History (1996): Interview of Richard Gruet, Floppy Disk Globetrotter

His job is like no one else’s in the world. He is hired by major floppy disk purchasers to inspect factories in order to verify the quality of the magnetic media being produced. Richard Gruet, a French national, has toured over 50 diskette plants, crossing Asia bit by bit, loaded down with some 150 pounds of inspection material. We asked the floppy disk globetrotter a few questions.

Computer Data Storage Newsletter: Your business card reads ‘Richard Gruet, Media Inspection,’ with a Hong Kong address.
Richard Gruet: Yes. I have a company in Hong Kong called Media Inspection Ltd. In France, I operate as an independent consultant.

What is your background?
I’ve worked in magnetic products for 25 years, first with Ampex, from 1971 to 1985. After that, I was a sales associate for the French firm MSI, which focused on production of quality control equipment for diskette manufacturing, obtaining some of its engineering from Rhône-Poulenc Systèmes. By the end of ’91, MSI was no longer working very well, so I decided to head into disk inspection, all the while thinking I could sell MSI testing equipment to disk manufacturers. I started my first inspections at the beginning of 1992.

Who are your clients?
The first door I knocked at was the factory of Albi’s Rhône-Poulenc Systèmes, which had a policy of buying Asian-made diskettes to supplement its own output. That collaboration lasted through mid-1993, until the European anti-dumping taxes aimed at Asian suppliers took effect.

So your customers are first of all media suppliers who can’t manufacture diskettes in sufficient quantities to meet customer demands. Who else?
There are also some software duplicators. I have worked, all in all, for something like 25 firms, such as Memorex, BASF, Rhöne-Poulenc, etc., but never for more than 3 or 4 at a time.

What exactly do they expect from you?
Most of the firms want to reduce all risks linked to quality. They tell me, ‘Okay, we just purchased a batch of a million diskettes from a certain place. Go and inspect them, and write up a report according to such and such criteria.’ Generally, these firms have placed an order and signed a letter of credit for the Asian supplier, with a quality protection clause requiring a certificate of approval from Media Inspection Ltd.

What are your fees?
At first, I was called ‘Mr. One Cent,’ because I charged one cent per diskette. That continued for a year and a half, two years. That was back in ’92-’93, when diskettes cost 50 cents. Certain customers don’t like that formula, preferring to pay me a fixed fee per inspection.

What is your responsibility in the event of delivery of inferior quality diskettes?
All my customers ask the same question. Actually, the situation has never arisen.

Your customers have never taken delivery on a single bad diskette?
Of course, some have gotten bad diskettes, but knowing full well ahead of time that they were bad. Certain companies accepted inferior lots knowing what the problem is. When I run an inspection, there are 3 ratings: 1) the lot is good; 2) the lot is completely bad; 3) all the cases in between where the lot doesn’t meet all required specs.

What you inspect, then, is the lot, and not the quality of the production line.
Exactly, I’m a guarantor. I run tests on a batch of diskettes, using a sampling of a sizable number, based on the US military’s MIL-I05D standard, which envisions 3 levels of sampling. For example, in level 2, the most widely used, on a lot of a million diskettes, 1,250 diskettes would be inspected from 125 boxes taken at random.

How many plants have you inspected?
A little over 50 in all, in China, Thailand, Singapore, Malaysia, Indonesia, the Philippines, Korea, India, Macao, Hong Kong and Taiwan. Nearly all of the countries in Southeast Asia. I traveled as much as 330,000km in a year. Once I even inspected a Bulgarian company, Magnetic Media, which is still around.

You never had any trouble entering all those different Asian countries?
Never, and yet I sometimes had more than 150 pounds of material to clear through customs. Once in a while I would get into arguments with the customs agents, but very rarely. Even when I traveled with 6 certifiers, which are diskette drives with diagnostic read-outs, and weigh nearly 80 pounds, two PCs, and a machine which verifies the track alignment for formatted diskettes.

How many diskette plants are there, worldwide?
Throughout the world, I have no idea. In Asia, there were nearly 80 at one point. But you have to make the distinction between Megamedia, Hanny Magnetics and Benelux, where more than a thousand people manufacture up to 20 million diskettes per month, and some factories half that size that produce only a few thousand diskettes. Some of the factories are fairly primitive, far from high-tech. Tomorrow any one of them could do something else, it wouldn’t matter. You could say, globally, that there are 30 to 35 veritable factories. The definition of a factory isn’t always the same as ours, except for the larger firms, and yet I have a hard time conveying that fact to my customers. Another problem I have is with the actual origin of a given batch, for example. You can never be certain that a lot of one million diskettes you’re shown have all been manufactured on site. In actual fact, they may have been partially, entirely, or not at all.

How long does it take you to audit a million diskettes?
It takes about a half-hour to get an idea of what your dealing with, given my experience. 4 to 5 more hours to have a better sense of statistics on the rejects, which is to say, to run 1,250 diskettes through 6 drives, at 30s per diskette, along with additional tests.

What are the most common problems you come across on a diskette?
There are a good dozen components to a diskette, which are becoming lighter and cheaper with each passing month. With fewer and fewer materials. I’ve seen shutters that you tear off like a piece of paper. The problems can be mechanical. Or they can be problems of form or size, for example with striped or deformed shutters, but also of badly soldered plastic, or bad pressings. Of course, the problem that most clients are interested in is for a missing pulse, which directly affects the diskette’s capacity to restore pre-recorded data without error. Not to mention other tests which the standardization bureaus established over 15 years ago. These consist of modulation, resolution, peak shift and overwrite tests. But these tests are less and less appealing, which is not surprising, given the evolutions in technology that have taken place since those standardizations, whether at the level of drives, magnet, heads or cookies. A certain number of players, including me, have been lobbying to replace the current analogic methods with a digital method (window margin test) in order to approach more closely the actual conditions of diskette use. In the past, these methods had trouble catching on, but the considerable reduction in the cost of the necessary components, along with more compact sizes, has made such equipment more widely available. Furthermore, the latest gen of certifiers incorporate the windows margin test online although the latter should be considered a complement to rather than replacement for more conventional methods. In the diskette industry as elsewhere, some are reluctant to change, but it hardly seems reasonable to hang on to the outdated methods of the past, far more costly to implement, when demand is constantly pushing the price of diskettes lower.

How much do diskettes cost now?
Coming out of an Asian factory, a brand-name diskette with printed shutter, in a 4-color box wrapped in plastic, a 3M for example, comes in around 23 cents. For diskettes in bulk for duplication purposes, in bunches of 50, around 19 to 21 cents each. In a blank box, formatted but not certified, somewhere around 17 cents. It’s better to get back a million bad diskettes at a good price than a million good diskettes at a price that is too high.

What does it mean when you get a diskette in box marked 100% certified?
Not a whole lot, it’s a commercial pitch, behind which both the best and worst can hide.

Are brand-name diskettes, 3M or Verbatim, for instance, justified at a higher cost?
Sure, since a reputable firm is behind them. That doesn’t mean there won’t be some bad diskettes. But there is certainly less risk.

Even when they’ve been sub-contracted to another company who offers them at a third of the price?
You can always find a diskette of just as good quality, even better, at a third of the price. But with the brand less diskette, in the blank box, it’s like the lottery. It could be excellent, it could be trash. When you go with the known brand, you have a minimum guarantee.

What percentage of brand-name diskettes are in fact subcontracted?
I couldn’t say precisely, but it’s significant, and becoming more and more the case. Some subcontractors do everything, up through the packaging and there isn’t any difference, for example, from those that emerge directly from 3M. The big firms don’t really have an interest in continued investment in diskettes, but are more concerned with new technologies. On the other hand, as long as they continue to operate at all in the floppy sector, they have to maintain their brand-name’s presence and the illusion of a full product line, and one way or another, continue to include diskettes in their strategy.

Why so many diskette factories in Asia?
The main worldwide diskette makers, that is to say, the first wave, made a serious mistake in their calculations. They sacrificed the long-term for the short-term. Which is nothing new. We also saw it with audio and video cassettes. They divulged their technology by selling their cookies to any and all buyers, all the while pretending they themselves weren’t arming future competitors. At the very least, they could have linked their cookies to sales licensing and royalties agreements. Sure, the Asian firms would still have tried to produce their own cookies by whatever means, but given the investment required, the number of new factories would have been severely limited, and you would have seen diskette manufacturing in conditions that were far less appealing. In that respect, you have to give 3M their due, they’ve at least had the most coherent attitude. Unlike other firms, the Saint Paul-based company always had a reputation for model management.

You don’t think the low cost of Asian labor has played a significant role in the widespread delocalization of diskette production?
You need to distinguish between different manufacturers. Certain plants have dragged and continue to drag their entire history with them. Despite some belt-tightening, they still have too many personnel, along with backwards or at least disorganized production facilities. In Europe, the Italian operation CSI can be taken as a model. They came along after everyone else, with everything to learn, but without the same heavy past to lug around, and look at them, they’re among those doing best in the sector. In Asia, to get ahead, or even to survive, you have to focus on what’s cheapest, i.e. labor, and in the countries where it’s cheapest. Above all, you can’t favor automation machinery that’s costly to acquire and maintain. Of course, this approach assumes you’ll back it up with excellent organization and solid management. In that respect, one of the best examples I know is that of Benelux, a Hong Kong firm with a delocalized plant in Indonesia, which can manually produce 20 million diskettes per month. With the low pricing of diskettes, it’s just not feasible to install automated production lines, and far too risky to go with used equipment. The only process that’s fully automated is that of the center core assembly on the cookie. Yet too many diskette purchasers allow themselves to be impressed by factories with a high-tech look. There were, by some accounts, a certain number of diskette factories in China, a large number of which have since disappeared.

What happened?
I never got confirmation of the figures, but as far as actual factories are concerned, there were never more than 40 or 50, many of them at Shenzhen. You also have to be careful about what people mean when they say a factory shut down. Just because a given plant is no longer producing diskettes doesn’t mean that it no longer exists. In Asia, factories are often located in buildings where there are other business activities of the same company. When a new order comes a long, the factory may start turning out a different product. It’s true that China, however, has suffered in the face of European anti-dumping taxes.

Since you mention it, what exactly has been the various effects of these taxes?
The first of them dates from May 1993, and affected Hong Kong, China, Taiwan, Korea and Japan, with special terms for certain factories. Naturally, the countries in question were severely affected. Until that time, the companies had managed to accumulate considerable battle reserves with their 50 cent diskettes. They then tried to dodge the tax by investing elsewhere, outside the range of the tax, in Indonesia, Thailand and Macao. Things have been working that way for about a year, a year and a half. But these taxes haven’t at all affected the evolution of the market’s pricing. (I)

Do the taxes do anything, in fact?
In any case, they don’t serve their intended purpose. They haven’t kept the prices from falling, nor have they slowed the tide of Asian imports. I was personally surprised by the amount of resources the European institutions allocated to this cause, given the fairly minor European interests they were protecting. What limits the playing field for Asian manufacturers is not taxes, but lower prices. Thus, Far East business keep easily one step ahead of the bureaucrats in Brussels. Not to mention the questionable practices of those who instigated the taxes, European manufacturers, who dream that some day they will be applied to everyone but themselves. It was they, in fact, who contributed to the prosperity of Asian suppliers, by buying massive quantities of diskettes from Asian manufacturers when they couldn’t meet growing demand on their own. At the same time, they turn and demand protective taxes from Brussels. At the very least, they could have confined their demands to something more reasonable, since the current unrealistic charges of 40 or 50% can only encourage delocalization.

Who are your main competitors?
Companies that run industrial inspections, for example Veritas, or SGS, a large Swiss concern based in Hong Kong. I’m not even sure if SGS is still in this activity. I have a sort of monopoly, but on very little.

What will replace the diskette?
The market will continue to grow as long as we still see diskette drives on PCs. I think it’s a bit like Philips’ audio cassette, a product which has grown old and outdated, but which will never quite be extinct, whether by force of habit or because of the enormous installed base. I don’t believe in the future of high capacity diskettes, because they require a shift to servo-positioning technology, which is very complicated with flexible media, particularly in terms of guaranteeing interchangeability.

So you think you have a future in your current line of work?
In any case, not with diskettes. What’s killing my particular activity is the ever-declining value of the product. Today, it’s virtually treated like a piece of paper. You get it, you throw it out. Which means, ultimately, people are going to be looking less and less to quality control before sale. I think I’ll manage to evolve, and stay with the same kind of work for the CD-R, once it attains a certain maturity.

(I) The European Commission has decided to end a 9-month investigation of an anti-circumvention case in Hong Kong and 8 other countries. 24 Hong Kong companies were suspected of falsely labeling their 3.5-inch floppy disks with different countries of origin in order to avoid the high taxes imposed in China and Taiwan.

This article is an abstract of news published on the former paper version of Computer Data Storage Newsletter on issue 104, published on September 1996.

Note: It was the first time I met Richard Gruet who began to work for StorageNewsletter.com from 2010 as advertising manager and also then as web master. (Editor)