Link Rich get Richer

by Krishna on February 4, 2007

The Hawthorne effect is the principle that “when people are observed in a study, their behavior or performance temporarily changes.” As Google Search has become the de facto search engine, webmasters and web designers have changed their activities to meet the requirements of obtaining higher ranks in Google search results. Mostly, strategies have revolved around getting as many incoming links as possible.

One of the key components of the Google search engine is the PageRank algorithm. In Google's words,

PageRank relies on the uniquely democratic nature of the web by using its vast link structure as an indicator of an individual page’s value. In essence, Google interprets a link from page A to page B as a vote, by page A, for page B. But, Google looks at more than the sheer volume of votes, or links a page receives; it also analyzes the page that casts the vote. Votes cast by pages that are themselves “important” weigh more heavily and help to make other pages “important.””

In day-to-day life, this concept will be familiar to most people. Say, there was no Internet and everyone that we know were equally accessible (without constraints of time, money, distance and social status). If we were in college and we needed accurate information about the rings of Saturn, we would first contact the head of the Physics department. If he was not available, we would contact a lesser professor, followed by a teaching assistant, high-grade student, medium-grade student, etc. Why? Because each person in this pecking order would point to the higher authority as the first line of authority.

Now, if the Physics head had designated a student, John Smith, as an authority on Saturn. Then more queries would be directed to John. If a person below the Head had recommended John, he would still get more attention than when not recommended, but less than if the Head had recommended him.

So far, so good.

At this point, let us muddy the waters a little bit. Let us say that another student Mary Jones joins the school and is more knowledgeable on Saturn than John. The problem here is that her incremental knowledge on the subject is not of interest to most students. Hence the Head has no reason to change his recommendation of John. Other students also continue to recommend John, although they may benefit from Mary’s extra knowledge, because the main authority, the Physics Head, has not changed his mind.

Applying this to the web link structure,

  1. A better, newer authority (say “A”) on a subject has fewer incoming links compared to the older authority (say “Z”) because it has just started. Pretty obvious.
  2. A” will get fewer new incoming links compared to “Z” because the incremental gain is not important to the majority of users.
  3. If a person wants to link to an authority on the subject, “Z” will get the new link because “Z” appears first on the search results. So Z starts hoarding more links, making it more difficult for A to displace Z.
  4. Z” has a significant headstart over “A” in terms of web pages and web links. The only way that “A” can overtake “Z” is if “A” can consistently outperform “Z” in SEO (Search Engine Optimization) activities. This is not guaranteed. For all we know, “Z” may be more web and SEO-savvy than “A”.
  5. A” is very likely to abandon or pay less attention to the subject matter as it continues to sport lower search rankings and attract less visitors than “Z”.
  6. As competition from “A” & others becomes less, “Z” is also likely to reduce efforts to maintain the lead. Practically, this means producing just the necessary level of content to keep up the reputation.
  7. This does NOT mean that “Z” will sit idle while someone slowly builds up momentum. With tools like Google Alerts, it is very easy to stay on top of keyword searches to be aware of competition.

This is a phenomenon, not necessarily a problem. The first reasonable authority on a subject has a first-mover advantage that is very difficult to shake off. As that authority tops the ranking, more and more people link to it, making it a de facto “top” authority. Yes, there will be people who break the trend and link to other authorities (who are lower on the search results), but statistically speaking, they can be discounted. Very soon, the first authority has amassed a fortune of incoming links and can rest on that.

The other aspect of link fortunes is the issue of “sub-authority”. If I am the authority on the solar system, it doesn’t mean that I am the authority on Pluto. But with my first-mover advantage, I have enough links to make myself the first result on Pluto and then gather more links. You might say, “That is fine, but there will be Pluto experts who can produce more content than you do. And soon they will rise to the top.” So let me explain with an example.

Let us take Wikipedia. In most simple searches, the Wikipedia article will be the first result or at least on the first page of results. I have great trust in Wikipedia pages and hence my initial reaction (which I usually follow) will be to link to the Wikipedia article. Unless I myself am an expert in that topic, I cannot differentiate between a Wikipedia page and another web page on the same subject or issue. Because of my trust in Wikipedia (coming from other sources) and the fact that Wikipedia is high on the results already because of PageRank, I throw another link at Wikipedia. Suddenly, Wikipedia has one more recommendation for its page which reinforces its position in the search results.

As I mentioned before, this need not be entirely a bad thing. The greater authority’s benefit comes from the incremental knowledge that it brings. By recognizing the de facto authority on the Web, this website or page can concentrate on more specialized information that will benefit expert audiences. The current authority can link back to the niche or expert sites. Essentially, this will reduce duplication of content.

The point I was trying to make is that a first-mover advantage on a subject on the Internet coupled with a good understanding of SEO techniques is very difficult to trounce. The first mover quickly gathers links which in turn fuels additional links to the site. One suggestion is to perhaps have a “Random Results” box in the Google search results that are based on text content and have no relation to incoming links.

{ 1 comment }

Robert February 6, 2007 at 10:29 am

That certainly is a disadvantage with the Google PR (page rank) system. New sites that contain more valuable information but a lot less incoming links are often hurt by this. In this case, it's the older content that is more "valuable" in the eyes of the search algorithm, even though it might actually be outdated.

However, most information doesn't change quickly... and if it does, most of the high ranking sources usually stay up-to-date.

Blogs presented the biggest challenge because the best content was the newest content. In order to solve for this, search algorithms had to look at the blog source (author, number of people who read the blog, the typical content of the blog, etc.).

Comments on this entry are closed.

Previous post:

Next post: