360 Con Call Numbers
July 6th, 2007So this struck me as fairly interesting. If you read through the transcription of Microsoft’s 360 call - the one where they describe to analysts exactly how much this warranty program is going to cost them - you’ll see that they’re quoting a figure of about $1B
(as we noted earlier). But when the Friedman Billings Ramsey analyst asks Robbie Bach for the specific number of units affected, Bach demurs: Microsoft isn’t willing to share that information.
Fine. But let’s make a really rough guess. Say Microsoft has to replace every single unit that gets returned, because there’s nothing salvageable, at a cost of $500 per console. The $1B set aside, divided by $500 per unit, implies a total of about 2M returns over the three-year period, or 666,667 per year. Assuming that the total number of consoles sold remains constant at 11.6M, this would be a defect rate of: about 5.8%.
If you don’t like the $500 figure, let’s say it averages half that, or $250 per console. I don’t think this sounds too far off, given the variety of costs involved - logistical, support, personnel-related, material, and so on. In that case, you end up with 4M returns, or 1.3M per year. This is a defect rate of about 11.5%, so ultimately we end up with a defect rate of somewhere in the vicinity of 5-12%… a figure which is undoubtedly too high, since there will be more than 11.6M 360’s sold in years 2 and 3. But that’s neither here nor there.
What shocks me about these numbers is that they’re really not all that terribly far off from what Microsoft has been saying all along. It’s not 3%-5%, but it sure ain’t 30% either. And it’s highly unlikely that MS would speculate wildly or openly lie to Wal Street analysts… the long-term costs of the lost credibility would be substantial.
Could Microsoft be making a mistake in not publicizing the true defect rates, and letting those 30%+ rumors fester publicly on the internet? I’m open to the possibility that I’m making a mistake, either logical or mathematical (I haven’t thought this through all that much), but I’m starting to wonder.
Update: There’s a very good conversation going on in the comments thread below about how accurate these assumptions are. used cisco has me wondering if I’m overestimating the failure rate in the following years and underestimating the failures that have already occurred. If you assume that Microsoft has managed to engineer the 360 to reduce the problems in question, and that most of the $1B is to fix/reimburse problems that have already occurred, you get much closer to the 34% figure circulating.
Posted in Geoff, Microsoft, Xbox 360 |
July 6th, 2007 at 11:33 pm
There are a few things to note here. All that Microsoft cares about, as far as its expenditures go, is what the actual return rate is, NOT what the actual defect rate is. The return rate is probably not 100% for defective consoles. There is probably a fair number of people who might simply buy another one (or trade it in at their store using a product replacement plan, which have been very popular for 360s it seems) or forget about it altogether, thinking their warranty is out of date.
I’d also place the $250 a little high, as all they’re doing is providing a refurbished console that was probably from someone else’s RRoD and just had it’s motherboard replaced (but that’s pure speculation). I’d also imagine it’s less than that because the original repair cost out of warranty was $140, and I doubt they were really subsidizing that much, if at all.
In any case, given these two pieces of info, it’s not hard to see how you get into the 20-30% range, or even higher.
I do wish they’d be a bit more transparent at least about what the actual problem with the console is. Is everyone who bought the console before a certain date sitting on a ticking time bomb that is definitely going to go off at some point, whether in a few months or in a few years, just after my warranty expires? Was it just a certain component that was susceptible to die in some cases but not others?
I just hope that my 360 makes it to whenever MS launches their next console, and that their next console is actually 100% backwards compatible with the 360 and a hell of a lot more reliable. I mean, come on… we still have a working NES at home. This things should be built to last.
July 6th, 2007 at 11:41 pm
Ok, I’m editing this comment because I made another mathematical error (I think it’s correct now).
Actually, I’m also kind of confused by what you’re calculating. At $250 we end up with 4M returns out of the 11.6M that they’ve currently shipped. That gets us almost exactly to the 1/3 return rate that’s being thrown about around the web. If you’re going to divide the returns into 3 years to try and get a more accurate picture of what the next few years will be like, you also need to use the numbers sold for each year. So if we assume 1.3M per year, and the console has been averaging about 6.96M sales per year, we get a rate of about 19% per year.
July 7th, 2007 at 7:12 am
This analysis is flawed from the start. You’re making the assumption that MS does nothing to fix the problem moving forward and that the failure rate will be consistent throughout the 3 year period.
I think this is a very dangerous assumption and makes the inadvertent statement that MS are horribly incompetent. I think if MS plans to extend the warranty to 3 years, it almost certainly coincides with actions being taken in manufacturing to mitigate the failures, (like the heat sinks we’ve been hearing about). No company is going to admit a problem, boost their warranty and NOT attempt to fix the problem. That being said, I would rework your math (with leanings towards Jeffs calculations, $250 is way too high) to demonstrate a failure rate of HALF for the period POST warranty change to demonstrate the changes they MUST be making in manufacturing and you’ll get a 30 percent and higher failure rate really quickly. Personally, since EVERYONE I know with a 360 has had a failure, I think 30 percent is conservative.
July 7th, 2007 at 7:16 am
Also, to be clear, I’m saying I think the failure rate BEFORE the manufacturing changes is too conservative. I think MS will make changes VERY SOON, if not already, to get the failure rate down to respectable levels i.e.
July 7th, 2007 at 10:39 am
Hmm. This may be a bit long. This comment will just respond to Jeff.
Jeff: I’ve got the following complaints from your end:
- People may not take advantage of the warranty program
- The three-year period is inaccurately calculating the base of consoles
- The $250 is too high
I’ll disagree with you flat-out on the first two. It’s true that a handful of people may not take advantage of the program, but I’d bet even the people who are unaware of its existence will take the console back to the Best Buy or whatever they picked it up from and then be told about it by them. I’d say the true number of people who opt to actually buy another console is so small as to be insignificant.
For the second point, you seem confused about the fact that this warranty program is prospective, not retrospective. That is, it applies that $1B for the next three years, not the years that have already passed. So the failure assumptions are built around Microsoft’s estimates of failure for all the current consoles, plus all the consoles to be sold in the following two years. As a result, the 11.6M is definitely the right base to use, although it doesn’t account for any of the sales that will occur in years 2-3 (so it’s actually rather low).
For your third point, you may very well be right that the number is too high, but I’m not just talking about the costs of materials. You also have to factor in the salary cost of all the people who will be managing the warranty program, the customer support techs who will be on the front lines, the logistical costs of moving the consoles/inventory back and forth, the costs of stocking that inventory, reporting and maintenance costs, and so on. So there’s a lot there to boost that cost besides the motherboard or whatever that gets swapped out. You may be right, it may still be too high, but it’s more reasonable than you might think.
July 7th, 2007 at 10:45 am
used cisco: If I understand you correctly you’re saying that Microsoft will dramatically improve the console, so the majority of the $1B will go to repairing old 360’s.
If that’s the case, the 4M in repairs would be out of the 11.6M current consoles, implying around a 34% failure rate.
Am I getting that right? If so, I understand your point… and you may be correct. It’s true that they’re retroactively reimbursing people who have already had to pay for such things.
July 7th, 2007 at 12:25 pm
If I understood the warranty correctly, it actually is retrospective. People who had already paid MS the $140 in the past for the RRoD will have their money refunded, and everyone who currently owns a console has had their warranty extended for RRoD to 3 years from the purchase date. Thus, every XBox sold so far is covered under it for at least another 16-17 months, including the original batch known for it’s particularly high failure rate. This is likely where the majority of the 1B will be going (and just to nitpick a little, the actual press release says “$1.05 billion to $1.15 billion”. The “.15″ would translate into $150 million more, which, using your numbers, is an additional 600k units)
I also wouldn’t underestimate the number of people who have purchased product replacement plans from their stores. Even the IGN guys have mentioned how they’re glad they have the product replacement plans, since they’ve been able to get a new box without a hassle once it breaks. Whether places like Best Buy then send it back to MS and MS includes these in its numbers is unclear.
As for the $250, I understand the additional costs that you’re talking about, but I’m just giving you the number that MS has charged when out of warranty, and that’s $140. Was MS actually still eating $110 per console each time they were sent back before?
Since they didn’t really explain what was happening with the RRoD before, it’s difficult to speculate on exactly what needs to be replaced. I may have been wrong about the new motherboard, as the new heatsinks are apparently over the GPU. If the GPU is failing, that is one of the most expensive parts. Right around launch, Businessweek did a breakdown of the components and estimated that the GPU cost $141. I’d imagine it doesn’t cost more than $100 now. Is it another $150 in additional costs? Maybe, but I wouldn’t be surprised if it were substantially less too.
July 7th, 2007 at 12:30 pm
“used cisco: If I understand you correctly you’re saying that Microsoft will dramatically improve the console, so the majority of the $1B will go to repairing old 360’s.”
Yes, thats what I think. If they are facing even a 10 percent failure rate, they would be fools not to try to fix the problem an any future sales. Surely this is not an impossible task. It would almost have to save them money in the course of 3 years.