Something I always ask when a sales clerk is trying to sell me a warranty, “Why are you selling a product you expect to break?”. I am only half kidding. Products do break and sometimes warranties make sense. My Dad in particular has had terrible luck with weed eaters, snow blowers, water softeners, computers, and vehicles. Now I might advise he not buy the cheapest option, because when you price in the warranty, the better, sturdier, more dependable option might be cheaper. For those products that are not objectively crap, I have developed a simple formula to determine whether a warranty is worthwhile:
- Pretend you are a bulk purchaser and add the cost of 100 of the product. (100 is easy, add two zeros).
- Estimate an X% failure rate and add up the cost replacing the X% of the product.
- Add up the cost of 100 warranties on the product.
- If the warranty cost is higher than replacing failed products, the warranty is a money loser.
Determining the failure rate is the difficult part of this formula. We cannot assume a 100% failure rate like my Dad. Consumer Reports is a great resource determining the failure rate on product lines. Some items defy conventional failure rates, such as the well reported 30-60% failure rate of the original batch of Xbox 360 game consoles and the legendarily indestructible Nokia feature phones. Sometimes store clerks will be honest with you about a products return rate. Do the best guesstimate you can.
Let’s use Macs as an example, because the AppleCare program for Apple products is a particularly attractive warranty for buyers and the failure rate of Macs is well reported at about 17% within 3 years (or at least, it used to be). That means over your lifetime, 17% of your Macs will need to be replaced within 3 years. For a $1200 Macbook Pro, you’d spend $120,000 on 100 and $20400 replacing 17% of them. Buying 100 warranties at $250 each would cost $25,000. Buying AppleCare for a $1200 McBook Pro is therefore likely to be a money loser.
Also consider this: your stuff loses value over time, and the longer you keep something, the less expensive it is to replace. For electronics and vehicles, you can almost see the drop in value in real time. Most all products come with a 1 year warranty, so you only need consider the cost of replacing the product when it breaks in the second or third year versus the cost of the warranty. Let’s assume a 30% drop in value in the second year and run the MacBook Pro numbers again: the cost of replacing 17% of them is down to $14,280 while AppleCare on all of the computers remains at %25,000.
Since we mentioned the Xbox 360, let’s use this game consoles as another example. Assuming an average of $400 per console in 2005 dollars, the above formula says that a warranty has value at $120 or less. However, in the second year of the console’s life, the price dropped as low as $300 and the reliability improved dramatically. This second round of consoles was the one to get (careful buyers always win). The lower cost of replacing the consoles means a warranty had value at $90. In actually, though, Microsoft extended the factory warranty of the Xbox 360 to 3 years, thus making any 3 year warranty a waste of money.) The legendarily unreliable turd like the original Xbox 360 is an anomaly, though and game consoles are notoriously reliable with have a failure rate of about 5%. (Microsoft sells a $60 3 year warranty for its $500 Xbox One, meaning it believes the failure rate on Xboxes is back to the industry average). Assuming the life span of a game console is about 6 years and an adult will buy about 7 game consoles in a lifetime the original Xbox 360 raises the average failure rate of a lifetime of purchases from 5% to about %10. That brings the warranty value down to about $9 for a console like the Xbox 360 that lost a quarter of its value in the second year.
Using this formula, it looks like a warranty never makes sense. That is why everyone is always trying to sell you one!