April 30, 2009    Business

Let’s Give the Other Search Engines a Chance

The artist Jenny Holzer said in her work: “The abuse of power comes as no surprise.” I share the same view. In this essay, I want to persuade you to use search engines other than Google because it could come as no surprise that Google’s dominance in the search engine market leads to abuse. Don’t get me wrong: I love many of Google’s products, and I use them every day. I’m not bashing Google as a company. My argument has to do with their dominance in the search engine market, and why it is in our own interest to help others be competitive with Google. Google essentially has a monopoly in this market and a monopoly can happen even if a company has no intention of monopolizing or have no predatory business practices. English as a language, for instance, has a monopolistic power, because our desire to have a universal language is strong. Nevertheless, a monopoly is a powerful and dangerous force because the abuse of that power can happen almost unconsciously to the holders of the power.

I would like to give you an example of a worrisome trend from Google. But before that, I need to explain some basic things about how Google measures significance of websites and pages. They use links as a vote of confidence. Their assumption is that if you link to some website, it must be because you like it. This leads to the next assumption that if a website has a lot of people linking to it, it must be important. So, Google gives more significance to the content of this website than it gives to a site to which nobody is linking. This is not a bad assumption. Suppose we take as an example an academic paper written by someone 10 years ago when people were using “hyperlinks” simply as a way to reference something easily and quickly. Some references (hyperlinks) may point to examples of evil corporation, bad writing, racist website, terrible design, etc., but I would imagine that a larger number of the references would be for something good, positive, superior, or exceptional. So, even though there is some room for errors, for the most part, this assumption worked in yielding more relevant results.

But the trouble began (in my opinion) when this medium called “search engine” became so powerful that their original model was turned on its head. This is a phenomenon described by Marshall McLuhan as “the reversal of the overheated medium.” When Google first came up with their algorithm for figuring out how to deliver more relevant results, they were observing the Internet to find the answer. They were looking at how the Web was naturally expressing relevance and significance, and they saw hyperlinks as a good way to capture it. But now that this medium has overheated, the cause and effect have flipped. Everyone is now conscious of how Google measures relevance and significance and is trying to conform to their ideas of relevance and significance. In other words, Google is no longer trying to figure out what’s significant or relevant, but they are telling us what is significant and relevant.

Do you know what “nofollow” is? Most people who are not webmasters or SEOs (“search engine optimization” professionals) have probably never heard of it. I myself didn’t know about it until recently. “nofollow” is an attribute that you can attach to your link which tells Google that it should not count as your vote of confidence. This in itself is not a bad thing, but what is troubling is that the failure to put this “nofollow” on your links can result in penalty, and when Google decides to penalize you, the damage to your site can be devastating because of their dominance.

Suppose you have a popular blog about restaurants in New York, and let’s suppose that you don’t know what “nofollow” is. One day, your friend launches a website for the preschool he and his wife own. As a token of support, you put a link on your website to their website, and encourage your audience to check it out. For doing this, Google could down grade your “PageRank” (the measure of significance) to zero and stop sending visitors to your site. Why? Because you forgot to put “nofollow” on the link. Why would Google do this? Because they see this as a way to manipulate their algorithm. They have different criteria for determining this but one of them is that this link is not relevant to the theme of your website. You are penalized for not knowing what “nofollow” is. Now, this new trend has webmasters from around the world scrambling to make sure that their websites meet Google’s new rule. If you are not a webmaster or SEO, and manage your own website, good luck.

This “nofollow” idea came about because many shrewd website owners started selling links on their websites, many of which had no relevance to the main content of the page or the site. It is a way to pass so-called “link juice” to other websites and get paid for it. This practice became widespread and resulted in poorer search results. To combat this problem, Google decided to penalize those who sell links, but exactly how would they figure out that you are actually selling the links for the purpose of manipulating their PageRank? There isn’t really a good way to determine this algorithmically. So, they have implemented a way for people to report paid links, but this too can be abused. For instance, in the example above, the owner of a competing blog about New York restaurants could report your link (the one referring to your friend’s preschool website), and Google will penalize you for that. Now your competitor will be more popular because his pages will come up higher on the search results.

“Hyperlinks” were originally like the footnote numbers you assigned to specific words and sentences in academic papers. It was a way of making references. So, if you were writing a paper about the unethical practices of Corporation X, you might “hyperlink” to the specific pages of their site as evidence/proof of your accusations, but this would not mean that you were giving them a vote of confidence. You could write a critique of something that you felt was poorly written and linked to it, but again, you linked, not because you approved it or liked it. Now, thanks to Google’s PageRank algorithm, this way of using “hyperlinks” have all but disappeared because we fear that we might get penalized for linking to sites we don’t like. If you are not aware of the “nofollow” flag and are still making this type of references/links all the time, you could be in big trouble.

As you can see, because of Google’s PageRank algorithm, hyperlink—what used to be an innocent device to make a quick reference—has become a cause for panic, an object of envy, conflict, greed, and bitterness. If you go visit Google’s discussion forum for webmasters, you will see many people panicking, frustrated, worried, angry, and feeling helpless.

A fairer solution to this problem would have been to implement the opposite of “nofollow” where all external links are assumed by default to be “nofollow”, and if you specifically want to give someone the “link juice” add the attribute “juice”. This way, you can add a link to your friend’s business website, or even add a paid advertising, just as you have always done without adding any extra attribute to your links, and you would not be affected. And, if someone specifically adds the “juice” tag, it would be clear that he is perfectly aware of its implication to PageRank. There would be no innocent victims.

Why wouldn’t Google do this? Because if they did this, their algorithm would stop working (at least not as well as before). Since most people are not going to bother putting the “juice” tag, the links will stop working as a way to measure the significance of web pages. But this is fine. In the world of free market capitalism, we should let the competitors (or Google themselves) come up with better solutions that would not require every website to have an SEO . But Google is so dominant in the search engine market that they can actually set the rules. Instead of changing their algorithm, they can force all the website owners to play by their rules. This, to me, is a worrisome trend, and it could get worse.

You might argue that someone has to fight these link Spammers and rogue SEOs, but let’s not forget that Google created these people. We have to deal with these people only because Google created opportunities for them. If Google’s algorithm did not use links as a way to measure significance, they wouldn’t be going around leaving spams in blog comments. There wouldn’t be much value in selling irrelevant links. This is a mess that they created themselves.

What disturbs me further is that, now because of their need to protect links as a gauge for significance (because they rely on it), they are trying to define what “relevance” is. If you have a link to a website that is unrelated to the main content of the page, they flag it as a potential PageRank abuse. But by doing this, they have to make an assumption about what “related” or “relevant” means. If your article is about fishing, and if you have a link to a fishing equipment store, you could get away with it. But if you have a link to a computer store, they might flag it as a potential PageRank abuse. Now, let’s think about what this really means. As McLuhan has said, the medium is the message. What is the message of this medium?

Google’s algorithm is forcing their definition of “relevance” on everyone on the Internet. Google being essentially a monopoly, we now have no choice but to play by their rules. I would argue that “relevance” should not be defined solely as being “on topic” in a literal sense of the world. Stumbleupon.com is a good case study. They try to present you websites that you might like by comparing you with others who appear to like similar things. When you keep hitting “Stumble”, you often get pages that appear to be completely unrelated, but they are not. The reason why Stumbleupon is popular among certain people is that, despite the apparent unrelatedness, there is commonality in terms of personal values and perspectives. This is what “webring” used to offer before we started worrying about PageRank. Webring originally was a way to connect your website with those of your friends, although they might not be related in the subject matter. Now webring is essentially a dirty word, and we are discouraged from doing it. Why? Because the shrewd people started using the same scheme to manipulate PageRank. Google deems the webring scheme as manipulation of PageRank, but it became so only because of the way their algorithm works. Again, if they didn’t use links as a way to measure significance, we would still be free to do webrings. There are obviously significant value in being able to browse through unexpected things that are connected by personal/human values and perspectives. Otherwise, Stumbleupon and social networking would not be so popular. When we read the “status update” on Facebook, or our own Twitter home pages, we see a collection of links that are completely unrelated topic-wise, but we still enjoy them very much. This is because they are links posted by our friends. The common thread is our own personal values. By enforcing Google’s definition of “relevance”, they are penalizing others who have different opinions of what “relevance” means. Soon enough, we could be forced to stay “on topic” on our Twitter and Facebook pages, and we’ll all have to have a very narrow interest in our lives, or risk becoming a social outcast.

Google search engine is a powerful medium which is transforming the way “relevance” and “significance” are defined in our culture. This is the “message” of the medium, Google search engine, and it is a potentially dangerous message since it has a bearing on our own relevance and significance as individuals. This is why I would like people to consider supporting other search engines. As we saw with the current economic crisis, the idea of “too big to fail” has proven disastrous. We need competition like the way we used to have in the search engine market. I think many people now use Google just because everyone else does, or they might be afraid of appearing stupid for using something presumably inferior. I personally cannot detect any measurable or tangible differences between Google and Yahoo, but that is not the point I want to make here anyway. Even if other search engines are indeed inferior, we should help them improve themselves. More importantly, we need to create an environment where small companies with creative and innovative people can come up with competitive algorithms. Google has now become so big and powerful that they have to spend at lot of money and effort in protecting what they already have. This can stifle innovation and competition.

As I said at the beginning, this is not about bashing Google as a company, this is about encouraging people to foster a competitive and innovative environment, and about protecting ourselves from a monopolistic power that can dictate how we should think and value. If you see any merit in my argument, please give other search engines a chance.

UPDATE:

I just discovered that Google themselves forgot to put nofollow on their own promotional links. This shows how easy it is for someone to make this mistake. If the company that created this nofollow rule can make this mistake, just imagine how many website owners must be getting unfairly penalized by Google, especially those who don’t know anything about SEO.

Google has quality guidelines that suggest we should ask: “Would I do this if search engines didn’t exist?” I certainly would not bother adding “nofollow” to any links if the search engines didn’t exist, but Google is forcing us to contradict their own guidelines.