To: Geoff Nunn who wrote (1845 ) 9/5/1998 6:20:00 AM From: rudedog Read Replies (3) | Respond to of 2578
Geoff - To achieve better coordination it also discloses chip prices to everyone well in advance of when they are actually produced. Intel produces a 'price roadmap' for chips indicating approximate dates and anticipated pricing for exactly the reasons you state. BTW does this not negate the claimed cost advantage of Dell's short inventory cycle when cpu prices fall? If CPQ knows in advance when CPU prices will fall, why would they get caught holding big inventories of CPUs right before a price drop, as has been claimed by Jim Kelley? In fact CPQ drives their CPU inventory as low as possible before a price drop.1. CPQ's willingness to commit in advance on these "futures contracts" doesn't really solve the problem of CPQ's uncertain demand because CPQ buys some of its chips in the spot market, right? The only time CPQ would buy on the spot market would be if their actual demand was higher than forecast and their price on the spot market was lower than the price they could get from Intel. In any event this is likely to affect only a small portion of CPQ's chip buys, a few percent at most. CPQ would more likely move previously committed buys forward and add to the downstream forecast. Intel still has a variability which is small in comparison to Dell's purchasing pattern, and Intel does not have to sit on the chip inventory.Don't you think Dell would get in on these deals if they were bargains? They are not bargains for Dell. Dell has one of the best managed cash models in the business. Therefore, they have determined that given their manufacturing and distribution model, it is cheaper for them, overall, to buy chips via the JIT model, even if the chips themselves are more expensive. This has certainly been true recently, since Dell's downstream distribution costs are so much lower. But it's silly to assume that Dell can get the benefits of their model and also the volume benefits of CPQ's model. Let me refine the point by exaggeration. Let's say CPQ could get an absolutely great deal by buying huge volumes of only 2 chips - say 450 Xeons and 333 Celerons. Lets say they cut a deal that gives them a 50% advantage over Dell in CPU costs on these chips. This would clearly, unmistakably give CPQ a much lower cost structure. Dell, on the other hand, would have the ability to flexibly adjust processor mix to meet customer needs and would probably sell more computers overall, though CPQ would sell more of the 2 particular models they had under this deal. So in this scenario, Dell would willingly chose to buy CPUs at twice CPQ's price. It would be the smart thing for them to do. But if, given all that, you asked me who had the lower cost structure, my answer would be CPQ. Dell is betting on JIT and optimizing for that, and minimizing their exposure to inventory costs. CPQ has been on the other end of the spectrum, optimizing for predictable high volume production. A more valid question is whether Dell will shift some of their production to a more volume model, which they need to do to satisfy some of their large corporate accounts going forward, and what will that do to their cost model? And as CPQ takes costs out of their distribution chain, will they want to try and get some of the cash management benefits of Dell's approach? Each of these companies would like to maintain the strengths of their current model while picking up some of the benefits of the other guy's approach, but it's easier said than done for both of them.