As often happens with the popular Nikon camera launches, first deliveries have been made and the product is sold out. Now we get the grumbling about "why didn't Nikon make more?"
It's a delicate balancing act that Nikon has to go through. We're not talking mass market consumer item here, where you build huge, permanent capacity to meet insane demand (the current Apple model). A D4 is going to sell a few hundred thousand copies in its lifetime, a D800 maybe a million+. Sendai's current capacity is 5,000 D4's a month, 30,000 D800's a month. Best case scenario (assuming that the parts are actually available, which is a huge assumption considering we're talking about precision sensors here), is that they could double production by doubling staff or making everyone work double shifts.
It should be clear that initial demand was far greater than what Nikon delivered. Let's assume for a moment that what Nikon delivered was two months of production. How many months do you think Nikon should make a product before actually delivering it to a customer? Three? Ten? Initial demand on these units was probably 3-4 months worth for the D4, 6+ months for the D800. Would you be willing to wait until June in order to guarantee a D800 on day one? Didn't think so.
There's a risk in building months worth of production and having it sit in warehouses: the product doesn't get user tested. As much as every tech company wants to ship a product that's perfect on day one, it's unusual that something wasn't caught. It's a numbers game. Even a few hundred testers testing round the clock aren't going to hit every last nook and cranny that tens of thousands of actual users will.
Nikon faced this problem more than once in the past. The D1x shipped initially with a real issue. It was already heading to user's hands when the problem was found. Nikon had to scramble to fix the problem in the field and it took a few weeks to get every unit dealt with; the volume back then was far lower than it is today, and that was without piling up many months worth of units before first ship.
Let's look at the other end of the demand chain: when the product's three years old the demand is low, because people are waiting for the "next generation" (which builds a new initial demand stream). You don't want to have a factory that can build far beyond late product cycle demand, otherwise you have idle capacity and lots of workers you're going to put into furlough.
Overall, Nikon tries to balance the problems it faces in delivering a new, high-demand product like the D800 (and to a lesser degree, D4). Build enough so that there's not a tickle at launch, but a significant number. Don't build and sit on too many and risk bigger problems and idle capacity afterwards.
If we look at Nikon's factory numbers, they add up this way: D4 will likely have a total product build in the neighborhood of 120,000 units (5k * 12 months * 2 years); the D800 might manage 1.44 million units (30k * 12 months * 4 years). That's assuming they stay popular enough during their entire life cycles to keep the factory at full capacity. Compare that to the Apple iPad's 3 million units delivered in the first three days. So all those writing me saying that Nikon should just do what Apple did are a bit off in their objection: the iPad is a mass market item with a one year life cycle that sells in the tens, maybe hundreds of millions of units a year. It's worth building additional factory capacity for that kind of demand. On the other hand, Nikon professional cameras are niche products that take a long time to reach even a million units (and that's assuming they're popular and state-of-the-art through their life cycle). It's not worth building huge capacity for that because it will go unused after initial demand dissipates.
Personally, I think Nikon has chosen a reasonable balance point. That doesn't make those waiting in line for a D800 any happier, obviously. I'm confident that Nikon will make some short term adjustments to try to quell some of the thirst for product. I noticed that they seem to be sending weekly shipments by air out of the factory at the moment, for example. They're not trying to amass a large number and put them on a slow ship to the customer. Instead they're delivering them hot off the line as fast as they can. They're likely to run some overtime and speed up their parts supply chain, too. They may even shift some D4 work to D800 work. What they won't do is try to meet all demand instantly.
Given what I'm hearing from dealers and users, I'm betting that D4 cameras will loosen up in supply within two months to the point where you can find one here in the states if you really want one. D800's, on the other hand, are going to be in relatively short supply for a longer period, especially once people realize just how good the camera is. There was always a stronger demand that went unfulfilled for a D3x at the D3 price, but we received a better-than-D3x-at-half-the-D3-price. Demand is high simply because of that, let alone the pent-up D700 upgrade demand.