The question if only the first link’s anchor text to a page counts or multiple anchors per page has been tested and discussed for quite a while now by a bigger number of bright minds, (see below) but Johannes Beus shows some VERY interesting results – especially in relation to image links.

So last Monday Johannes Beus ran more tests on Google.de to find out how and if Google rates the first link text differently from the second link text.

—< Johannes’ results translated by Bernd & Christoph of “CEMPER.COM”:http://www.cemper.com >—

I elaborated on the fact that Google rates only the first link text � if there is more than one link directing from one page to another � a couple of weeks ago. In response to that post several e-mails and comments were written with special cases about that fact. I tested those cases and would like to present the results to you. All the tests are still available under online, so please feel free to check it out yourself.

Test 1: Standard case as tested a couple of weeks ago (1. Link / 2. Link)

The result is still the same � if there is more than link directing to a target page, Google rates only the first link text. Therefore it is still better to store the important keywords for the internal linking within the first link text.

Test 2: First Link = No Follow-Link (1. Link / 2. Link)

As you have seen at the first test but with the difference that this time the first link was provided/devaluated with the attribute No Follow.

The result is quite surprising: Google neither rates the first nor the second link text and additionally, Google neither crawls nor accepts the target page to the index. To make sure that this wasn’t a random mistake by Google, I waited a couple of days and attempts to crawl by the Googlebot after the admittance of the remaining sites but the result didn’t change. The focus of Google on the first link of a page seems to go so far that all other links on that page will be damned if the first link shows a link text with the attribute No Follow.

Test 3: First Link = picture; No title, No alt attribute (2. Link)

The first link was set with a picture as link text although there is neither a title nor an alt attribute within the picture. With this combination, Google rates the second link text.

Test 4: First Link = picture; No title, but alt attribute (1. Link / 2. Link)

As in test 3, but this time the alt attribute of the picture was implemented. The result is the same as in test 3. Google rates the second link text and ignores the alt attribute.

Test 5: First Link = picture; With title but no alt attribute (1. Link / 2. Link)

As in test 4, but this time the title and not the alt attribute was implemented. That doesn’t change the result from test 3 & 4: Google chooses the second link text.

Summary

Johannes Beus summarizes the test in a way that we can say that especially the treatment of No Follow Links by Google might cause a bustle for website hosts & SEO people. Although it was the general assumption until now that Google would rate the second link anyway and the management of internal linking was built on this thesis, the test shows the complete opposite.

As every piece of information on how Google treats links will be tested and reviewed, so did it happen with this assumption.

Some of the reviews were positive such as Branko Rihtman, David Eaves and some of them were negative such as (who took also a closer look on the testing itself). Michael Martinez review of that topic is also currently been discussed on Sphinn.

-
-
-
-

What’s your take?