An Analysis of Hubs and Search Engines

Hubs and search engines join together in a symbiotic relationship. For many hubbers (authors), seeing their work appear "high" in the search engines represents a significant achievement. The never-ending struggle for more traffic (readers) begins and ends with Google, Yahoo, Bing, and numerous lesser indexing web sites.

This article examines the relationship between one hub and several search engines. The results may be surprising, given the prevailing wisdom regarding search engine optimization (SEO).

The hub under consideration is titled Coke vs Pepsi. It was published on October 12, 2009. According to the HubPages 'Stats' tool, it contains 600 words. It includes three color images and one comment widget. At the time of this writing the comment widget contained 16 comments; all but one of the comments originated from registered HubPages authors. It contains no eBay or Amazon widgets, but it is configured to accept Google advertising.

Relatively speaking, the hub is not tremendously popular. It has only received 1216 hits since being published. Only two backlinks have been manually created for the hub: one on a Twitter account and one on a Posterous account. Other than those two 'advertisements', no effort has been made to get the hub indexed by any search engine.

The Search Term Data, listed in the following screen shot, enumerates the keywords that delivered traffic to the hub. Keywords break down as follows:

• Google: 12 different keyword phrases
• Yahoo: 8 different keyword phrases
• Bing: 2 different keyword phrases

Note that this is not an indication of the total keyword phrases associated with this hub. Rather, these figures represent the keywords that caused search engine users to click on the page link when it was included in the search results.

Additionally, these results should not be construed as an indication of any "top 10" finishes in the search engine results races. The link could have appeared in any results page (Google typically provides 100 pages of results with 10 sites per page), although it is reasonable to assume that most search engine users do not browse beyond the first page of results.

The author-selected keywords, listed in the following screen shot, tell a much different story. Note that almost none of the keywords that were manually added to the hub at the time of publication appeared in the Search Term Data, discussed above. Evidently, Google, Yahoo, and Bing, have specialized algorithms that extract additional keywords and keyword phrases from web pages. These keywords may be dynamic and they may have some direct relation to the manually added keywords, but additional research is necessary to prove or disprove this thesis.

This should not be construed as any indication that judicious keyword selection by HubPages authors is not necessary, but the traffic delivered to the hub is a function of keywords that originated external to the publication of the text.

Referring back to the Search Term Data, note the keyword phrase "is coke preferred over pepsi in mexico". At least one Google user entered this phrase into the search engine. Google elected to include the hub in the search results and the user clicked on the link provided in the search results. Oddly enough, the word "mexico" appears in the hub only in a comment left by a reader. The word actually appears twice in the same comment.

The implication is that hub comments are also indexed by search engines. However, another question is begged: if comments help a site get indexed, can they also be detrimental? For example, will profanity in comments cause a search engine to lower its' opinion of a hub?

In a similar vein: will comments cause Google to serve up different ads? For example, the hub discussed here is obviously slanted toward Coke and Pepsi products, but not remotely related to Mexico. Given that Mexico is mentioned twice, will Google consider this situation when selecting ads to publish? More research is needed.

Hubs need search engines and search engines need hubs, but indexing may not be as straightforward as some might surmise.

1. Research the impact of manually added keywords on the (possibly dynamic) keywords generated by the search engines.

2. Research the impact of comments on search engine indexing, keyword generation, and advertisement selection.