March 12, 2001

3:06 AM

This nonsense has to stop: " One of the most heavily guarded secrets in the computer business and the closely related consumer electronics industry is how many products are returned by customers because they are defective or the customer cannot figure out how to use them."
posted by jhiggy (25 comments total)

Perhaps the customers who were every bit as important to the technology explosion of the nanosecond '90s as the companies that built the stuff are finally rising up like Peter Finch did in the film classic "Broadcast News": "I'm mad as hell and I'm not going to take it anymore."

so, i can return this column if it proves to be defective?


posted by lescour at 6:39 AM on March 12, 2001

If one out of every twenty high-tech products is a dud, then somebody else must be having some terrible luck. I've built half a dozen computers from parts and upgraded many more and I've never had anything DOA. I had a motherboard fail on me after a year once, and one time a modem went zombie, but i think it got fried by lightning. As far as hardware failures, that's it.

If the 5% includes products and devices the consumer couldn't figure out, then I wonder how many are actually defective? Some of those user's guide translations are pretty ridiculous...

I'm sure I've seen a collection of humorous hardware manual translations, but I can't find the URL right now. During my search I did find this paper that goes over a few of the difficulties of machine translation. I can definitely imagine that impenetrable user's guides contribute to a lot of frustration and unnecessary returns.
posted by techgnollogic at 6:48 AM on March 12, 2001

I actually was thinking that the fact the Trib ended up with three "defective" review units says something more interesting about the PR departments of the companies involved than about the technology being produced... How stupid do you have to be to send a REX unit that doesn't even work to one of the nation's largest newspapers? On the other hand, how Coates became the technology reviewer at the Trib is beyond me - he's as brainless as the day is long and has demonstrated on quite a few occasions that he really hasn't a clue what he's talking about. Perhaps if the column were pitched as "Guide to Technology for Non-Technology People Written by a Non-Technology Person" that would be acceptable... but alas, it's not.
posted by m.polo at 6:51 AM on March 12, 2001

I would say about 70% of consumer electronics I've ever bought has been DOA or seriously defective, or broke within a year of purchase. I thought I was jinxed. Guess it's just a fact of life.
posted by prolific at 6:55 AM on March 12, 2001

The difference between today's products and those that were made during the old west, is without a doubt, complexity. I do agree that once an engineer makes a design, it doesn't go back too often once it's 90% effective or fault tolerant. But with electronics, the rate of failure is so high because of the internal dependancies and many many "links in the chain." I've experienced a great deal of RMAs, mostly from Iomega and early AMD. Iomega had a problem with the drive motor for their parallel zip drives until they went back to the drawing board. AMD seemed to suffer from consistant cache damage (static electricty?). I bet there are tons more that people can relate to...like sending back a defective product only to get a replacement with the same problem. There's a good possibility that a majority of defects occur during the shipping and handling process...that is, if I'm right in assuming that techgnologic is closer to silicon valley than prolific :)
posted by samsara at 7:06 AM on March 12, 2001

To add on to that, it doesn't help that UPS tends to play World Soccer with packages traveling across the U.S.
posted by samsara at 7:07 AM on March 12, 2001

I don't know about hardware being so defective -- maybe I have good luck -- but I do know that software corporations routinely ship buggy software and that they know there's something wrong with their product before it goes out the door.
posted by pracowity at 7:27 AM on March 12, 2001

how Coates became the technology reviewer at the Trib is beyond me - he's as brainless as the day is long and has demonstrated on quite a few occasions that he really hasn't a clue what he's talking about

No kidding, he can't even be bothered to get his stupid shining example right. Everybody knows that <filmgeek>Peter Finch rose up in Network, not Broadcast News. Would it have been so hard to go to imdb and double check that? Geez.</filmgeek>

That said, I've only experienced bad results with technological products when I've made bad choices. Proprietary PC box that can't be upgraded easily or well? Sucks to be me later on when I can't upgrade any software because my processor and platform are antiquated. Off-brand external storage drive? Great, and when it breaks and I return it the week before the company goes Chapter 11, I'm SOL.

In any case, I wonder what level of product "failure" is really incompatability with the consumer's planned use, or with their other devices/systems? If I didn't know that my processor couldn't support a bit of software, and found that I couldn't run it once I purchased and installed it, I might characterise that as a product failure, when in fact it was my lack of knowledge in my purchase.
posted by Dreama at 7:41 AM on March 12, 2001

Bought a Sony DVD player. After a week the disc tray stuck. Bought a Fuji digital camera. The lens stopped retracting after two months. Exchanged it. Same problem in a month. Bought a digital camcorder. The transport ate the tape.

Five percent sounds low indeed.
posted by lileks at 8:00 AM on March 12, 2001

"Some of those user's guide translations are pretty ridiculous..."

Tech writers are wonderful, wonderful people to have on staff. We fix those things.
posted by jennyb at 8:51 AM on March 12, 2001

Oh wait, let me amend that:

Good tech writers are wonderful, wonderful people. We fix those things.
posted by jennyb at 8:52 AM on March 12, 2001

This, in the computer hardware industry anyway, is because computer companies tend to use cheap parts when building their cheap machines. People buy these $500 systems, or even $1000 systems, and they get what they pay for.

For this reason, I will never buy a prebuilt PC. I want to buy all the individual parts myself, so that I can get picky about what goes into my box. I can skimp on the hardware that doesn't matter as much to save myself money, and pay a little extra for the more important pieces. I've never had hardware seriously fail me.

In terms of things like the broken REX, iPaq, and Powerbook DVD drive...that's a twofold issue. One, well, shit happens. Parts make it past inspection that shouldn't. However, it should all be inspected again before shipping them to, say, the Chicago Tribune.

And pracowity's right, software companies habitually release shoddy software. Rather than spend the time to debug properly, they just want to release something on the deadline.
posted by CrayDrygu at 9:03 AM on March 12, 2001

> Good tech writers are wonderful, wonderful people.

Go, Jenny, go!
posted by pracowity at 9:04 AM on March 12, 2001

A long long time ago (circa 1988), I was doing QA for a computer game publisher, and one of the other guys I worked with also worked part-time for a now-defunct PC manufacturer. That particular manufacturer had a "bad-out-of-the-box" ratio of 50%!!! In fact, so I was told, this guy knew of machines shipping with NO COMPONENTS IN THEM to meet production deadlines, and nobody was concerned because they expected them to be returned anyway.

I also have been told by individuals with cred inside the PC industry, that even such big-name makers like a certain international business machines corporation and a certain firm named after fruit expect a return rate of 25-30%.
posted by briank at 9:39 AM on March 12, 2001

Some manufacturers have more quality problems than others. For example, I've had so much difficulty with Sony equipment than I don't buy their products anymore if I can at all avoid it. Their video equipment tends to be OK but I avoid even that just on principle. Their audio stuff develops intermittent connections and other weird problems shortly after the warranty runs out. (Had three Discmen which would spin the disc really fast backward and then claim there was no disc in the player, for instance. The third Discman was what really put me off Sony stuff.)
posted by kindall at 9:49 AM on March 12, 2001

I'd guess the actual bad-out-of-the-box rate is higher than 5 and lower than 90 percent. And that would be an unacceptable rate of failure if you were talking about automobiles or the post office. But one needs to remember that we're dealing with an emergent technology. Yes, PC's have been around for a while now, but they've changed so quickly that manufacturers, etc. don't really get a chance to catch their breath before they're working will substantially different components. You should expect growing pains. (I'm always trying to explain that to people in my office when they want to know why things fail so often; they don't listen either.)

When we get a period of relative stability, when we're not always looking to the (n+1)86 or the Pentium n+1 or whatever, then we'll see reliability improve markedly over time. I hope that the time is approaching when a computer will become obsolete after five years instead of one to one-and-a-half years. And no, I'm not holding my breath.

Businesses already complain greatly about the amounts they invest in technology. But they're willing to continue because they see enormous productivity gains. I think that's slowing down, although I'm sure it varies by industry. I think there are plenty of sectors of the economy that don't really need further upgrades to function. When enough businesses feel that their systems are good enough, they won't upgrade, and the manufacturers can concentrate on making existing models better rather than creating the next great thing.

posted by anapestic at 11:01 AM on March 12, 2001

techgnollogic... man, where are you shopping?! I wish I had the luck you did. Usually with every machine I build, something ends up being defective. I would say 9 times out of ten it's either the memory or the motherboard, however.

Also, as a former network admin I have to say I used to get pretty frustrated at the amount of hardware that would fail on a regular basis. Of course, we had a lot of Compaq stuff there when I started. One thing's for sure--I'll never buy any Inspiration Technology in my lifetime.
posted by fusinski at 11:11 AM on March 12, 2001

I would say about 70% of consumer electronics I've ever bought has been DOA or seriously defective, or broke within a year of purchase.

What in God's name are you buying?

When I worked in IT for one ad agency for about two and a half years, we problaby received $1 million dollars worth of computer equipment (it was a small company). I never, repeat, never had anything DOA.

Of course, at least half of it was Apple-branded.
posted by Mo Nickels at 11:41 AM on March 12, 2001

The other distinction that Coates doesn't make in his article is why the items are returned, though he does hint around it. How frequently, for instance, are things returned to CDW because the user couldn't operate it? The human-machine interface is demonstrably a key factor in the peceived "success" of a product - I've personally stood in line behind people returning things at CDW's Vernon Hills warehouse location where the "reason for return?" was answered, "it's too complicated for me."

And then there's people like me: since CDW is based here, I have the opportunity to purchase things and then return them if they're not exactly what I need or want. There's nothing wrong with them, per se, but since CDW is so great about taking things back - they are the Nordstrom of mail order - why not? I'm sure it works in their favor, because I'm also sure I spend more there than I might otherwise since I know I can return things...
posted by m.polo at 11:59 AM on March 12, 2001

What's CDW's return policy on Software? Because if it's one of the "x-Day No Questions Asked" type (rare for software, I know, but Price Club/Costco used to have it - they still might) then I can tell you why a good portion of that software was returned.

It's not terribly uncommon for people near a store with such policies to buy and burn as often as possible, returning the software and getting an essentially free copy.

I seriously pity the people who are having problems with 90% of the consumer electronics, but at the same time, I thank you profusely. You're likely the reason I've never had a problem I couldn't fix myself when purchasing new toys. :-)
posted by cCranium at 1:37 PM on March 12, 2001

CDW doesn't accept opened software packages as returns, they refer all issues with software to the respective vendor.
posted by m.polo at 2:40 PM on March 12, 2001

I recently bought and returned a Handspring Visor Prism. There was nothing wrong with it, it just didn't suit me very well. I imagine a lot of returns are like that, especially when companies offer unconditional return policies.
posted by kindall at 3:58 PM on March 12, 2001

While field failures do happen (and are sometimes common) a surprisingly large number of returns are due to cockpit error. Once I was in an office while a computer dealer was there delivering a new monitor to a guy who claimed his was broken. So the new one got plugged in and he exclaimed "This one is broken, too! See? Look at those black areas around the side of the screen." The computer tech didn't really say anything, but I kind of stared at the guy, and then said "That's how it's supposed to be."

The businessman was judging on the basis of how his TV looked, and he was expecting "overscan" on his computer monitor. I've run into that with new computer users many times, though less so in recent years.

In a lot of cases a company (like a cellphone company) finds that it's just easier to exchange a completely flawless product than to spend time trying to teach the user what they did wrong. (Doing the latter involves telling the customer "The product's OK. It's just that you're stupid." This is not good business. It also takes a long time and labor costs more than products do.)

There really isn't any solution to this. As products become more technically sophisticated it will happen more and more often. Most companies get used to the fact that a fair proportion of the returns have nothing whatever wrong with them.
posted by Steven Den Beste at 5:36 PM on March 12, 2001

What tweaks me is when, as an experienced software developer, I get a terrible piece of software and I try to let the company know about serious problems. It is impossible to get past the tech support front line to try and make a difference because our goals are at odds: mine is to have a product that works to my satidfaction; theirs is to make sure the customer has a working product.

Recently I purchased a scanner. No big deal except that they clearly never ran Macsbug when they wrote the driver software, else they would've seen the errors that they generate by writing to address 0. My choice, since the scanner manufacturer doesn't respond, is run without Macsbug (which I won't), run without the scanner, or use it only on a different machine.

Recently I purchased a composite video digitizer. The driver also has some serious problems that I was able to debug to a reasonable point, but instead I had to go through every step that the tech support people wanted until they got to the point where they affirmed that it was not a hardware problem and not a configuration problem and couldn't do anything else. It's ridiculous. This one is actually progressing because I disassembled the driver code and found the author's name and company and have been in touch directly now.

The problem is that there are a couple levels of tech support needed: most will be for novices, but what happens if you're not a novice? There is no sanctioned way to communicate with someone who knows more than the support scripts. It seems like there should be a liason between support and at least a QA tech.

Software does go out with bugs. Some companies simply ship and hope for the best. Others make more careful decisions about what is "good enough". One company I worked with rated its bugs from 1 to 10 where 1 is least severe and 10 is most severe. Their criteria for shipping was that the sum of the squares of the severity had to be less than 100. My company had a different rating system and the differing systems turned out to be in conflict at times. Neither were perfect. We were aghast that they would allow a product to ship with 1 severity 10 bug, but in theory they would. They were aghast that we would allow some bugs and not others.
posted by plinth at 6:17 AM on March 13, 2001