E-commerce Sites Edit Customer Reviews

Carl Bialik from the WSJ writes "Online retailers have a wide range of approaches to customer product reviews, with some struggling to balance candor with the desire to sell product. The Wall Street Journal Online has an overview of sites' policies. Newegg 'says it has a team of eight people who monitor reviews and reject submissions if they are too vague, mention competitors or criticize a brand without specific product insight, among other reasons. From July 1 to Aug. 2, the site received 18,188 reviews and rejected 15% of them, according to a Newegg spokesman.' Meanwhile, Overstock recently changed its policy: 'The Web retailer had been relying on its merchandising group -- the employees responsible for deciding which products to sell on the site -- to monitor reviews submitted by customers, but found that the group tended to approve only positive reviews. In January, the Salt Lake City-based company changed the monitoring responsibilities to its marketing team. The company now says it posts both positive and negative comments, as long as they are constructive.'"

Comments

Some edits can be insidious.

[Anonymous Coward]

I once wrote something to the effect of: "I can't recommend this laptop backpack for anyone who travels a lot," and the site neatly editted out the "can't". Never filling out one of those things again.

Heh...

[BlackCobra43]

The company now says it posts both positive and negative comments, as long as they are constructive.'"

Nothing's easier than saying "Sorry, I won't do it again" and pulling up your pants after getting caught. It doesn't change the fact you WERE caught and you DID do what you were caught doing.I also in no way guarantees that this behavior will not resurface at a later date.

Re:Heh...

[Ingolfke]

Nothing's easier than saying "Sorry, I won't do it again" and pulling up your pants after getting caught. It doesn't change the fact you WERE caught and you DID do what you were caught doing.I also in no way guarantees that this behavior will not resurface at a later date.

They had a problem... people would post useless reviews "it's good", "I hated it", "Buy me an Ipod", etc. and this was detrimental to the customers who were trying to figure out whether or not they wanted the product. So they tried a solution... to have the product team screen the reviews to make sure they were accurate based on what the product teams new about the products. A good idea, in that who is better to validate a review than the people who manage the product. Of course, the unforseen consequence was that the product team didn't want to look like they were buying crappy products and not doing their job (or they just though they had the best products) so they biased the screening process. The marketing team has in theory less bias (and less knowlege of the product perhaps), and in theory should do a better job with the screening process this time. The point is that the company is interested in providing the customers with valueable meaningful reviews and in order to do that they are willing to continuously improve their process.

As long as they keep working to make it better for the customers and then that is a good thing.

Re:Newegg rev 01

[antifood]

Whats wrong with Newegg? I have always found that they have competitive prices, and have always taken care of any problems I have had. Any insight is welcome.

Fry's

[DarkHand]

Fry's Electronics owns Outpost, and I know someone who works at Fry's... At the penalty of losing their job, they're not allowed to say ANYTHING negative about any product whatsoever. They can't specifically say one product is better than another either. I wonder if this policy will change now as well?

Re:Newegg rev 01

[BearInTheWoods]

This hits close to my experience, since I purchased 2 external RS-232 modems from NewEgg recently and submitted reviews for both (after some test time with both). The cheaper one (an 'Amigo' [generic Conexant-chipset-based] modem) worked great. The more-expensive one (a Diamond SupraMax) was horrible -- wouldn't work with 'kppp', had no on/off switch, wouldn't remember (after an ATZ) S register settings that had been written to non-volatile RAM (AT&W), etc. Keep in mind that the Amigo modem had none of these flaws.

I submitted a positive review for the Amigo (see the one by 'External Modem User, 7/24/2005 2:52:49 PM') which got accepted and a negative (but constructive) review of the Diamond, which was rejected.

Here's my original Diamond SupraMax review (which I may tweak and re-submit to NewEgg sometime):

I bought this Diamond SupraMax (Model 'SM56E', NewEgg part #N82E16825116109) modem along with the 'Amigo' modem (Model 'AME-CA95', NewEgg part #N82E16825137104) to replace a Creative Labs external RS-232 'ModemBlaster' that recently died.

The Diamond SupraMax modem is a huge disappointment, mostly due to its poor design. The Diamond modem is also more expensive than the 'Amigo AME-CA95' (Conexant-chipset-based) modem even though the Diamond modem is of lesser quality.

For starters, unlike the Amigo modem, the Diamond SupraMax has no 'On/Off' switch. You must physically disconnect the 9V plug in the back if you want to power it off! Yuk!

The Amigo modem has a nice On/Off pushbutton on the back right side.

Furthermore, the LED indicators for the Diamond modem are the worst I've ever encountered! First off, there are only two LEDs -- one for 'PWR' (Power On) and one for 'OH' (Off-Hook). Despite what the photo on the front of the Quick Start Guide shows, there is no "DATA" LED, let alone a specific 'RxD' (receive data) and 'TxD' (transmit data) as one would expect on any decent external modem. Secondly, the meager 2 LEDs which _are_ present are very difficult to read, since they're not typical LEDs with permanent, silk-screened text near the LED but are of a design where the LED light shines through to show red-colored text ('PWR' and 'OH') and the text is almost unreadable at any angle except the exact "proper" angle. The 'OH' text is almost unreadable at _any_ angle, actually.

The Amigo modem, on the other hand, has dedicated, well-labeled (viewable at all angles) LEDs for 'RxD', 'TxD', 'OH', and 'READY'.

Annother annoyance -- the Diamond modem does not retain certain settings (e.g. 'ATS95=1', which causes the modem to report DCE [modem ISP] speed instead of DTE [PC modem] speed) through a simple 'ATZ' (modem 'soft' reset) even when they've been written with 'AT&W' ('store user profile' command). All other (good) modems I've encountered (including the aforementioned Amigo modem) will retain the 'S95=1' register setting through an 'ATZ' command as long as the 'AT&W' command was used once after setting the register. This isn't a "show-stopper" problem because you can always instruct the modem to 'ATS95=1' (or 'ATW2', which is similar and may work with some modems) as part of the initialization string in whatever application you use (Windows or Linux) to dial out, but it's annoying and indicative of the bad design of the Diamond modem.

The Diamond modem works well enough in W98se with the driver supplied on the CD-ROM.

Linux use (Slackware 10.1, currently) is a different story, unfortunately. I could not get 'kppp' (the KDE 'ppp' dialer application) to fully connect to my ISP using this modem. After lots of testing, tweaking, and cursing this modem (which is the poorest-designed of all the external RS-232 modems I've ever used), I finally got it to fully connect (and start the 'ppp' session) by invoking the 'pppd' process manually and using a 'chat' script. I still d

My own examples were more arbitrary than that

[ianscot]

Any bestselling item will never have an average review of less than 4.0/5.0 stars.

There is a much higher standard for poor reviews than good ones; and even excellent reviews of a product may disappear if they are unfavorable.

I doubt Amazon is really able to pull off that sort of thing consciously, but we'd have to know more about how their process works to say for sure. From my limited experience -- okay, mod me an embarrassed loser, but over several years I've posted a bunch of Amazon reviews -- things seem much less calculated than that.

I've never had a review disappear entirely, and really most of the "editorial" changes to my reviews have seemed like arbitrary, almost nonsensical elisions made by rigid formula. Two easy examples I can think of:

• I compared the (godawfully calculating and soulless chick flick) movie "Love Actually" to a well-produced episode of "The Love Boat." The Amazon editors removed the name of the series, but left in the names of the characters - Julie, Doc, Gopher, and Captain Steubing.
• For one history book I said something about how people wanted "bastardized" history. They removed the bad word and stuck in a "..." for it. That's happened a few times, in places where my use of the language was not at all offensive.

So, okay, I can see a simple filter catching the bad words, but when did "The Love Boat" become a bad word? Did they think it was a copyright problem? Or what?

Most of my negative reviews are left as-is, but you know, I tend not to post "This SUX."

The overall effect might be to push products, in sort of the same sense that the overall effect of our court system can be racist. I don't think individual decisions within either system are rational enough to amount to a conspiracy, though. You'd have to look at how the process works to figure out why that happens.