On SEO and Search Mediation

(hmm. . .where did September go? And October? And, um, November?)

I’ve been meaning to write about search engine optimization for a while in this space. It’s not exactly a topic that I’ve thought much about before the last 12 months or so. In fact, it’s kind of a topic that if you had mentioned to me 13 months ago I would have rolled my eyes.

The fact is that when you work in academia, discussions about how to “market” your “content” online aren’t wholly appropriate. Talking about how to make your way to the stop of the SERPs is sort of jejune. We don’t like to think about our online intellectual output as merely “content.” And we don’t like to consider that the size of our audience really matters. After all, if we judged the worth of academic research by whether or not a lot of people were interested in a topic, a lot of other people would be in trouble. That’s definitely not a bad thing. IMHO, universities and academics should be interested in exploring and shining a light on the hidden crevices of human thought and experience.

So, like I said, SEO and academia make uncomfortable bedfellows.

But about a year ago I started to do some freelance Web design work on the side, and, out of necessity, found I needed to develop a deeper understanding of how search engines worked.

I wasn’t completely clueless about the topic. I had an interest in it, you know from an intellectual standpoint. But understanding SEO intellectually and actually practicing SEO techniques and being successful are two very different things. I’m definitely still learning and am by no means a master of these techniques. But I think I’m getting better at understanding what’s going on under the hood.

By far, however, what I think is most interesting about SEO is the way in which it underpins so much of our experiences these days — and how little we realize it.

Most people (including me) are probably guilty of not thinking very often about where those search results come from. We place some inexplicable faith in Google’s magical algorithm, believing that because it is algorithmic it is therefore unbiased and magically pure.

Increasingly, however, I’m seeing stories of people who are taken in by search engine results. For example, the NYT story yesterday about an online merchant who realized that he could get to the top of the Google hit parade even if all of the Google Juice he was getting was, in fact, generated by negative reviews and feedback. The comments on the article are interesting — there are quite a few people who suggest that the consumers who blindly bought a product from a merchant simply because he was the first Google hit are just stupid. Didn’t they know to do more research into the seller’s reputation? But the hundreds of people who’ve been taken in by this particular merchant (and those are just the ones who have shared their story online through sites like Get Satisfaction), are evidence that, whether or not it’s stupid, people don’t necessarily do this kind of research. And Google search results carry a kind of unspoken, unchallenged weight for many people.

Then, check out the story from earlier this summer about how the conversation about the Islamic Center in lower Manhattan was essentially hi-jacked by the way search engines work. Once enough sites (bloggers and main stream media) started referring to the Center using the term “ground zero mosque,” attempts by news outlets to correct the misnomer were useless. Why? Because the masses were using that particular term to search for information about the story. Even if a news organization issued a memo insisting reporters use another term, those stories weren’t as likely to be found because they weren’t using the, now popular but incorrect, term.

These stories highlight a silent part of our online interactions. Every time you search, your experience is being mediated. Just because that mediation is algorithmic doesn’t mean it is pure. In the case of the NYT story, the merchant found a way to work the system. It may not work forever, but in the short-term he seems to be reaping huge rewards. In the case of the Islamic Center, no one seems to have deliberately tried to manipulate what you’re seeing in your search results, but the effect is still dramatic.

Truly, the notion that algorithmic results can be somehow agnostic is misguided, I think. Even if you argue that sure the system isn’t perfect, but at least we’re all working with the same imperfections, you need to realize that those algorithms are the product of some kind of human intervention. Choices are made regularly about how to value the content of the Web–how to index it, how to rate it, and how to display it.

The freelance work I’ve been doing has been primarily with small businesses who rely, primarily, on local customer bases. Most people probably don’t know about the minefield that is local search results. You’ve all seen them — they’re the results that show up and are associated with a pin on a Google map. They’re displayed when Google notices that you seem to be searching for a term that is likely to have a local business presence (based on your self-reported location or the location of your ISP).

These results link to Google Place Pages — essentially mini home pages that businesses can claim from Google. Lots of businesses do claim these, but most have not. That doesn’t mean that they don’t get listed. Google generates it’s business listings not just from user submitted information but from a network of Web sites that provide local data (yellow pages, business directories, local directories). The Place Pages are automatically constructed and populated with data from these sources. Even if you claim your business and enter the correct data, it can be overwritten (or supplemented) by incorrect data from another source: Google doesn’t necessarily consider you to be the final word on your business’ information.

Why does this matter? Troll the Google Places help forums and you’ll see business after business caught in search engine hell. Listings go missing. Listings get merged with competitors’. Listings have incorrect information (like phone numbers and addresses) that can’t be fixed through any discernable means.

Because Google has become the Web resource for searching and because so many of these businesses rely on customers who rely on Google, these businesses are at the mercy of Google’s system. They can’t just opt out — because opting out means opting out of the main conduit into the online marketplace.

More and more I hear conversations (and concerns) about the power of Google in academic circles; these tend to focus on the power it has over our personal information. That’s an important conversation, but I’m not sure we’re talking enough about the bread and butter of Google–the search mediation that has come to define many, if not most, of our online activity.

I’m not suggesting that we can change the way that mediation occurs or that we should avoid it. Rather, I think that as educators we have a responsibility to educate ourselves and those we teach about this other power of Google.

UPDATE: Google has responded specifically to the NY Times article about merchants realizing that developing a negative reputation online can help them as much as a positive one. The post on the Google blog is interesting. They actually walk us through the various scenarios they explored in order to demote merchants like those mentioned in the article. Interestingly, even though they state that they always seek ways to adjust search algorithmically, the solution they seem to have hit upon sounds like it relies more on human judgement:

Instead, in the last few days we developed an algorithmic solution which detects the merchant from the Times article along with hundreds of other merchants that, in our opinion, provide an extremely poor user experience. The algorithm we incorporated into our search rankings represents an initial solution to this issue, and Google users are now getting a better experience as a result. [emphasis added]

This particular “algorithmic solution” sounds like it simply involves humans at Google identifying this and another other badly-behaving merchants and penalizing them in the larger search algorithm landscape.

IMHO, this post sounds like it is trying to respond to the specific concerns raised in a high-profile article agains a specific vendor (and “hundreds of other merchants”) while also trying to convey to users why it is that systemic approaches to these kinds of problems within Google’s algorithmic mechanisms aren’t really viable: “We’ve got no way to really solve this, but we’ll take care of this bad guy (and a few others we know about.)”

In any case, further evidence that search is not agnostic. Humans at Google’s make value judgements that effect both the larger algorithm and “one-off” cases like this one seems to be.