Comments On Response To SEO Questions By Rand Fishkin

I just finished reading theGoogleCache’s response to seo questions by rand fishkin and have to appreciate the time and effort that went towards running the experiments necessary to prove their points, however like most SEO experiments this one is flawed to an extent. First, I’d like to summarize the results as I do believe most are correct in nature, however, are not applicable in the majority of real world examples.

  1. The Diminishing Value of Anchor Text
  2. The experiment proved that subsequent unique anchor text linking to the same page carried less weight.

  3. How Far Does Synonymy Go?
  4. The experiment proved that Google gives weight to mispelled pages when the mispell is not stemmed from the correct spelling. I agree with this result.

  5. Can Link Removals Hurt Rankings?
  6. The experiment proved that this statement is true as pulsing an important link from Wikipedia may knock you in and out of the SERPs. I agree with this result.

  7. Does Sharing Registrants with Spammers Hurt You?
  8. The experiment proved the answer to this is no.

  9. How Much of the Original PageRank Formula Still Matters?
  10. The experiment left this one inconclusive, however I do agree that the relationship between PR and ranking is synergistic and dependant on other factors such as domain age.

  11. Yahoo!, MSN & PageRank
  12. The experts replied with “Who cares?”. I agree with this one 100%. But on a little bit more of a serious note, each engine ranks sites very differently and I believe this will continue to happen for the years to come.

  13. Text Placement Weighting
  14. The experiment proved that the placement of text has a bearing on the weighting put on it. I agree with this and can add that text above the fold makes a difference in the keyword weighting.

  15. Higher Links = More Weight
  16. The experiment proved this to be true, and funny thing is I noticed this experiment in Google along time ago. It worked wonders for linkbait and agree with the experimental result to an extent.

I would like to comment on the experiment itself. Logically, the techniques used to complete the experiment seem reliable, however in many cases, as the diminishing value of anchor text test, the anchor text and links are incoherent jumbled letters, not actual english words. Well, how often are you trying to rank for rtgfghjfd? Then you might counter my argument with “Well, it doesn’t matter because in theory whatever holds true for rtgfghjfd holds true for my keywords”. Wrong. Google is smart, very smart.

First, Google has certain keywords and key phrases flagged so that a specific algorithm will rank web sites in those areas (ie. Travel, Gambling, Hotels, Pills). Secondly, they set threshold values for most of the variables of the ranking formula, and exceeding the threshold may harm your website rather then benefiting it. Therefore, if a very high percentage of incoming links are top links of other websites, Google is most likely going to humanly examine your website and maybe even flag it. That is why the results of the experiment mentioned above are considered true in general, but can’t be set in stone, because had half your incoming links appeared in the middle area of websites, you may have avoided being flagged.

Add this article to your favourite bookmarks:
Del.icio.us
Del.icio.us
Digg
Digg
Reddit
Reddit
Co.mments
Co.mments
Windows Live
Windows Live
Yahoo My Web
Yahoo My Web
Stumble Upon
Stumble Upon
Squidoo
Squidoo
Google
Google
Newsvine
Newsvine
Netscape
Netscape
Technorati
Technorati
Furl
Furl
Ask
Ask
Ning
Ning
Ma.gnolia
Ma.gnolia
Spurl
Spurl
Raw Sugar
Raw Sugar

No Responses to “Comments On Response To SEO Questions By Rand Fishkin”

No comments yet

Leave a Reply