Google’s machine

After years of taking a fairly laissez-faire attitude toward googlebombing, Google has now taken action to stop the practice. It has incorporated into its search engine a googlebomb-sniffing algorithm that somehow identifies and neutralizes any concerted effort to skew search results for a given phrase (such as, most famously, tying “miserable failure” to the White House site). That’s fine with me. It’s like scrubbing graffiti off the side of a subway car.

But that’s not the whole story. What’s particularly revealing here is Google’s explanation for why it has finally decided to tweak its search engine to defuse googlebombs. Did the company act out of a desire to present better search results? Nope. To counter the willful distortion of results? Nope. To serve the public interest? Nope. The driving reason is that the company had come to fear that googlebombing was tarnishing its image. As Google’s Matt Cutts explains:

People have asked about how we feel about Googlebombs, and we have talked about them in the past. Because these pranks are normally for phrases that are well off the beaten path, they haven’t been a very high priority for us. But over time, we’ve seen more people assume that they are Google’s opinion, or that Google has hand-coded the results for these Googlebombed queries. That’s not true, and it seemed like it was worth trying to correct that misperception.

So the company is allowing its concerns about its public image to influence its search results. The upshot in this case may be salubrious, but what kind of precedent is being set here? And, perhaps more important, what does it tell us about what’s inside the Google black box that determines how most of us find information on the Web most of the time?

Three years ago, when Google was first asked about googlebombing, it gave, as Danny Sullivan notes, the corporate equivalent of a shrug. A spokesman said, “We just reflect the opinion on the Web, for better or worse.” Google’s search engine was, in other words, just a passive feedback machine that reported the people’s wisdom – or stupidity – back to the people. It was resolutely democratic, with all the strengths and flaws of democracy. Google itself had little control over the machine it had built.

That perception of Google’s search engine – that sense that “we the people” control its workings – continues to hold sway among the public. But while it may have been true once – and while it may in fact have been the company’s founding ideal – it’s not true anymore. Google’s engine is a meticulously hand-crafted, continually optimized machine that does precisely what Google instructs it to do – even if that means filtering results to protect the company’s reputation. Google may have good in its heart. It may, for the time being anyway, be fighting on our behalf against the forces of distortion that it has unleashed. But let’s not forget that Google’s machine is not our machine. It’s Google’s, for better or worse.

6 thoughts on “Google’s machine

  1. Dragos

    Gradually, Google’s search engine is starting to be perceived more and more like the Windows of net.

    Although proprietary technology has key benefits in the end it’s lack of transparency when used by large communities becomes a road block. After numerous legal battles Microsoft accepted to present the Windows source code to important partners (mainly governments). Soon it will be Google and Apple turn.

  2. Skip McCoy, American

    Not our problem attitude, until it makes us look bad? If Google is to the quality and safety of its results what GM circa 1972 was to the quality and safety of its cars, what does that say about Google’s future?

  3. Alan Levine

    Google bombing is small potatoes. My spam watcher on my educational technology blog snags daily 300 comment spams featuring thousands of links for pills, porn, casinos, mortgage scams, people doing bizarre things with animals, shoes for sale on ebay– Google’s “machine” provides the monetary incentive that makes link spammers inflict this on hundreds of thousands of small time bloggers.

    “Do No Evil?” Pffffft. I see fruits of Google’s evil every day.

  4. tmcmh

    If spammers try to exploit the automated nature of Google search results, we expect Google to resist that. Why would we not expect the same of Google in regards to Googlebombing? They’re both efforts to exploit the automated pageranking algorithms to distort the ‘real’ Web (whatever ‘real’ means in this context).

    The main reason we expect this is because we expect Google to protect its reputation. It’s Google’s reputation for useful, reasonably representative search results that keeps us using them. It’s bizarre to think that a reluctance to “look bad” is somehow a poor strategy, or unethical. A proper care for one’s reputation (as an individual or as a corporation) is to be expected.

  5. George Johnston

    Based on the popularity Bush’s bio being the first result of the miserable failure search I would argue it was also the most relevant result. They let it go for years and within a few months of starting a PAC they disabled it. I believe it is another instance of Google’s China Syndrome.

    They didn’t get rid of googlebombs. Less famous googlebombs still work.

  6. Matt Cutts

    Hi Nick, this is Matt Cutts, the author of original post. I stopped by to comment a few days ago, but my Firefox ate the paragraphs that I had drafted. After I saw the Guardian article, I thought I’d stop back by.

    In your post, you say “Did the company act out of a desire to present better search results? Nope. To counter the willful distortion of results? Nope. To serve the public interest? Nope. The driving reason is that the company had come to fear that googlebombing was tarnishing its image.”

    As someone with first-hand knowledge because I worked on this algorithm change, I don’t think your statement is accurate. All of those reasons played a role. It’s true that a reason mentioned in the post was to correct a persistent misperception that we saw from some users, but I was drafting a short heads-up blog post, not writing a lengthy background-and-policy piece that included a full history of Googlebombs.

    If I were writing a longer post, I would have discussed how the first Googlebomb became widely known in 2001 (“talentless hack”). Back then, Google was a few hundred people, and putting resources into Googlebombs would have been a mistake. In 2001, it was much more important to work on ramping up Google’s index size, scaling the company, improving core quality, etc. There are less than 100 well-known Googlebombs, so looking at the potential impact, we judged that it wasn’t worth targeting Googlebombs back in 2001 compared to other goals, and I think that was the right call.

    I’ll close out by noting that I don’t think listening to feedback is automatically a bad thing. I think one of Google’s strengths is that we listen to our users and try to respond. A good example is our webmaster console, where the features that we provide are driven to a large degree by what we hear on blogs, at conferences, and in person from webmasters.

    I believe that if Google didn’t keep an eye out for comments in the blogosphere and elsewhere, we would miss many chances to improve. So listening to users is not by itself a bad thing. I think that Google would suffer if we only ever worked on our own projects without listening to the outside world. I’m sorry if my blog post was unclear, but misperceptions by some of our users was certainly not the sole reason behind the change. The other reasons that you mention certainly did play a role in the desire to improve our handling of Googlebombs. Thanks for the chance to clarify that.

Comments are closed.