What You Learn When Your iMac Goes Down

We recently had the hard drive in our iMac go out.

Drive Reliability

See this link for a very interesting article on drive reliability

http://www.pcworld.com/article/129558/studyharddrivefailureratesmuchhigherthanmakers_estimate.html

This article reinforces what we have experienced first hand — the MTBF numbers produced by drive manufacturers are false. Carnegie Mellon’s lack of differentation among vendors in this study indicates their research was likely polluted by vendor pressure and or contributions.

What we learned is that iMacs are not designed to be serviced by users. The design of the iMac looks great, but has a very strange assembly that makes it even more difficult to work on than a laptop. The iMac has not screws or other fasteners on the case (except on the bottom for memory replacement). A hard drive is a major sub-component in a computer and tends to be one of the more problematic. It is something that not only should be designed to be easily replaced but should be designed to be swappable. As with media like CDs, there is no reason a door could not be added to any computer, and different hard drives could be added and removed to give the user maximum flexibility in booting to different drives. With a spare drive, this would mean that no computer could be brought down due to drive failure.

“Swappable” drives have been used in servers for some time, and are now available for home disk centers (which allow for RAID configurations) such as the Acer model above.

However, while no personal computer actually makes it as easy as we think it should be, Apple has designed a case with no entry through the back, so the user or service technician must actually pull off the glass cover with a suction cup and remove the display (delicately) to expose display. Next the display must be removed to reveal the hard drive. Several specialized tools are required for the task. Waiting for tools to arrive from eBay, as well as the Apple Store’s $420 quote for the work, is why our iMac is sitting unused right now.

Long Term Trend

This is part of a long term trend in consumer items to hide the fasteners in order to increase the “coolness factor.” This trend extends to a number of different categories. If one looks back to the cars of the 1930s, one can see that they were more modular, and the rivets, pins, screws and other fasteners were more apparent. What this meant was that cars were more serviceable.

The Bentley Speed Six was a very serviceable car. The engine was easy to get to, the fenders were easily replaceable, and the exposed fasteners allowed the replacement of many parts by shade tree mechanics.

By the 1950s, almost all cars had moved to integrate the trunk and fenders into the body, and fasteners were no longer observable from the outside. This resulted in a smoother look, but also in a more complicated design and more expensive automobile to work on.

The 1950s Cadillac Series 62 was representative of cars from this era, in that it had an integrated body and hidden fasteners.Bodywork on this type of car is more time consuming and expensive and must be done by professionals.

Serviceability Trend

The long term is to decrease the serviceability in items. While this may be good for corporate profits, it is actually bad for consumer and bad for the environment. The more difficult and expensive it is for items to be repaired, the more quickly they are simply replaced by new items. The problem is that companies do not seem to have an incentive to build long lasting and easily serviced items. The finance area of the company seems to think it reduces sales of new items (which it does), and new product design and marketing seem to think it reduces the “coolness” factor of products. Marketing and finance have come to dominate US corporations, so it is no surprise that their values have become the values of American business. This is not going unnoticed. industrial designer Victor J. Papanek,
.while American products once set industrial standards for quality, consumers of other nations now avoid them due to shoddy American workmanship, quick obsolescense and poor value.

Historical View

There is this common impression which is reinforced by advertising that this year’s model is better than last year’s, and that in general we are on a on an upward slope. This is not actually true. There are many business practices and products that were “better” – better for the consumer and better for the environment — in the past. In addition to serviceability, many products were simply designed to last longer half a century ago. For instance, there is a lively market for classic toasters from the 1950s on eBay. These 50+ year old toaster still work, because they were built to last. The concept of a 50+ year old item is unheard of today.

This 1950s SubBeam is still working, and adjusted for inflation, is probably selling for more on eBay than it did back when it was purchased in a store in the 1950s. Why can’t more items be built to last and be built to be serviceable?
Noticing Changes

I suppose the question to ask is what has changed? How did American business go from offering many durable and serviceable products to offering products designed to be thrown away? Secondly, how did both American and international consumers become habituated to this new consumption pattern? Thirdly, does anyone think that this trend can actually be reversed by “the market?” Actually, it would appear that on broader goals such as environmentalism (which the life-span of products are a contributing factor towards) that the market will drive product development in the opposite direction, towards planned obsolecense.

References

One question we have is if purchasing specialized drives, such as surveillance drives — which are designed for high usage video applications are more reliable than normal consumer drives. Seagate makes an very price attractive model.

http://www.provantage.com/seagate-st3320410sv~7SEGS1UN.htm

This is an interesting article on planned obsolecense in hard drives. We quote from it below.

http://www.driveservice.com/bestwrst.htm

For a long time, I was a big supporter of IBM drives and recommended them at every turn, but now not so. They too have had enormous numbers of drives returned to them recently, and I am sure that is what spawned the Hitachi buyout. I have noticed over the last couple of years that manufacturers have stopped putting little mini in-line fuses on the electronics of the drives. I often asked myself why they were doing this, as the fuses could not cost 1/2 cent each. I have since found out! This is a little known fact that is not limited to hard drives alone unfortunately, but also incorporated into cars, electronics of all sorts, and everyday things that we the consumer use. This little known fact is called “built-in obsolescence”! This is a very little discussed problem in today’s society, but we all face it at some point or another.

The is an excellent excerpt on this exact point in the book Waster Makers, published back in 1960. We have copied it below. This relates to deliberate changes that made items less serviceable.
Beyond all these factors of quality debasement and by repairmen there were several objective factors about modern appliances that helped make them expensive to maintain and that helped increase the business volume of servicing agencies or replacement-parts manufacturers, and, in some cases, the manufacturers hoping to sell new replacement units. There were more things to go wrong. Those added luxury accessories that so delight copy writers were adding to the problems of products to break down. The rush to add extras on washing-machines in the form of cycle control, additive injectors, increased the number of things that can develop ailments. The Wall Street Journal wrote: “Parts and accessory dealers naturally are pleased with the added extras put on new cars.” They should be. I have two neighbors who bought station wagons in 1958. One bought a model with power steering, power brakes, automatic shifting, and power windows. The other—a curmudgeon type who doesn’t think that shifting gears and raising windows by hand are too much of a strain—bought a car without any of the extras. His years of ownership of the car have been relatively trouble free. (And by spurring the extras he saved several hundred dollars at the outset.) The other neighbor who bought the car with all the extras moans that he got a “lemon.” His car, he states, has been laid up at the garage seven times, usually because of malfunctioning of the optional equipment. Replacement parts were costing more. The gizmoed motorcar was a good case in point. A creased fender that in earlier years could be straightened for a few dollars was now, with integral paneling” and high-styled sculpturing, likely to cost I $100 to correct. The wrap-around windshield was likely to last three to five times as much to replace as the unbent/ windshields that motorcars had before the fifties. Ailing parts were increasingly inaccessible. In their pre occupation with gadgetry and production short cuts, and perhaps obsolescence creation—manufacturers often gave little thought to the problem of repairing their products (or alternatively made them hard to repair.) Sales Management dominated that “products are not designed for service.” It was of steam iron that could be repaired only by breaking it apart and taking out the screws. Some toasters were riveted together such that a repairman had to spend nearly an hour just getting to the right part. This is to replace a fifteen-cent or a ten-cent spring. Product analysts at Consumers Union told me that air-conditioning units in automobiles were often cluttering up the engine compartment so badly that it took an hour or two to remove a rear spark plug. Built-in appliances—which were being hailed as the wave of the future had to be disengaged from the wall before repair work could begin. Many of these built-ins were simply standard.

Manufacturers often failed to provide in provide information that would facilitate repairs. Recently The Boston Globe protested that appliance manufacturers were getting so “cozy” with service manuals that customers seeking them got the impression they were “censored as if they contained obscene material.” The Electric Appliance Service News likewise expressed indignation on behalf of servicemen, or at least independent servicemen. It said, “Our mail is loaded with gripes daily from servicemen throughout the country lamenting their inability to obtain service manuals from certain manufacturers.” Often this coziness has sprung from the desire of the manufacturer to keep the repair business to itself and out f the hands of independents. The News charged that “some manufacturers do not make service manuals available to all independent repairmen and therefore it is almost impossible to make repairs easily and properly—and at a time-saving expense.

This exerpt is from the book The Break Through Illusion and is related ho how R&D was changed to be less integrated and more specialized, and how service-ability as well as manufacturability were reduced.
R&D was also separated from other corporate activities such as product development and manufacturing. In 1925 Bell Labs was organizationally separated from Western Electric “to permit more effective specialization in research and development. “R&D now became the first step in a specialized assembly-line process of innovation. According to the historian George Wise: “At subsequent workstations long that assembly line, operations labeled applied research, invention, development, engineering, and marketing transform that scientific idea into an innovation.”

As this process moved along, projects and products would simply passed over the transom from R&D to product development, from product development to pilot production, and from pilot production to manufacturing Once a project was handed on, the receiving group vas confronted with a fait accompli, their freedom of operation constrained by earlier decisions. For example, engineers working on the body of a car might design it in such a way as to make proper placement of the engine and steering difficult. The engineers assigned to steering and motor development would then change the design based on their needs. By the end of this process would be expensive and difficult to manufacture. Typically, this yielded results that were expensive and frequently of low quality, for example, the Ford Pinto and the Chevrolet Vega, cars that were designed as lemons.“

He consequences of all this were both profound and disastrous. The connections between R&D and production were irrevocably severed. New ideas and inventions were stranded in a “twilight zone” between R&D and production. American industry went from a system in which innovation and production were closely linked to one in which it became increasingly difficult to produce the research labs’ developments economically. Some companies, like DuPont, responded to the growing gulf between the R&D and manufacturing by creating internal “venture” divisions designed to turn promising R&D into new products or in >me cases into new businesses. But few of these new venture visions proved successful. For example, none of DuPont’s major new internal ventures or spin-off companies amounted to much. The reason for this was basic: new venture divisions simply added another intermediate level to an already overblown and unwieldy R&D bureaucracy. Here again; large corporations showed that they were oblivious to the need for more fundamental kinds of restructuring—

New Information

In a previous post we wrote about the inefficiency of automotive service parts networks.

http://spplan.wordpress.com/2009/05/15/auto-service-part-networks-are-a-mess/

The line of reasoning of the article was that manufacturers were unnecessarily outsourcing the management of service parts too low in the supply chain – at the dealer level and the auto service parts could be greatly improved in their management through a national and regional system of service parts management. Furthermore, that dealers were incapable of creating effective service parts websites and that this function should be centralized as well.

Structure of Auto Industry

What we learned from the book Who Really Made Your Car, by Thomas H Klier and James Rubenstein, is the following:

• 70% of the parts of automobiles are made by suppliers
• Manufacturers are actually now primarily assemblers
• Much of the intellectual property and complex component manufacturing is owned and provided by the supplier / component manufacturers

Suppliers Actually “Make” the Car

Suppliers are producing most of the car and providing many different manufacturers with similar items. This is explained in the graphic below which provides a great insight into the many different places that the car’s major components are coming from. The sourcing pattern seems identical to, although far more complex than that of laptop manufacturers. (although laptop manufacturing is even more outsourced, with contract manufacturers producing HP and a number of other major brands out of the same factory and sometimes the same production line.

From Automotive Weekly

We took the example of one vendor called Dura. A visit to their website demonstrates that they make numerous automotive components, which they sell to many different manufacturers.

Dura’s Part Distribution Model

Dura does not sell parts directly to retail customers, but they do to dealers and independent shops. (however, dealers do have a stranglehold on the industry, and many parts are carried only by dealers) This is one of a number of areas where business are opposed to “free markets,” and instead select tying agreements and monopolistic competition.

______________________________________

Why Doesn’t eBay Own the Auto Aftermarket?

eBay is the largest service parts database in the world. However, for some reason, eBay is not prominent in automotive service parts. The fact that automotive service parts are expensive, yet only a modest service part market has developed on eBay is an indication that there are significant restrictions to who can get access to parts, and that there are in all likely strong restrictions on part suppliers, as part of their agreements with manufacturers as to who they may sell parts to in the aftermarket. No such restriction exists for computer components, where anything can be found and purchased on eBay.

Even the most esoteric service parts for computers are available at low cost on eBay. For more on eBay and their success in service parts see the link below.

http://spplan.wordpress.com/2007/07/22/parts-databases-and-ebay/

______________________________________

What This Means For Service Parts Network Design

What this means is that the dealer system for distribution is even less efficient then we originally thought. People are going to dealers to get parts they think are made by manufacturers (Honda, Toyota, etc.), that are actually made by suppliers. All of these middlemen could be eliminated from the system and actually should be. These suppliers are the creators of these components and they should not be controlled by manufacturers, much less have to go through dealers – so dealers or independent repair shops can add an extra markup with no value add – to service parts.

References

Who Really Made Your Car,Thomas H Klier and James Rubenstein, W. E. Upj0hn Institute, 2008

The degree to which dealers are “taking it easy” is evident in the latest Consumer Reports survey where despite the overwhelming advantages of being part of a dealer network, dealers on average provide a customer experience that is 7% lower than that of independent maintenance shops. However it gets a lot worse when actual repairs are needed. For those that required repairs, only 57% of customers were satisfied with dealers vs. 75% who were satisfied with independents.

The consumer reports survey is a stark condemnation of automotive dealers.

Why Do Dealers Perform So Badly?

So the natural question is why are dealers performing so badly. The traditional concept is that dealers provide better, although more expensive service, and that they provide better service because of the following:

• They are trained by the manufacturer
• They have information available from the manufacturer
• They are more expensive
• They know the cars better because they work on the same make over and over again.

Clearly the outcome (service performance) does not match the bullet points above, and in fact, a number of the bullet points above are dated. For instance, Honda stopped sending its mechanics to its own internal training program several years ago, which by most accounts was excellent, and has instead outsourced its mechanic training to a trade school in Arizona, which is nowhere near as good, and which does not specialize in Honda. Knowing little, mechanics are now increasingly reliant upon Honda’s remote service technicians that are available by phone out of Honda’s Southern California main office. Honda at one time had a sterling reputation in service maintenance, and it now no longer does. If you bring your Honda into a dealer now, you can expect a technician trained by a generic trade school which was a low cost bidder to Honda.

The Monopoly Explanation

Most likely, automotive dealers are not better because they do not need to be to survive. This is the best explanation, when everything else is tilted in the dealer’s favor, and they still cannot perform in a manner competitive with companies with far fewer advantages. Of course, this does not even include the cost of dealers charge, which is widely known to be exorbitant and far more than independent stops.

Service Parts Website Incompetence

We first found how bad dealers performance was when we tried to find service parts on their website for online purchase.

See this link for the full article:

http://spplan.wordpress.com/2009/05/15/auto-service-part-networks-are-a-mess/

As we recount in the article, we had a devil of a time finding a very simple part (the right interior door handle cover to a 97 Honda Accord), and after visiting many sites for hours, we can state with confidence that dealers have no idea how to put together a service parts transaction website. Furthermore, we question the logic of having manufacturer’s outsource their parts management to dealers, when manufacturers are much more capable of doing themselves or outsourcing it to companies that actually know how to manage large service parts inventories, (Amazon.com is extremely capable of creating a service parts shopping site.) Is this a strategy designed around enhancing the customer service, or a compromise thrown out to the dealers to enhance dealer profits?

Additionally, since cars should be built to order items, and ordered online, our out of small showrooms that just stock test models, combined with the fact that dealers can neither maintain websites nor provide service superior to independent shops, the dealers value-add to the car buying and maintenance cycle is not apparent.

Dealerships typically have very nice buildings. However, aside from architectural flair, dealers are not a value added part of the purchasing or service chain. Wise automobile manufacturers of the future will offer their cars direct from their website (or from a small retail outlet with test models) – saving tremendous money in reduced inventory (not having cars sitting around on lots), and allow independent shops to flourish though both providing a top notch service parts website (for both dealers and customers), and through offering extensive service documentation with the creation of a service parts portal which publishes and builds on maintenance information by allowing mechanics in the field to contribute to its content. (see our example of just such a portal below)

http://www.servicepartsportal.com/

The New Dealer-less Model

Any car company that were to operate under the dealer-less model, would be extremely cost competitive with the current manufacturers running the cost heavy and inefficient dealer model that have to base many of their decisions not on what is right for the customer but what makes dealer happy. In fact, they could not be effectively competed against on the basis of either cost or service parts management or overall service level.

Conclusion

Dealers appear to be cruising on their monopoly position and the fact that customers come to them without considering all of the alternatives. The evidence for this is their very poor showing in terms of customer satisfaction for repairs, and in their inability to create or maintain service parts websites. Given the technological development of the web, dealerships are no longer necessary and should be done away with.

References

This is actually a very good blog on autmotive dealers. However, this article is symptomatic of the problem in terms of thinking about dealers.

http://wardsdealer.com/ar/auto_whoopee_bankrupt/

In it the author, who seems to know a  good deal about dealers bemoans the closing of dealerships by the large American automotive manufacturers. However, the analysis seems misplaced. While the author is focusing on the number of lost jobs, what he is not focusing on is how little value the existing dealer structure adds to the process and to the consumer. Jobs that do not add value in an economy do not deserve to be saved (of course this applies to those with degrees from Harvard and Wharton who work on Wall Street over people that work for dealers of course). Still,  why is there no mention in these articles of the atrocious behavior on the part of dealers, and how dealers perform far worse than independent repair facilities, or how their inventory heavy push model of distribution is completely archaic? For us the question is not why the current dealers being closed are being closed but rather other dealers are even in business.

In our previous post we discussed the problems with how automotive service parts websites are dominated by dealers. We also discussed how this is inefficient and why these web sites should be centralized and either managed by the manufacturer, or simply outsourced to a company that has this as a focus.

http://spplan.wordpress.com/2009/05/14/auto-service-parts-websites-a-problem/

However, after further research it turns out automotive service networks have even bigger problems than this. When consultants for service parts planning software company MCA goes into an account and uses its SPO software to perform inventory re-balancing, they often find that parts are kept too low in the supply network. This is often because fill rates are only being locally managed and and local managers are attempting to move parts to where they will eventually be consumed. The problem with this is that transferring parts from a forward location to another forward location is less efficient than moving parts from the parts depot to the forward location. Secondly, there is no reason to move a part to a forward location unless there is a high probability of consumption, or unless transportation lead times are particularly long. This analysis of where parts in the field should be located goes by a number of names including multi-echelon inventory optimization, redistribution and inventory re-balancing.

See the diagram below.

See these posts for more on part redistribution.

http://spplan.wordpress.com/2008/04/23/drp-vs-multiechelon-planning/

http://sapplanning.wordpress.com/2009/04/23/inventory-balancing-in-spp/

Generally, the independent dealer model has worked against, and appears to be working against rational inventory pooling. AMR Research (who we are not big fans of because they are so frequently wrong on their forecasts) does have a good point when they bring up this point in their paper Service Parts Planning and Optimization.

During the course of this research, we found SPP applications tended to be very tactical in nature, solving specifc inventory, fll rate, or service-level goals. Oftentimes service is still being viewed as a cost center, and SPP applications are not necessarily viewed as the keys to a greater world of service nirvana.

One explanation is that the buyers of SPP software tend to be planning managers or director-level planners who have no jurisdiction over service and repair or other areas of the SLM model. Other reasons include outsourcing, where OEMs have outsourced the service process but retain the planning aspects, or the fact that the company was never in charge of service in the frst place—think of an auto OEM and the dealers that actually provide the service.

During the course of this research, we found SPP applications tended to be very tactical in nature, solving specifc inventory, fll rate, or service-level goals. Oftentimes service is still being viewed as a cost center, and SPP applications are not necessarily viewed as the keys to a greater world of service nirvana. One explanation is that the buyers of SPP software tend to be planning managers or director-level planners who have no jurisdiction over service and repair or other areas of the SLM model. Other reasons include outsourcing, where OEMs have outsourced the service process but retain the planning aspects, or the fact that the company was never in charge of service in the frst place—think of an auto OEM and the dealers that actu-ally provide the service. – AMR Research

Better Service Parts Planning Begins with Cooperative Planning

Rather than having every dealer attempt to manage its inventory, a much more rational and effective setup is for the dealers to pool their parts at a local depot and for the depot to manage the parts for them. Daily local “milk runs” would ensure part flow to the dealers, and would reduce the poor inventory turn of parts at the dealer location. A series of these depots can then be large enough to be electronically connected and to have their inventory represented in a web order fulfillment system that can better match supply and demand than can a series of disconnected dealers all trying to manage a smaller amount of inventory locally. Honda (for instance) could manage this themselves, or instead could outsource the management to a company like Amazon.com, that really knows how to produce transactional web sites and knows how to match supply and demand. This solution would be vastly superior to the current one where small dealers attempt to manage their own service parts websites (and where it took us 2 hours searching various dealer sites to find that we would have to call in to order a part)

What is happening in the dealerships is a disinterest in making changes or becoming more flexible in order to adopt new technologies. Companies can make a lot of money in the short term by simply living off of monopoly power. GM was the poster child for inept management, inward thinking, abusive supplier relations and unresponsiveness to customers. A good catchphrase for management consultants could be “Don’t be Like GM.” While Honda quality is much much better than GM’s ever was, Honda’s dealer network with respect to their service parts management is not all that much different. In fact most manufacturers seem to employ the same inefficient system. This demonstrates the restrictive influence of the dealership system that no matter how good the car company, the dealer system remains anachronistic.

Conclusion

The current dealer centric automotive service distribution system is an anachronism and is probably one of the reasons that dealerships have such high costs. Instead of attempting to reduce these costs, dealers are simply passing on their inefficiency to the consumer. However, dealers should be wary. While they have used political finagling to prevent web based car purchases, this will eventually come to pass. The only thing that the dealers are really necessary for is for providing local service. They should do what they can to make their service operations, which includes service parts planning and management as efficient as it can be. A big part of the answer to this is to begin cooperatively or centrally planning and pooling inventory.

Parts Hub

The parts hub concept has also been proposed by John Snow, at Enigma, which is a software company focused on parts procurement decision support. The post on this topic can be found here.

http://www.uptimeblog.com/aftermarket/how-fewer-dealers-can-sell-more-parts/2009/05/

Reference

Service Parts Planning and Optimization, ARM Research 2007

Post-Reference

After this post was published, we found that auto dealers have a considerably poorer track record on customer service for repairs than independent shops. This promoted us to write this article that questions the validity of dealerships generally and proposes a dealer-less model.

http://spplan.wordpress.com/2009/05/18/automotive-dealers-mostly-useles/

John Snow has some interesting things to say about this concept at the link below.

http://www.uptimeblog.com/electronic-parts-catalogs/simplifying-parts-sales-make-it-easy-for-the-customer/2009/05/

It is always amazing to come upon a technology that is so amazingly underutilized. This would be the case for service parts online databases.

The Story

We needed a door handle assembly part for a 1997 Honda Accord. First we started with eBay, which really had a pretty small inventory. We could only find the door handle assembly for a four door, not for a two door. This was a dealer only item. The trouble began when we started looking through dealer websites for the item. The experience began to get us thinking that the dealer value-add is seriously in question. Dealers are not necessary to buy cars (they could be bought online, but tested at a manufacturer sponsored center in a mall that had just a few models). The care could then be either transshipped from a different location, or simply build to order. However, instead of this we have this medieval auto dealer system that holds massive amounts of inventory so buyers will make impulse purchases “that day.”

Service Databases

When looking through the websites of dealers, it was absolutely maddening to try to navigate them. Most the sites are caught in a time warp and exhibit the worst of web navigation and design. Some of them ask for contact information so they can treat the desire to purchase parts as a “lead.”

San Francisco Honda, like 99% of the dealerships, seem to seriously misunderstand what the web can do, and how it can help automate transactions. Now we will be calling to the dealer, just like we would have back in 1940.

Why Has Online Parts Supply Demand Matching Been Decentralized to Dealers?

Why does Honda allow dealers, who lack the interest or size to develop competent transactional websites to sell auto-parts on-line? Why are Honda, and other major manufacturers, not managing this with a single website and a national network. It appears as if the dealer network (a way for manufacturers to sell franchises and not have to worry about retail, is interfering with the new realities and efficiencies of the web. Automobiles may have to be serviced locally, but there is no reason, with our fast shipping network, for parts to be managed at dealer locations. And especially when a customer wants to order a part, there is absolutely no reason they should have to a dealer to do so.

Its does not have to be this way. The fulfillment could be performed by dealers, but Honda could manage the front-end, much like Amazon.com.

Learning from Amazon.com

The lesson from Amazon is that the web based supply demand matching no longer needs to be performed by the same organization that performs fulfillment. See this article on Amazon.com and how they serve as a supply demand matcher.

http://sapplanning.wordpress.com/2009/05/07/gatp-atp-trees-and-amazon-com/

IT and Monitoring Competence and Fourth Party Logistics Providers

The concept of multi-partner coordination enabled by monitoring tools is a concept in logistics called fourth party logistics and is covered in this post.

http://fourthpartylogistics.wordpress.com/2009/05/23/fourth-party-logistics-providers/

It’s a sad fact that there is simply not a lot of thinking going on in the management of service parts.