Edgefactor

What do you mean my AppleCare doesn't cover against the horrible scratches?

Recently, every time one of my friends, acquaintances, neighbors, co-workers, or clergy-people have decided to drink the white and silver Koolaid, one of life’s most enduring and difficult questions has been presented to me:

To AppleCare or not to AppleCare?

Slings and arrows aside, I have actually given this question a reasonable amount of thought. Upwards of 15 minutes at least. Eventually I came to the following profound and potentially Earth-shattering conclusion:

The AppleCare Protection Plan (APP), like most “product protection plans,” is basically a scam. I’m sure it’s tremendously profitable for Apple. I’m also sure that if anyone ever gets around to posting a comment on here, it will be to refute my argument with extensive quantities of anecdotal evidence. Oh well.

The key to most product protection schemes and extended warranties is understanding what exactly is covered by the policy you are purchasing. As a general rule (APP included), product protection plans take the existing warranty that comes free with the product and make it last longer. This, “manufacturer’s warranty” guarantees your product against “defects in material and workmanship” that occur within the specified time frame. (APP does include a couple other things, but those aren’t usually what the salespeople focus on during the pitch.)

Product warranties expressly disclaim any coverage against things like: “normal wear and tear,” “theft,” “accidental damage,” and any failures whose root causes can be traced back to something within the realm of control of the end user.

Ok, so let’s make sure we’re all on the same page at this point. Here’s a few situations for us to examine:

• My one-year-old iBook has accumulated a bit of grime below the keyboard where I rest my hands when I type. It’s a bit unsightly but has no functional impact on the operation of the product. This would constitute “normal wear and tear” and as such would not be covered by my original warranty or by APP.
• Rumors and anecdotes are swirling around the internet about how scratch-able the new iPod Nano’s are and how susceptible to breakage the screens are. There are certainly two separate issues here. If you are unhappy about how scratched up your iPod is, tough beans. No warranty or APP is going to save you here. (This would probably be your best bet.)

  But the issue of the spontaneously breaking screens is completely different. In this case, Apple has admitted a defect in a small percentage of units from their supplier. This is certainly covered under the product warranty and if, for some reason, your screen spontaneously broke after you had already used the thing for a year, APP would have your back here.

• One of my friends told me about how they knew someone whose screen on their PowerBook started to give them trouble a couple years into their usage. I admittedly don’t know the exact circumstances of this but in that case, APP did replace the screen for free. However it is not entirely clear to me whether or not they actually were required to by their agreement.

  The fact is, almost any problem you have with your computer or other product after a year could be blamed on “normal wear and tear.” Apple basically pulled this maneuver with some of the original iPod’s whose battery life became greatly diminished after a year or two of use. (This is actually fairly normal for high-capacity rechargeable batteries like those found in laptops and MP3 players.) Until a class-action lawsuit got their attention, Apple basically said “Normal wear and tear, not our problem.”

Ok, so we’ve seen some situations where APP would seemingly be a good investment. The real crux of the problem with APP and it’s ilk is the statistical unlikelihood of ever needing it. The failure rates of mechanical (and electro-mechanical) devices is typically described as a “bathtub curve” which looks something like this:

The horizontal axis here represents the amount of time the product is in service. The vertical axis represents the likelihood that the product will fail.

Observe the red line. This is the line that most closely represents failures caused by “defects in material and workmanship.” As you can see, these problems are much more likely to crop up early in the product’s service life.

The green line represents the likelihood that a product will fail for a completely random reason. These product failures are caused, not by some defect in the manufacturing or assembly process, but generally by an imperfection in the “raw” material used to create the product. These imperfections could be microscopic and virtually impossible to detect. If you’ve ever seen a Mean Time Between Failure (MTBF) rating for a mechanical product and wondered what it meant, wonder no more. MTBF indicates the rate at which a product fails due to these random, uniformly distributed problems.

So if the MTBF for a hard drive was, say 200,000 hours, that certainly doesn’t mean that the average unit lasts for 200,000 power-on-hours. It just means that once you are past the initial “infant-mortality” phase, the odds are something like 1 in 200,000 that your product will spontaneously fail in any given hour. Most likely, some mechanical component will wear out long before one of these spontaneous failures occurs. (The “wear out” phase is represented by the yellow line in the figure.)

People’s lifespans follow a very similar pattern. Early in life you are much more likely to fail than when you are a healthy adult. However, as you age, eventually your body begins to “wear out” and the probability of your “failure” (death) begins to increase. In this case, a person’s MTBF would approximate the odds that an seemingly healthy adult (let’s say between 25 and 40) would spontaneously die.

So what does all this have to do with AppleCare? Well, you see those two dashed, vertical lines in the figure? Let’s say that the first line represents the end of the first year of product use and that the second line represents the end of the third year of product use. If this is the case, then the first phase of the products life (where failure rate is comparatively high) is already covered under the default, 1-year warranty. The APP would cover the second phase (where the failure rate is lowest) and nothing would cover the last phase.

Now, it may seem like I just arbitrarily ascribed temporal values to those lines, but the reality is that Apple (and every other purveyor of similar plans) has employed some very intelligent people to try to get those lines as close as possible to the scenario I described. That’s what makes them the most money. It’s not profitable for them to be having to fix everybody’s computer for free all the time.

The figure I provided may make it seem like the green “constant failure rate” is fairly substantial. In reality, it’s not. For one of Seagate’s recent SCSI hard drives, the MTBF is 1.4 million hours. So in that case, the green line would represent a 0.00000071% chance of failure in any given hour. An iPod or PowerBook probably has a significantly higher failure rate than this given the complexity of the device, but it is still anything but likely.

For my money, I’d much rather protect against stuff that I’m actually likely to do to my laptop. Fire, theft, accidental damage. Insurance provides peace of mind. AppleCare doesn’t.

What the AppleCare Protection Plan does offer (in addition to the extended warranty of dubious value) is extended telephone technical support, on-site repair service (sometimes, read the fine print), and a couple other things. If you decide that these items are worth the 0 to 0 APP costs, that’s fine. I just want to make sure everyone knows exactly what they are and are-not getting from APP before they buy it.

I’ll leave you with a little anecdote of my own that I think of whenever a salesman is trying to upsell me on a protection plan.

My brother bought a late model used car about a year ago from a local dealership. After my brother had talked the asking price down about 25% and they were beginning to hammer out the deal, the salesman began talking about the extended warranty that the dealership offered. It was obviously a well rehearsed exercise that even included props. The salesman proudly presented some obscure part from a similar vehicle and stated that this one part alone cost more to replace than the price of the warranty he was offering. At that point, my brother astutely asked the salesman, “Well, is that part covered?” The salesman’s face fell, “Well, uhhhh, no.”

There was no more talk of the extended warranty.

Update 1/3/06:

Macintouch just released the results of their iBook and Powerbook reliability survey which are quite interesting. They break things down by model and also by the year of ownership in which a repair was required. There seems to be quite a variety, with some laptops seeing a dramatic dropoff in failures after the first year of ownership whereas others actually had an increased failure rate in the second and third years. Overall, the failure rate decreased from 21% in the first year to 18.3% in the second and third years.

Those failure rates might strike you as high, but keep in mind that the figure includes basically any type of failure requiring your laptop to be repaired. To me, “failure” means it doesn’t work anymore, which is certainly not the case if the latch stops working or something.