete with fully virtual design, layout and even schematic groups in some cases. Still, these companies must identify their requirements when approaching possible service bureau vendors.
Is time-to-market your only requirement? Accordingly, are you so speed driven that tool choice is irrelevant, thus the service bureau will make the tool selection based upon bandwidth and your requirements? Is there a possibility that future partnerships may mean that legacy work must be available on a certain platform? (This is an oft-overlooked consideration.)
If you know that your company will end up as a purchase target for OEM "X," it makes sense to investigate what that company's tools of choice are. Deals literally could be shot down if an OEM in pursuit of your patents can't actually determine how to implement their application. This alone is a good reason to avoid "non-standard" or archaic layout and design tools. By the time you are showing off seven or eight circuit boards to a potential suitor, it might have $500k or more invested in design and layout alone. These are unrecoverable funds if someone should decide that "these designs are useless to us...."
Another important thing to consider when selecting a service bureau is history. If the service bureau can't provide numerous references and/or doesn't have a multi-year history, you should consider if this is the right provider for your company. This requirement feeds the last concern you should have. If the service bureau is small, will it be able to handle binding legal issues or the rider required to assure you that faulty design and layout, schedule issues or other pitfalls won't cost your company money as well as time.
You should ask to see the service provider's credit information in order to ascertain that it is, in fact, a legitimate business (if there is any doubt at all). Some clients even ask for proof that the software tools being used are rightfully licensed to the agency. While this may appear extreme, consider that a bootleg software tool could irretrievably corrupt a database, rendering it useless in the future. Where would you turn should this occur?
Mindreaders and fortune-tellers
Potential clients have many misconceptions about service bureaus. I don't think it is quite as bad as this section heading may imply, but it can get weird. Many clients don't adequately understand the relationship between the graphical part of the service bureau task and the intelligent, electronic part.
Essentially, you deliver a "stupid," non-graphical collection of information to the service bureau. The first step on the part of the service bureau is to add "intelligence" to this in the form of graphic decals (part types) with embedded attributes and information. The next step is to convert this now-intelligent collection of data into three dimensions in the form of a PCB database specifying layers and connectivity.
You need to understand that certain things must be provided in order for the task to go smoothly. Many clients assume that approving the service bureau's quote to accomplish a design means that the service bureau is then prepared to help with schematics, create a BOM and do AutoCAD drawings. While many service bureaus are very adept at a wide variety of services (and many, in fact, seek these additional tasks), they are not able to perform them for free-nor do they know to perform them unless asked. You must clearly and accurately respond to what is typically a very specific set of questions asked by the service bureau.
The most common misconception or mistake on the part of the service bureau-and every agency in the country is guilty of it-is assuming a client understands what they do. By taking for granted your knowledge of its abilities and skills, a service bureau can virtually assure itself of getting off on the wrong foot.
A service bureau should clearly articulate in numerous ways (in its brochures, on the Web site, etc.) what they are capable of and at what level. "Schematic capture work" is not nearly as valuable to a potential client as "OrCAD and Innoveda schematic capture work." As with many industries, information is very, very valuable. Those agencies able to intelligently disseminate it will dominate in the future.
Spec'ing and bidding
There are numerous methods of spec'ing PCB design and layout tasks. There are a few clients (and hopefully you're one of them) who consistently deliver excellent, relatively complete data packages. (I am mercilessly chided about the "flawless data" my personal clients provide our agency.) But in reality, much of the spec'ing process is an initial attempt to decipher the caliber and level of work required by you, the client.
Will we be using an existing database? Will we be updating schematics? Will we be subcontracting out the fabrication? If you don't know what the service bureau is capable of or you don't understand the process, you will not deliver useful data to work with.
As much as it is the client's responsibility to deliver viable data, it is the service bureau's task to describe, in detail, what this data must consist of. Many modern contracts will actually charge the service bureau for compromised delivery schedules; the service bureau has no recourse unless they have indicated to the client that delays in receipt of required information will result in delivery delays. Binding agreements concerning delivery, payment and performance schedules are becoming the norm-even for smaller clients and service bureaus.
Frequently, in order to save time for the client by removing entire approval steps, schematic updates become part-and-parcel to the design task itself. In the case of a larger card, this may be a separately billed task. In the case of a smaller board, it may be a "built-in" fee. Schematic translation flexibility, symbol creation knowledge and the staff to perform these tasks are realizing higher and higher ROI for service bureaus.
Numerous other services are now becoming available through the traditional service bureau. Signal integrity analysis, HDTML and Verilog core-writing skills are also "value-added" services that today's growing service bureau may offer. Many of these services are sold on an hourly basis, however some are not. Check with your service bureau.
The key to choosing a service bureau is to ask the right questions-and be as detailed as possible. That way, you will be aware of all the services and processes that a particular service bureau can provide. The key to working successfully with a service bureau is communication. And the best type of communication is two-way communication that involves both parties. Only then can client and service bureau become partners that produce truly great products.
Alex McPheeters oversees all aspects of sales, customer relations management and business growth for ACDI, a 15-year-old service bureau based outside Washington, D.C. He has spent almost 20 years in sales of electronics at the services level and the finished-goods level.
© 2001 CMP Media Inc.
6/1/01, Issue # 1806, page 34.
Reproduction and distribution of material appearing in PCD and on the www.pcdmag.com Web site is forbidden without written permission from the editor.
Contact information for Reprints.
NOTE: The articles presented here contain only the text originally published in Printed Circuit Design magazine. Any accompanying graphics and illustrations have not been recreated here. You may view the article in its entirety in each printed issue of PCD.