The bike biz

With Interbike right around the corner I thought I'd write a bit about how the bike business works—not everything about it, but the nuts and bolts of how a bike gets from the "paper napkin sketch" to your local bike shop's showroom floor.

If after reading it you think I've injected unneeded and gratuitous sex be advised that, as they say in Hollywood, "It was necessary to move the story along." There is a soft underbelly to this business—the dirty little secrets that nobody in the biz particularly wants end users to read about. I'm convinced that Slowtwitch readers are, by and large, adults and can "handle the truth."

The first thing you need to know about the bike biz is that there is a stateside component to any bike project, and an Asian component to it as well. This is true whether or not the bike's frame is built in the U.S. It is very rare, except in the case of a small, custom frame builder, for any bike to be sold in the U.S. without a significant contribution, or even control, from the Asian side of the Pacific. This is because the Japanese first, then increasingly the Taiwanese over the past 20 years, have become so expert in making certain components and sub-assemblies that even those bikes made entirely in the U.S. are not "entirely" U.S.-made.

One example of many that I might point to is my own Yaqui Carbo, my current tri bike. It's made by Ves Mandaric in San Diego, out of American-made Easton Scandium tubing. U.S.-made through and through, right? But it's got a carbon seatstay which is made in Taiwan, as are the chainstays, the bottom bracket shell, the head tube, as well as the Syntace aerobar and brake levers, the hubs and rims, and of course the Dura Ace components are Japanese made.

Over the next days I'll publish chapters describing each element of the process, after which I think you'll know much more about why your bike gets built the way it does.


In the bike biz the "product manager" is a staple. This industry is different than others in how a product manager works. In the pharmaceutical industry, for example, it is common for a drug to be managed by one person or team from the clinical trial stage through FDA approval, through the manufacturing process, and to continue on managing much about how the drug is marketed and sold. Not so in the bike biz.

A bike product manager is generally a younger fellow who will husband a model from its earliest stages through to its completion, that is, through the manufacturing process. It'll be a finished product, ready for marketing. The product manager rarely has input in how the bike is sold, just in how it is made.

Who is a product manager? Will he know anything about tri bikes if he's tasked with managing a model of tri bike? It varies from case to case.

American Bicycle Group's Jeff Menown is typical of product managers, but with some differences. He's in charge of spec'ing the parts on Litespeeds and Merlins when they're sold as complete bikes, but he doesn't have control over the spec of the frames (other Litespeed employees handle titanium frame geometry, tube spec, etc). Menown is a classic product manager when it comes to Quintana Roo's complete bikes, in that he handles the whole process. Many of QR's complete bikes are made in Taiwan, and for those Menown will rely on help from the agent ABG uses in Taiwan, who will also be the agent for certain other U.S.- and European-based bike and/or component and accessory companies. He'll also husband the entire manufacturing process of a Litespeed aluminum bike since—like QR—it is built outside of Litespeed's Chattanooga factory.

Menown is not a triathlete himself, but he's close to triathlon. He came over to ABG from Quintana Roo, and his long-term companion, Cherie Touchette, is a pro triathlete racing primarily on the XTerra circuit. He came to QR when it purchased component maker Real Design, and Menown's strength is in his knowledge of Taiwan-built componentry. For any complete bike that'll sell for less than $2000, it's imperitive to have a product manager who understands what Taiwan is capable of. Taiwan has made good frames for a long time, and can now make exceptional composite products, such as forks, cranks, and aero seat posts. Only in the last several years has it been possible to buy a high-quality rear hub and crank from a Taiwanese factory, and Taiwan is just now getting its arms around how to make a race-quality brake caliper. It's not quite yet up to speed on cassettes, and it's got a long way to go when it comes to matching Japanese and Italian shifters and derailleurs. For saddles, seat posts and stems, Taiwan is increasingly the place to shop, and its tube factories are so good that several (or even most) of the well-respected bicycle tubing brands have much of their midrange and less exotic tubing made in Taiwan. When you can spec a Taiwan-made part in a bike, you'll save money. Menown knows just how much "Taiwan" he can stick onto a bike and still have it perform properly.

I don't want to overplay how much autonomy Menown or his contemporaries have. In fact, he and they have very little. The big decisions at ABG are made in committee. Menown's job is to present the options and suggestions, and then when everybody signs off on the final spec Menown will "chaperone" the bike through manufacturing and point it toward its "date" with the end user.

Trek's situation is somewhat different. It has a product manager who oversees the entire bicycle division, including Gary Fisher and Klein bicycles, and his name is Joe Vadeboncoeur. Underneath him is John Riley, who is product manager for just the Trek brand. Neither "Joe V" nor Riley are triathletes. But Trek is cognizant of the need to have those in charge of a project that actually are experienced in the avocation for which the bike is designed, and it'll assign pro-ject managers to a particular line. That has been the case when Trek has made a recumbant or a tandem. Mark Andrews is an engineer at Trek, and is himself a triathlete, having competed in Kona a half-dozen times and has six Penticton's under his belt. He was the project manager for Trek's Hilo line.

In the case of Trek, every bike model has its pro-duct manager (Riley) and also maybe a pro-ject manager (Andrews for the Hilo).
There is also a design engineer, which in this case is also Andrews. The term "engineer" is not used to make the employee feel more important—i.e., sanitation engineer—this person is a trained engineer. Then there is the purchasing agent. None of the people mentioned thus far are involved with grinding the poor sub-assembly salesman into sharpening his pencil to the nub. Trek, like any successful company, employs "pros" to do that. This differs from ABG, where Menown will probably hold a fair bit of sway over the purchasing process, whether it be outsourced complete bikes or parts.

Interestingly, Trek's marketing department oversees all issues having to do with paint, graphics and model names. A product manager or project manager may not know what the bike is called, or see what it looks like, until after a prototype is finished. This is also the norm at ABG.

Trek has a separate quality (testing) department, and the larger companies like Trek and Cannondale have extensive testing facilities. Finally, the finished bike goes to Trek's assembly engineer—again a real engineer—who will calculate the size of the box needed for the bike, decide if the packing materials are adequate to withstand the forces of shipment, how long the cables need to be, whether the brake levers are properly shimmed for the size of the handlebar, and so forth. You might say that the design engineer is the "input engineer" and the assembly engineer analyzes the "output" and makes required changes in the bike, its parts, or anything that relates to how the finished product operates.

An example of how things can change from the "input" stage to the "output" is in the case of Trek's initial WSD bikes built in the late '90s. These are women-specific road race bikes, and in the smaller sizes have quite steep seat angles (up to 76 degrees). Trek found that the bikes weren't shifting properly, and they needed a front derailleur bracket that placed the derailleur back to a more normal 73 (or so) degrees. I was running Quintana Roo and Merlin back then, and QR in San Marcos was making a custom clamp-on front derailleur bracket for Merlin's new Aerial tri bike. It was a complicated piece that was made on an expensive Fadal horizontal CNC mill bought mostly to make this particular part. Trek approached us and wanted to buy the piece for its steeper WSD bikes, but we quoted them a price that was too high for them to pay. They ended up performing a secondary machining operation to an existing Shimano front derailleur bracket, and it solved the problem for them. The story just indicates how bikes can be drawn up a certain way, but don't always operate as planned (I'm sure Trek's assembly engineer has many such stories, as would his counterpart at any bike company).

One final word about product managers. They are the most likely to be wined and dined by component, frame and sub-assembly factories. Sure, you can ply Trek's John Burke or Cannondale's Scott Montgomery with food and drink and yes—they're their companies' Big Kahunas—but they'll still defer to their product managers when it comes to spec and factory decisions. Joe Vadeboncoeur is therefore one of the most powerful figures in the bike biz, as is Bob Margevicius at Specialized, and their counterparts at C'dale, Giant, Raleigh and elsewhere. The only people weilding as much power are the Taiwanese agents, and even those inside the bike "beltway" do not on the whole realize how much power they have, and how much wealth they create and transfer. More about them in another installment. (Note: Margevicius, aka the "king of spec" in the bike biz, has within the past year moved over to head up purchasing for Specialized).


What I'm going to write about Taiwan will not only be news to you, but to 95-percent of those in the bike industry. Only those who've actually spent significant amounts of time over there, looking under the hood of that country, understand what's going on.

I don't care who you are, or what your preconceptions are about Taiwan. The reality of Taiwan would blow you away. If I wanted to build a space shuttle or a nuclear missile, and I had my pick of cities in which to build it, I'd probably choose L.A. But if I had to simply build a whole spitload of reasonably technical and intricate widgets, I wouldn't pick L.A., or Chicago, or Pittsburg. I'd take Taichung. And not just because of pricing. Taichung, with its 1 million people, might simply outproduce (if you measure output by weight) L.A. and its 10 million people. This city is styled an "education and cultural" center, but as far as I can see it's built around the idea of producing a maximum amount of manufactured stuff. The city is like a huge mitochondreon. It's one big powerplant.

You'd never know it by looking at it. It just seems like any reasonably dirty city in any emerging country. Choose any sidestreet in Taichung, walk halfway down, and ignore the fact that there is no fancy corporate sign outside announcing that fat blowhards sit inside behind oak and cherry desks, the way it would be in America. Just open the door, walk in, and be prepared to be bowled over by rows, and rows, and rows, and more rows, of CNC mills and lathes. And then walk into the adjacent room and hear the pounding of a half-dozen giant forging presses.

Next door might be the tubing mill, with a dozen draw benches and a room full of CNC tube benders. Down the street from that guy is the extruder. Or perhaps just a small shop with a few machines.

And so on.

Impressive as all that is, what ought to scare the bejeesus out of "high tech" artisans in any first world country is what Taiwan can do with composites. I've toured the cream of the crop in composite factories in the U.S. When I go to Taiwan and tour those factories, they look exactly the same, except three times the size. Instead of one million-dollar walk-in autoclave on the floor, there are three our four. Funny thing is, they sit unused. Why? Because they're used for curing commodities like carbon golf shafts, which are now so easily made that even Taiwan can't make them cheaply enough! It can't compete with Mainland China!

But the Taiwanese aren't complaining, because they own many of the factories on the mainland as well. Your Kestrel Talon, for example, is built by a Taiwanese company, but not in Taiwan. It's built in that company's mainland factory, as are many of the carbon forks you're riding, including many which are made by American composites companies.

What Taiwan lacks is U.S. and European design sensibilities. Kestrel's Asian contractor can make fifty times the Talons that Kestrel can make in it's Watsonville, California factory, and it can make them for half the price. But it could never have made the first Talon without Kestrel's design expertise.

There is one company in Taiwan, however, which bucks the trend and it is Giant. While I doubt that anyone from Trek is interested in commenting for the record, I'd be shocked if this American powerhouse doesn't see Giant—not Cannondale or Specialized—as its chief long-term competitor. Giant is not just a contractor. It has its own ability to invent and streamline processes, and it's got just enough ability to keep pace with its competitors on the design side. It's telling that so many start-ups in Taiwan—like Kinesis and Topeak—are headed by ex-Giant execs. In fact, in Taiwan it's termed "Giant U"—as in Giant University. When you hear that a certain firm is headed by graduates of Giant U it means that they've come up through the Giant system and were employees at that firm, where they learned how to compete.

All that is what makes Taiwan impressive. What has happened in the last five years, however, is that Taiwan has become an absolute juggernaut, and ironically it's because Taiwan was almost put out of the bike business. The downturn in the MTB market which started a few years ago, combined with the upsurge in the ability of mainland factories to produce bike commodity products, meant that Taiwan bike factories were working at only about half capacity at best. Then, all of a sudden and with no warning, Schwinn went bankrupt, leaving Taiwan firms owed as much as $10 million per firm. Schwinn, the mainland "problem," and the MTB downturn was a triple blow to Taiwan, yet it was all forseeable. The same thing happened in Japan fifteen years ago, when cyclical market forces combined with the emergence of Taiwan put so many Japanese firms on the brink.

As a result, many Taiwanese companies realized that if they wanted to keep from folding their tents they had to do one of three things: 1) They could become brands themselves, like Giant, Topeak, KHS, Race Face, FSA, Titec, and many others, and in so doing cut out a layer or two of distribution and protect their margins; 2) They could invest in, or start up, mainland factories, or even factories in Vietnam or South Asia; 3) They could upscale their technical abilities, which is how Shimano survived the crisis in the bike industry in Japan.

Point three above explains why, within the short span of three or four years, Taiwan has become adept at making intricate things like cranksets (trust me, it's very hard to make a bicycle crank), and to work very adeptly in composites. Trying to get a Taiwan company to make something better than it wanted to make was very tough ten years ago. Dragging them up-market, while they're kicking and screaming, was a chore and was usually a fruitless exercise. Not so now. Nothing like an empty bank account and an idle factory to motivate someone.

I write all of the above to explain to you how important Taiwan is, even to a "made in the USA" company like Trek, Cannondale or ABG. Take C'dale as an example, which alone among the three companies makes all its frames in the U.S. I have frequently visited small, back-alley aluminum forging companies in Taiwan and noticed, lo!, a Coda crankset popping out of the forge and getting its flash knocked off. C'dale is desperately dependent on Taiwan for its business, since it owes so much of its margin to the making of its own components. Likewise, many or most of Trek's Bontrager stems, seat posts, saddles, rims, hubs, etc., are Taiwan-made, and it's like that with every company's logo'd bike bags and handlebar tape, and the no-logo stems and handlebars that go on entry-level bikes.

If you're going to get your bikes, or your parts, made in Taiwan, you've got to have a presence over there. If you don't, you're just asking for trouble. Many Taiwan factories will make you a first article and yes it'll look great, and then in production they'll make unmitigated crap for you. You've got to employ a babysitter, and I suspect that would be true in any country. Furthermore, even with the best intentions a factory will make a mistake. Often, when mistakes are made in manufacturing, they are not made singly. It's like when a strand of DNA falls prey to a random mutation. Not just one strand of protein is malconfigured. They're all malconfigured, as long as that mutated DNA keeps on churning out product. You need an agent to watch the process, and that agent needs to be there in person, and it needs to be an Chinese-speaking agent.

You may own your own agency. Trek and Ritchey have company-owned firms over there tending to their business. The great majority of western companies contract with independent agents, however. These are Taiwan-owned companies manned by Taiwanese, and they take (let us say) 5% of the cost of the product as their fee. In theory.

One of the "benefits" to having an agent—one of the "duties"—is for the agent to find you your factory. Yes, this is necessary. You'll find, though, that lo and behold your product, which you thought was being made just fine, is now all of a sudden being made by a different factory. Why is that? "Because I just didn't like the quality that was coming out of the old factory," your agent may tell you. "I had to reject too much." Well, shucks, okay, you're the agent, you know best.

Perhaps this is exactly what was happening. Or perhaps another factory promised to pay the agent an additional 20% on top of the 5% the agent was assumed to be making, whereas the first factory drew the line at 10%. Or perhaps the new factory gave a particularly aggressive and attractive price on some job entirely unrelated to yours if it could also get your business—without, of course, you being a party to any of these negotiations. You'll never know, because the way it works is that you don't pay the factory for your product. You pay the agent and he pays the factory. Or you pay the factory, and the factory rebates to the agent. You never pay the agent only his cut, so you don't really know what the agent gets paid.

The main reason Trek and Ritchey have their own company offices in Taiwan—though they may not say so—is in order to break what one executive termed to me the "Taiwan mafia." Personally, though, I've come to view all this differently. If a manufacturer really considered what 5% of his business amounted to, he'd wonder why a Taiwan agent would work so cheaply. The truth is, he may not. What we may term payola or mordida is what many of them have to do to earn a living. There are agents—not very many—who're up front, and will say, "I can't do this for 5%, I need 25%." The only difference is that this agent is telling you the truth. Your pricing may not be any different whether you use a 5-percenter or a 25-percenter. If you've got some piddly job you want done, any agent you use might be a "25-percenter" (one way or the other) and I wouldn't blame that agent.

When I first went over to Taiwan, I brought my American sensibilities with me, and I injected them into my Taiwan dealings. I would have none of these shenanigans. I don't think I solved anything that way. Nowadays, if I were to do any business over there I'd go with the flow and, like sausage, I'd just enjoy the product and not worry about how it's made.

Perhaps my impressions were unique, and I've mischaracterized how it works over there. It seems, though, while chatting up my contemporaries at other companies in the bar at the Taipei Hyatt across the street from the convention center where Taipei hosts its annual bicycle trade show, that my experiences were typical.


There are two kinds of factories in Taiwan: those which make parts for the bike biz, and those which just make parts. It is eminently preferable to find the latter, and the best agents in Taiwan are those which can ferret out a factory that can forge, machine and polish a hub shell; or extrude, draw and butt a #7005 tube, without knowing that it's for a bicycle.

The latter is important because there are economic precedents associated with the bike industry. If you go to a crank maker and ask that shop to forge and machine a crank, and broach and coin the square-hole, you're going to pay a certain "industry" price. Likewise, if you go to a stem maker and ask for a stem, you'll pay for a stem. If, however, you find yourself a good quality 3-D forging factory that can forge you a rather complicated tube, that factory doesn't need to know how a stem works in order to perform that job.

What you'll need, however, is your own engineering capabilities, your own testing facility, and your own CAD-CAM-literate designers. This factory can make you your stem, but it can't tell you if it'll break. If you need your factory to tell you that as well, you'll have to get your stem made at a factory that makes bicycle stems, and you'll pay a premium for that.

Likewise with any other process. There's the bike industry price, and there's the not-bike-industry price, if you know where to look. This is where hard work and pavement-pounding comes in. I've been lucky to know certain people who were good at pounding the pavement and finding the best factory and the lowest price.

This dynamic doesn't work with frames, however. You've got to get your frames made at a frame factory, because bike frames must be incredibly precise. It always makes me chuckle when I see somebody come into the bike biz from some "higher" industry, thinking that because he's been making car parts or tank turrets or airplanes or speedboats it's just got to be simpler making bikes. What these people don't realize is that the tolerances one must hold in this business are at least as close as as in just about any other industry. Consider this: I (when I was in the biz) would reject cranksets if the total runout exceeded one tenth of one millimeter at the large chainring. That means the bottom bracket, and the squarehole cut into the crank, must be absolutely perfect, and the crankarm absolutely straight. It means the metal must be sufficiently hard at that joint so as not to deflect under the heavily leveraged load applied at the pedal. Very, very hard to do.

Likewise, frames are hard to make, if you want them to be straight. It's very hard to make a frame in which the rear wheel is centered inside the chainstays, and centered underneath the rear brake hole, and where the front and rear triangle are true to each other, and where the head tube is parallel to the seat tube. It's so hard to do that, that I've seen manufacturing companies that make very precision car parts just throw up their hands after years of trying to make frames and say, "It can't be done."

That's why cycling is its own industry, and those who do it well can command a premium. The smart product manager or agent knows what it's safe to have have manufactured at a non-cycling facility, and when you've got to go to a factory that knows the bike biz.

Probably even more intricate than a frame factory is an assembly factory. Assembling a bike isn't easy if you need to do it in a hurry. I've got a full bike shop in my garage, and it'll still take me two-thirds of a day to build up a complete tri bike. I don't know how many man-hours go into assembling a bike in a Taiwan assembly factory, but I'd guess it's measured in minutes. I wouldn't be surprised to know that a bike is assembled in 15 minutes, though those "minutes" are spread out over ten or fifteen different workers along an assembly line.

When I told my friend Steve Hed that I was writing this series, he laughed and said, "I wonder what the environmentally-minded among your readers will think about cycling after you tell them about Taiwan!" Of course it's "green" to ride bikes in place of cars. Making bikes is another story. It's a dirty business. Certain rivers in Taiwan are absolutely sterile of life because of what factories dump into them, and the bike factories are no exception. It's sorry to see, because Taiwan is, in its natural state, not unlike Hawaii. It's a tropical paradise with mountains that reach 13,000 feet into the sky. But when I go running in Taipei I have to take a taxi 2000 feet up Yanming Mountain, just outside of town, so that I can run above the thick belt of smog.

That's frankly one of the allures to having your bike made in Taiwan or Mainland China. Your paint is on that frame for good. There's no way any shop in America would be allowed to use the sort of paint, or the processes, that are used in the Orient. What it does to the environment, or to the workers who make the product, is another story.

But it must also be noted that the investment in bike-building infrastructure is much greater there. There is a terrific amount of automation. The sorts of robots you see in film clips of auto factories are in place in Taiwan, making bike frames and parts. It would be rare to see anything like that in the U.S. Bike building in this country is limited to the realm of the artisan, with the exception of small factories and the very few large ones, like what you'd see in Trek's Waterloo factory, or where Cannondale makes its bikes in the Eastern states.

Perhaps you're getting the picture that many bike "manufacturers" don't do very much manufacturing. While Fuji bikes are made at Fuji's Japanese factory, and KHS's bikes are made in KHS's Taipei factory, and while Litespeeds are made at Litespeed's Chattanooga factory, Specialized doesn't, to my knowledge, own a factory at all (outside the few bikes it makes stateside, which are a small fraction of what it sells overall). I could be wrong about that, but if I am, it's only in recent years that that's changed. This is no slam against Specialized. Nike doesn't make footwear. Neither does Reebok. Most of these companies are just design houses, with sales and marketing offices, and smart CFOs that keep track of the money in and the money out. Those who actually make bikes have, in the main, never heard of Lance Armstrong or Eddy Merckx, and in any case they don't live in towns where recreational cycling would be advised.


How important can this step in the process be? In fact, it is often the assembler who handles all the important money transactions. If you're a bike manufacturer in the U.S.—that is, if you're the sales and marketing office that owns a "headbadge" and you have your bikes manufactured in the Orient—you either pay your agent for your bikes or you pay the assembler directly. In either case, the assembler gets the money. Then he pays for all the parts, and he pays the frame company.

These are pretty important people, because they make spec recommendations that are taken very seriously by product managers. There's nothing worse than having your bike all ready to go at the assembler, and then you can't take possession of the darn thing because the stems didn't show up on time. Therefore, when the assembler says, "Better if you spec Profile Design or Bontrager, because [the other stem maker] is an unreliable deliverer," that carries a lot of weight.

Assemblers have their arses on the line. I do believe, if I remember right, that both Merida and FritzJo lost a lot of money when Schwinn went bankrupt year before last. In particular, FritzJo's American arm, Omnium, was hurt pretty badly according to what I heard.

Speaking of Omnium, this is an interesting company. Assembly is a pain in the rear, it's very hard to do. One reason why it has been difficult to make a complete bike in the U.S., and that it's desirable to make one in Taiwan, is it's hard to get the parts hung on the frame for a decent rate. Omnium is an assembly factory in San Luis Obispo, California that has really been a godsend to mid-sized manufacturers, in particular some of those in the sport of triathlon.

If you've got a U.S.-made frame and you're not doing very many units—say, a thousand to four thousand a year—you may not want to take your project offshore. But that's too many frames on which to hang parts here in the U.S. But it's not enough frames to justify building an assembly line. So you send your frames to Omnium and have the parts hung there, and they'll then pack the frames and send them to the retailers. They'll even warehouse them for you.

Specialized got to having more and more frames built in the U.S., and it invested in its own assembly factory in Utah. Trek of course does its own assembly in Wisconsin, and Cannondale assembles its own bikes. But it's rare to find that one company that can be used as an outsource vendor, and Omium is it. I don't know of another in the U.S., although I'm sure others exist.

The only other option is to push the process off on the dealer. Quite a few small manufacturers have an agreement with a parts distributor for a modified OEM arrangement. One New York-based parts distributor called Security Bicycle Accessories pioneered this. You'd call them up and say, "Hey, I sold a frame to Mission Bay Multisport, send them a 'tri kit' today, so that the frame and kit show up at the same time." This is the way it works in theory, but in practice sometimes details fall through the cracks. I don't know how many times dealers have said to me, "The frame is here but it's thursday, my UPS has already come, and the customer is coming in expecting his bike on Saturday morning." Perhaps SBA didn't ship the kit. Perhaps the frame maker forgot to order it. Fortunately, UPS is always a convenient entity to blame. Either way, it's the dealer who has to white-knuckle it.

That's why dealers prefer to have the bikes come into them already complete. But not every dealer is like that. In fact, I've come across some dealers who disassemble and reassemble every bike from the mid price range up, because they don't trust anything but a pro assembly job.


End users see bikes on the showroom floor of their LBS, or in the pages of mail order catalogs. They don't see the route these bikes take to get there.

There are several ways bikes, or anything your LBS carries, can make their way from welder and painter to you. As the business has gotten more competitive some of the dynamics have changed. But in all cases some form of distribution involved, and the only difference is whether the manufacturer does it himself or whether he relies on a third party to do it for him.

If a distributor is used, the manufacturer will give up a lot more margin than if he uses reps. Therefore, if a company feels it can reach a significant amount of customers by itself, it'll forego selling its product to distributors. Where distributors become almost imperitive to you as a manufacturer is if you're trying to reach retailers in a country other than your own; or if you make a smaller, ubiquitous commodity that needs to be purchased by a very large number of retailers—a much larger number than you have the ability to contact.

A manufacturer like Topeak, for example, is in both categories. This is a Taiwan-based company that needs to sell virtually all its products in some country other than its own. It also needs to sell to a lot of retailers. It might have its small bicycle tools in 2000 retail stores in the U.S., whereas Kestrel might have its bikes in around 200 U.S. stores (Kestrel needing to be pickier about its retailers).

Topeak uses as its distributor a company called Todson, owned by a gregarious East Coaster named Neil Todrys. Todson is one of a rare breed of distributor that also includes Reno, Nevada-based Sinclair Imports, which imports and distributes Carnac cycling shoes. These distributors are what I call "brand builders," because they do more than simply warehouse and ship product. By definition, they are good at building your brand. Deda's components were first introduced into the U.S. market through Sinclair, and while Carnac has given its loyalty to Sinclair in return for having that distributor build Carnac into a robust trade name in the U.S. (against such competitors as Sidi, Nike and Shimano), Deda was, in my view, less appreciative of the good job Sinclair did.

Deda is therefore also now sold by the other kind of distributor, perhaps best exemplified by Quality Bicycle in the Twin Cities area of Minnesota. Steve Flagg has built his company from the bottom up into a juggernaut of efficiency. He is not a brand builder (in my view), but he's just as important to the industry. He's a master of having what you, the retailer, want, and of shipping it on time, and selling it at an aggressive price.

If you're a niche, high-end company, however, like Serotta or Zipp, you might find that you're better off without a distributor between you and the retailer. You might be better off with representatives. These people take a smaller margin than do distributors—like only a fourth to a third as much. This is because distributors buy the product from manufacturers, and warehouse and ship it. They use their own money. Reps are like real estate agents: They don't use their own money to "warehouse" your house, they don't "deliver" your house, and they don't buy and and re-sell your house. A bicycle rep's only expenditure is in the soft costs he incurs, chiefly in traveling from shop to shop.

These are independent reps. They are self-employed, and they'll have anywhere from four to eight "lines" that they represent in a territory, which might be as large as several states, or as small as a couple of Southern California counties. Typically it's the apparel, footwear and accessory companies that employ reps, though a very few high-end companies that make high-priced products have become masterful at gaining and using an independent rep force, and Serotta and Zipp are two of them.

Larger manufacturers, like Trek, have company reps. These are employed directly by Trek, as Trek doesn't want its reps to do anything but sell its product. I feel the cut-off point in the bike industry for employing company reps is $100 million in annual sales. That means that Trek, Specialized, and Cannondale—and maybe Giant—are the only bike companies in the U.S. that can afford company-employed reps (if my formula is any indication).

One reason you don't see as many bike manufactures making headway in countries other than their own is the sheer economics of it. Forget the issue of import tarifs, and even shipping. Having a distributor inbetween your product and the retailer adds a huge multiplier to the retail price. If you wonder why Colnagos and De Rosas are so expensive in the U.S., it's not just because of the artisanship. It's because of the extra layer of margin, which adds several hundred dollars to each bike. Unless you've got enough muscle to come in and set up shop with your own distribution in a foreign country—as Trek and C'dale have successfully done in Europe and Canada—exporting a low-margin product like a bicycle is tough.

Not so, however, an image-driven product that has a low cost basis, or a product that might have high development costs but high gross margins. An Oakley sunglass, for example, or a Giro helmet, might have a high cost associated with the development and marketing of a model, and for the cost of the mold, but once you finally get to the point of flipping them out of the molds like Big Macs it's not difficult to absorb an importer's margin and sell them into another country.

When you realize the dynamics of how and where a product is manufactured, and the class of product you're talking about, you can understand why a certain manufacturer might enjoy a robust overseas market, while another is locked into a chiefly domestic customer base.


Bicycles are devilishly hard to engineer and test, both for performance and for safety. On the performance front, you soon see after spending a bit of time in the wind tunnel that the forces and vectors and all that stuff are unpredictable by simple virtue of the application of mathematics. You affix little strings to points on the bike and you see that in several places the wind actually points in the same direction the bike is traveling! Rolling resistance and wheelsize? Just as hard to test for.

You'd think that testing for safety would be a lot easier. You just put a fork or stem or complete bike into a testing machine and it either breaks or it doesn't, right? Thinking that, my company bought a $10,000 bike testing machine from Taiwan—state of the art—what the other companies use—problem solved.

The first bike we tested was a new model of Merlin off-road bike, because we'd been experiencing failures in the machined unistay in the back—an aluminum piece that clamped the seat stays below it and a single unistay above it. We put the bike in the machine to see how many cycles our new, revised clamp would last. The chain stays broke first. Great, problem fixed. Except afterward we put the old clamp in, and again the chain stays broke first. Regardless of what clamp we affixed to the seat stays, the chain stay broke first. We never could break a clamp. It was obvious that the machine didn't apply forces to the bike the forces are applied "in nature," because Merlin never did break any MTB chain stays in this model. Only the seat stay clamp.

This is what makes testing and engineering bikes very hard. You never know what is going to break. That is why it is only of dubious value to hire an engineer to do your engineering, if that person is not familiar with and experienced in bike manufacture. Einstein couldn't calculate the forces on a bike. I'd rather hang my hat on a guy who's been making bikes out of various materials for fifteen years then a rookie engineer just out of college. Of course it's nice to have an experienced bike designe actually be a trained engineer, and there are a few of those out there, like the Cervelo folks and the maker of my own personal bikes, Ves Mandaric. But when push comes to shove, I'll take Keith Bontrager, who to the best of my knowledge isn't a college degreed engineer, over just about anyone who does have a diploma hanging on his wall.

There are some companies that really do go the extra mile in testing. I suspect Cannondale is the best among the big bike companies. If you're a parts maker who wishes to have your product spec'd on a C'dale, your part really has to jump through hoops. Not only does C'dale put your component through all sorts of structural tests, it has an oxidation test as well. If your part'll rust too fast—rejected.

Reynolds Composites allowed us to photograph a typical fork testing machine, which uses a protocol quite similar to others in the industry. Reynolds does a prodigious amount of testing both on its forks and on those of its competitors. The fork being tested is pictured on the left, with a piston across the bottom of the photo that pushes the fork blades, deflecting them what looks to be about a centimeter or so. There is a readout on the top center of the photo that shows the amount of deflection per cycle, and the counter on the top right shows how many cycles the fork has gone, not having yet failed. Once a certain amount of deflection is reached, the fork (or bike, or component) is deemed as having reached "failure."

Certain companies set a standard for testing, and it's nice when the industry can agree on a protocol, because then you know that your new stem (if you're a stem manufacturer) is safe because it's gone 300,000 cycles before it failed, and 150,000 cycles is deemed safe. Unfortunately, there isn't that much commonality in protocols. I know fork makers, for example, that just test a lot of their competitors' forks, and their way of assuring that their fork is safe is to make sure it lasts longer than the other forks in the industry.


Funny thing, I'm helping a brand new company with some advice on how to get going with Asian manufacturing. When they read this final installment they may think twice. Admittedly, though, I chose what is certainly the sexiest, juciest example I've come across in a while. Most of the time things go more or less well. But there are exceptions.

Like the time I had bikes made overseas and did so without the aid of an agent. I flew over to watch my production. I was there for a week—the week my first production of bikes were slated to be made. The bikes didn't get made. Each day there was another excuse. "The chainstays didn't show up. The derailleurs didn't show up." I left without watching them weld a single frame. I got the bikes some weeks later in America. They were awful. I rejected more than half. That's the way it goes, or can go.

In the case I'll write about today, Hed cycling made a handlebar. John Cobb was employed to aid Hed in the design and testing of the bar. Cobb made several trips to both Europe and Taiwan in helping get the bar going. Cobb also had other products on which he was working, including a new fork.

Meanwhile, a longtime friend of Steve and Annie Hed, Morgan Nicol, had just left after a long tenure as the European head of Ritchey. The split was acrimonious. At the same time, Ritchey went in another direction concerning its Taiwanese agency—it decided to hire its own people and set up its own captive office in Taiwan. Ritchey's just-abandoned agent, Joseph Chao, went into partnership with Nicol, forming a components company called Oval.

Steve Hed had never done much manufacturing in Taiwan before, and needed an agent in Taiwan. His friend Nicol recommended Chao.

Nicol was also slated to be the Hed's new European distributor for its wheels, and for other products on which Steve Hed and John Cobb would collaborate. All was friendly, one big happy family.

Just a few weeks before the Interbike show, however, the Heds come to learn that their carbon aero bar is no longer theirs. It was to be an Oval product. They understandably felt betrayed by their Taiwanese agent, Chao, because they'd been dealing in good faith with Chao, and had sent money to Chao for Chao to pay the factory for tooling costs. The Heds did not have very much leverage when trying to appeal directly to the manufacturer which made the Hed's bar for them, because that manufacturer's contact was Chao. Plus, agency relationships with manufacturers are strong, since it is common for an agent to have other products manufactured in that factory. Simply put, the Heds felt they'd been outmaneuvered.

What about Nicol, Chao's partner, and Hed's longtime friend? Things had soured a bit after an agreement couldn't be reached for distribution of Hed's products in Europe. The Heds would not be able to appeal to Nicol for help. It must also be said that Nicol had a set of grievances against the Heds which revolved around how the European distribution deal for Hed wheels fell apart.

I'll not get into the particulars of that, nor take sides. I write this just to show you how things can end up, and it happens with every company, big and small, at one point or another. Things went sour with Cervelo (several years ago) and a frame vendor. It happened with Trek and Rolf Diettrich. There are many examples I could cite. Sometimes relationships don't work out. Sometimes things end up in court. Generally when they do, both sides lose.

This one has a happier ending. The Heds were able to reach an agreement with Nicol and Oval at Interbike, less than a week ago as of this writing. The Heds got their aero bar back. Oval had a very good showing of its (Cobb's) fork, of its aluminum version of an aero bar (not unlike the Deda bar Lance rode in the Tour), and of its other products. Oval will in the future very likely be a force in the components business. John Cobb remains friends with all sides. Hed's bar was one of the hits of the show.

The moral of the story is, there is no moral—there are no morals. Or ethics. Or friendships. None that you can count on. Not when you're getting a product manufactured. There's just protection, and you have to make sure you have it.