"Utterly noxious retail" as Search Engine Optimization

« previous post | next post »

David Segal, "A bully finds a pulpit on the web", NYT 11/26/2010:

Today, when reading the dozens of comments [at getsatisfaction.com] about DecorMyEyes, it is hard to decide which one conveys the most outrage. It is easy, though, to choose the most outrageous. It was written by Mr. Russo/Bolds/Borker himself.

“Hello, My name is Stanley with DecorMyEyes.com,” the post began. “I just wanted to let you guys know that the more replies you people post, the more business and the more hits and sales I get. My goal is NEGATIVE advertisement.”

It’s all part of a sales strategy, he said. Online chatter about DecorMyEyes, even furious online chatter, pushed the site higher in Google search results, which led to greater sales. He closed with a sardonic expression of gratitude: “I never had the amount of traffic I have now since my 1st complaint. I am in heaven.”

That would sound like schoolyard taunting but for this fact: The post is two years old. Between then and now, hundreds of additional tirades have been tacked to Get Satisfaction, ComplaintsBoard.com, ConsumerAffairs.com and sites like them.

(Cartoon version here.)

An excellent argument, as David Segal suggests, for incorporating some kind of "sentiment analysis" into (at least some) search-engine ranking methods. Of course, there are some kinds of searches where this would be exactly the wrong thing to do.

In any case, the publicity associated with this article seems to have spurred Google to do something about this particular case: the search for Lafont eyewear (with which Segal leads his story) no longer turns up DecorMyEyes.com anywhere on the first five pages. Similarly for Bing and Yahoo.

[Update — Amit Singhal responds that "Being bad to your customers is bad for business", The Official Google Blog, 12/1/2010:

We were horrified to read about Ms. Rodriguez’s dreadful experience. Even though our initial analysis pointed to this being an edge case and not a widespread problem in our search results, we immediately convened a team that looked carefully at the issue. That team developed an initial algorithmic solution, implemented it, and the solution is already live. I am here to tell you that being bad is, and hopefully will always be, bad for business in Google’s search results.

I suspect that this "initial algorithmic solution" had not been put in place by Nov. 28, when I originally ran some eyewear searches that failed to turn up the offending company (in the first few pages of results, at least), because the same searches on Bing and Yahoo had the same negative result. Rather, I suspect that a much earlier process, described in the same blog post, had already fixed this particular problem:

First off, the terrible merchant in the story wasn’t really ranking because of links from customer complaint websites. In fact, many consumer community sites such as Get Satisfaction added a simple attribute called rel=nofollow to their links. The rel=nofollow attribute is a general mechanism that allows websites to tell search engines not to give weight to specific links, and it’s perfect for the situation when you want to link to a site without endorsing it. Ironically, some of the most reputable links to Decor My Eyes came from mainstream news websites such as the New York Times and Bloomberg.

There are several possibilities here. One is that my specific searches for eyewear brands (the ones mentioned in the NYT article) failed to find DecorMyEyes, but some similar searches would have succeeded. Another is that the NYT article described a state of affairs that was true several years ago, but no longer obtained (for any of the major search engines) by the time the article was written. ]



8 Comments

  1. Chris said,

    November 28, 2010 @ 3:11 pm

    I'm not sure I would want sentiment analysis to affect the basic ranking of results, but rather be incorporated into the results (bad sentiment results are coded red, or something like that). It just seems to me that, in principle, ranking should be agnostic with respect to how people feel about a site. But I realize that the whole ranking process is fraught with philosophical perils (i.e., what exactly constitutes a "good" match?). Sentiment analysis just adds more peril (i.e., what exactly is "negative" sentiment?).

  2. Ken Brown said,

    November 28, 2010 @ 4:11 pm

    For some years the top Google hit for "Mild Cheddar" was a sort of very negative and (I hoped) mildly humorous rant by me. Might still be. Not anything by a company that sells cheese. Might still be. Never worked out why.

    [(myl) Language Log was once #1 for "stupid ideas". On the other hand, neither you nor we were trying to sell people overpriced counterfeit merchandise.]

  3. Lucy Kemnitzer said,

    November 28, 2010 @ 4:40 pm

    Ken Brown: it still is, apparently. I checked:
    http://www.google.com/search?q=mild+cheddar&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a

  4. Helen DeWitt said,

    November 28, 2010 @ 4:49 pm

    Ken Brown! Ken Brown! Someone is RIGHT on the Internet!

  5. Evan said,

    November 28, 2010 @ 11:20 pm

    Before I trust a new e-commerce site with hundreds of dollars I spend the 15 seconds necessary to search for user reviews of said site. These people are aware enough of user review sites to use them to complain after the fact, but not aware enough to use them beforehand? Does the lure of a deal too good to be true tempt them to bury their head in the sand and/or willfully deceive themselves?

    I do think a search engine could help solve this dilemma*, but why try to use sentiment analysis when it already has the apparatus in place for direct scraping of numeric user ratings?

    *if needing to act like a responsible adult is a dilemma

  6. harry said,

    November 29, 2010 @ 12:42 pm

    Sounds like getsatisfaction.com and the like should learn about the "nofollow" value in html.
    http://en.wikipedia.org/wiki/Nofollow

  7. Adam said,

    November 30, 2010 @ 6:46 am

    @Ken Brown:

    You're still right about the cheese.

  8. Pomplemoose said,

    December 1, 2010 @ 1:19 am

    It should be very easy to implement some form of sentiment analysis, to be applied only to sites with an internet commerce component, and wherein the 'sentiment' analyzed is drawn only from certain trusted consumer advocacy sites. The concern in the article over the White House disappearing from Google is a complete red herring.

RSS feed for comments on this post