Does The Second Link From A Page Pass Value? – SEO FAQ

Author: Gab Goldenberg

A little while ago, Willy Franzen of One Day One Job was doing a 1-hour phone consultation with me, and he asked whether a second link from any given page would also boost the page’s ranking.

My answer was that the second link would pass anchor text relevance as well as pagerank/other markers of ranking value. I was speaking instinctively rather than scientifically from an a-b test experience, and today I want to share 2 theories why that should be so.

But first, for those not up with the blogosphere’s every rumination, let me provide some context. My friend Branko Rihtman has a good history of the whole second link debate, and contributes some original research himself. Other relevant posts come from Rand, Mike van de Mar, Dave Eaves and Michael Martinez. Some partial information was also provided by Matt Cutts to Debra Mastaler.

But rather than pay attention to the tests that are / are not valid (depending who you believe) I thought I would ignorantly pontificate in the total absence of any study to back me up. All I’ve got is a pair of logical theories on this, so take this for what it’s worth.

Clicky web analytics screen resolution Attribution-ShareAlike License by adria.richards Theory 1 – Prominence / Likelihood to Drive Traffic.

Google was founded on a “random walk” algorithm – a person continuing to click links ad infinitum, randomly, would get to sites with more links and more prominent links, more often. So any additional link, should pass both anchor text value and ranking factor value.

That said, Matt Cutts specified in the post linked to above, that typically the second link wouldn’t pass anchor text weight – if the anchor was identical to the first one. That makes sense, since all else being equal, two identical links should get the same click through rate, and not be materially valuable to the random walk theory.Adsense featuring successful copy

However, the random walk is mostly dead, at least to Google. After all, adsense ads don’t pass ranking value, and they’re prominently visible throughout the web, sending millions of visitors all over the place!


Theory 2 – Content Consolidation Via 301s.

In my latest guest post at Graywolf’s, on SEO for WordPress blog categories and ecommerce site categories, I referred to consolidating content and 301ing it, as well as using a table of content a la Wikipedia to help people navigate the new, longer page.

Well, suppose that each post you consolidated into the main one got an external link from another site’s page. E.g. Some blogger is thinking “I found this great series of posts on social media at SEO ROI Services. I should share each of those with my readers!”

Are each of those links no longer going to count, once all the links’ destinations are now consolidated on a single URL? Just because every page got rolled into one, doesn’t mean each link should stop passing value. Besides, if it did, then SEOs would just get craftier and break content out across multiple URLs.


Sidebar Story


  1. I remember Rand talking about this, and I think it's test time Gab. I'll let you know the results!

    Comment by Lindop - March 19, 2009 @ 6:02am
  2. You probably won't get the second link to pass value unless you use a well-established, trusted site to do the test. I believe (and other people have suggested or implied) that I was able to get a second link to pass value because I used my own personal site (Xenite.Org) for the test. Xenite has been around since 1997 and is a trusted domain with lots of history and link reputation. Personally, I don't recommend that people try to use double-linking. The SEO community puts way too much value on links and needs to look at other aspects of optimization.

    Comment by Michael Martinez - March 19, 2009 @ 1:22pm
  3. David, that would be awesome - and great linkbait for you, I'm sure :). Please let me know when the results are in!

    Comment by Gabriel Goldenberg - March 19, 2009 @ 6:35pm
  4. Michael, I'm flattered to see you here, and thanks for sharing your thoughts. The possibility that this is a trust-threshold thing makes a lot of sense, and would be a possible problem. is also pretty trusted, so I can try and test it vs David's test on a new domain, and see what comes of it. We do put a lot of emphasis on links, you're right. But if you see my posts on brand building and the evolution of link buying in 2009, you'll see that I also look at bigger picture considerations like demographics and traffic :). Thanks again for sharing your knowledge - I'm looking forward to more!

    Comment by Gabriel Goldenberg - March 19, 2009 @ 6:38pm
  5. How did i miss this one ? Probably because it didn't send enough visitors to my site :))) I'm not sure I get the second point. The fact that you are consolidating a lot of pages into one means that you will be getting a lot of links that used to point to different pages to this one single page. So how is that similar to the situation where you have two links from one page pointing to the same target page ? Another thing is that you are talking 301s here. A whole different animal than straightforward links. BTW, my two-link test is still up and running and the second link is still not passing anchor text relevance to the target page, despite the fact that the first link has been nofollowed for months now. This has brought up a whole bunch of other questions, mainly about the situations in which google will ignore their usual practice of indexing/ranking and behave in a way that contradicts what we think we know about the way it usually behaves. Tests running, will write :))

    Comment by Neyne - March 24, 2009 @ 8:43am
  6. Oh, what i meant is that the first link IS passing anchor text relevance, despite the fact that it is nofollowed, while the second link is not passing anchor text relevance.

    Comment by Neyne - March 24, 2009 @ 8:51am
  7. Fascinating insights there Branko, and thanks for sharing your research! The second point is exactly as I said. The context is just that the links all come from one page (i.e. the links to the various pages that are now being consolidated). You make a fair point about 301s, which may affect the value.

    Comment by Gabriel Goldenberg - March 24, 2009 @ 4:55pm

Leave a Reply