Social Links DO Matter – Especially for Social Sites

by Peter Young on August 5, 2010 · 13 comments

The post below is a guest post from Mindy Gofton from Manchester based digital marketing agency I-Com

So how much value do blog comment links and social bookmarks actually have? We’re talking about links with the rel=”nofollow” attribute here. Links that are not meant to pass any sort of PageRank value, links which have been so labelled so as to tell search engines “I don’t recommend this site – the link is either advertising or has been added by the site owner and I haven’t checked it to decide if it’s spam or not.” Some SEO commentators have humorously referred to “nofollow” as being a link condom in that putting it onto your link allows you to connect anywhere you like without having to deal with the consequences!

Theoretically, then, these “nofollow” links should be worthless which means all mentions across sites such as Twitter and Friendfeed, et. al. should be fairly worthless from an SEO perspective. So why, when we downloaded a list of links to the I-COM blog from Webmaster Tools, did Google report links such as the following:

  • http://twitter.com/SearchDesignDev
  • http://twitter.com/behappy2day
  • http://twitter.com/searchdesigndev
  • http://www.blogstorm.co.uk/black-hole-seo-dont-get-sucked-in/
  • http://www.readwriteweb.com/archives/friendfeed_hotter_than_ever_or_fading.php
  • http://twitter.com/SearchDesignDev/lists/memberships
  • http://twitter.com/loggedout/profile/SearchDesignDev
  • http://www.problogger.net/archives/2008/10/09/is-twitter-selling-links-and-what-in-the-world-is-a-promotion
  • http://digg.com/music/Online_Reputation_Management_Ryanair_vs_Franz_Ferdinand
  • http://sphinn.com/story/119344
  • http://sphinn.com/user/sphunn/philipnorton42/
  • http://friendfeed.com/searchdesigndev/34670708/google-chrome-people-are-reading-about-it-but
  • http://friendfeed.com/tinotriste/8502cea1/google-chrome-people-are-reading-about-it-but?service=digg
  • http://sphinn.com/story/109531
  • http://friendfeed.com/searchdesigndev/358d4dce/google-why-screw-up-with-indexing-and-not-admit?login=1
  • http://friendfeed.com/searchdesigndev/358d4dce/google-why-screw-up-with-indexing-and-not-admit?login=1&comment=358d4dce-65b5-ccf1-7528-887913044c4a
  • http://friendfeed.com/searchdesigndev/34670708/google-chrome-people-are-reading-about-it-but
  • http://friendfeed.com/tinotriste/8502cea1/google-chrome-people-are-reading-about-it-but?service=digg
  • http://sphinn.com/story/109531
  • http://friendfeed.com/searchdesigndev/358d4dce/google-why-screw-up-with-indexing-and-not-admit?login=1
  • http://friendfeed.com/searchdesigndev/358d4dce/google-why-screw-up-with-indexing-and-not-admit?login=1&comment=358d4dce-65b5-ccf1-7528-887913044c4a
  • http://twitter.com/Justhipper/status/13546530474

You get the idea – all of these are links marked with a rel=”nofollow” attribute whether they are blog comments, tweets, or other types of social bookmarks of our various posts.

If these have no worth, then why is Google reporting them in Webmaster Tools? It makes no sense to me why they would report a list of useless links to distract SEO consultants from reverse engineering the algorithm. I mean, why bother reporting anything at all? Why give us any sort of incentive to clog up useful social media sites with crap even further, in hopes it’ll boost rankings when it won’t? This behaviour would, essentially, be Google giving SEOs the green light to spam up good sites for nefarious purposes and I don’t see what Google has against social media sites that the company would wish this level of badness on them.

This leaves us with a second option – even those “nofollow” links have some value. They may not pass PageRank, but they are telling Google something that’s useful for them to know.

Bear with my logic for a minute here:

The major search engines created rel=”nofollow”, to stop comment and forum spam from getting out of hand. Webmasters were being inundated and by adding a rel=”nofollow” link to anything that wasn’t added to the site by the Webmaster; it made it pointless for SEOs to use this form of link building to achieve any sort of result.

Then Matt Cutts told us to also use rel=”nofollow” on advertising. Surely, by putting an ad on your site to someone else’s site it is some form of recommendation. Alternatively, to put it another way, would you put an ad on your website to a product or service that you knew was a scam? No. You wouldn’t. Well, at least I wouldn’t – nor would any of my clients. So that rel=”nofollow” is a little unfair to the sites being linked. Surely, there is some signal going back to Google, even if it isn’t factored into PageRank.

Then clever SEOs started using rel=”nofollow” to “sculpt” their own internal PageRank. Google claims to have counteracted this by making rel=”nofollow” links take a percentage of the PageRank flowing away from the other outbound links on a page without passing any PageRank or anchor text weight to the target page. So, the rel=”nofollow” has started to take on a variety of uses for which it was never intended (some of which Google has had to counteract with an algorithm tweak).

Spammers being what they are, social/user generated content sites have been slowly queuing up to add the rel=”nofollow” attribute to all external links in order to stop themselves turning into unwitting link farms.

Whether I label a link “nofollow” or not; when I post a link on Twitter, it is a recommendation; when I bookmark a link on StumbleUpon, it is a recommendation. When I link any site from an active account with lots of connections and I regularly recommend content from a variety of sites, then that’s an even bigger signal that it’s a legitimate recommendation.

Google folding in real time social results into the SERPs is another suggestion that they consider these valid recommendations. Spending time and money on this development and cataloguing all of this data makes no sense otherwise, considering the scrolling results have little value to normal searchers.

The use of citations as a local search-ranking factor suggests that Google is also starting to understand and analyse the many ways people recommend things online. Why wouldn’t they build this into the main algorithm in some way, using social mentions as a signal for quality – even if it doesn’t contribute to the PageRank section of the algorithm?

The way people use the internet and share content on the internet has changed. Everybody has a say now, not just those savvy enough to understand how to build a web page, and increasingly people are looking for recommendations for good websites, products and services conversationally through social media sites and those people don’t really care if the recommendations pass PageRank – they’re given freely between “friends” to be social.

It follows that search engines have had to change and adapt the way they rank useful websites, and that these changes will continue to affect the way we need to promote sites and the way in which we think about links and what makes them good. If Google is telling webmasters that they’re cataloguing rel=”nofollow” social links and we know they aren’t passing PageRank, then they must be passing something; so maybe it’s time to ask Google to give us some indication of why they’re reporting these links and what value they might have?

Google+ Comments

{ 11 comments… read them below or add one }

Jon Hudghton August 5, 2010 at 6:47 am

Interesting post and I do agree with your sentiments; I think people have been asking Google to answer this for a while though! A couple of years ago, Matt Cutts was quoted as saying the following:

“we don’t follow nofollow links, but we do include nofollow links in the webmaster console just to give webmasters a full picture of all the stuff that’s linking to them” (taken from http://www.webmasterworld.com/google/3630386.htm)

It’s a bit of a wishy-washy answer I agree. What we do know for certain from various SEO tests is that a no-follow link won’t pass the anchor-text as you say.

But yeah, if you’re getting lots of links from “quality” sites Google really loves (Wikipedia for example) or from Facebook/Twitter then I’d agree they are probably factoring them into the algo somewhere, even though the links are no-followed.

idietcola August 5, 2010 at 8:30 am

Our research leads us to discover that google does indeed go down a nofollow linkie – after testing we indeed discovered that the site behind the link was indexed by google after pointing nofollows to it…

Essentially a ‘link is a link’ – and social media sites are no more than popular/ authoritative websites – there has always been an interest in having your links come from authoritative websites… be them follow or not…

Mindy August 5, 2010 at 9:12 am

Thanks for the responses guys. I think it’s well within the remit of the “nofollow” attribute for Google to use the links to discover new pages even if the links don’t actually affect the PageRank part of the algorithm – this isn’t to say they don’t affect something else.

What I haven’t made a point of is that those links are social links pointing to a blog which is, by nature, a social site, so may have more relevance than they would for a page on a brochure site describing a service or something similar in that they’re more in keeping with how blogs and the social side of the web works.

I still say it’s bizarre that Google would report links they found if they’re not using that information for anything themselves. Just doesn’t make logical sense. Why would it matter that we know, unless they’re about to launch a piece of software to track reputation/buzz?

Rhys August 5, 2010 at 9:20 am

It does make a lot of sense what you are saying & I totally agree. Google I believe wanted something similar to the “citation” system used in disserations & thesises (sp?) which a few years ago meant that links were all the rage.

Now with social bookmarking & networking sites, it makes sense that “citations” are now carried out on a lot of these websites. Therefore if Google is still following the same methodology in listing the most popular web resources top, it makes sense to follow social media traffic.

It makes me think whether other factors, such as authoritative & relevant citations (say for example I ran a music blog, and was retweeted by Lily Allen) will help ranking in the future?

Jon Hudghton August 5, 2010 at 9:30 am

Yeah, Google will of course crawl a no-followed link (think they’ve stated that a few times); the issue is more whether or not that link then carries any weight and affects rankings (and by how much), even if it’s not via the traditional passing of “pagerank” or anchor text.

idietcola August 5, 2010 at 9:36 am

I wouldn’t concern myself personally with the passing of pagerank – but going along the lines of a natural link mix powers rankings… then building some nofollows into your mix is an essential part of the process….

Simply use links as they are meant to – i.e. a blog comment link should be nofollow – so searching endlessly for a comment link that is do follow – you’ve probably cost your self more time than the worth of the link…

heck – one of the primary reasons i commented here – was for a link. plain and simple. There is a bit of juicy seo to be had from here – im linking to link building – there is the word ‘link’ in the url above… for that matter this whole post is about links… so i think it has been well worth it – as for the impact of one link… good luck figuring out that. there are simply too many other factors to take into consideration.

Andy @ FirstFound August 5, 2010 at 2:24 pm

Despite what Matt says, they’re keeping that data for a reason. I agree with the blog. Nofollow or not, it’s doing something.

hydroponics November 23, 2010 at 2:51 pm

Great post thanks for taking the time to explain this to me as i thought no follow was a waste of time.

mo @ small business internet marketing February 4, 2011 at 12:53 pm

Good discussion. Fair points raised on both sides. Personally I always use a range of different link building methods when doing SEO, and will often build link on sites that use no follow.

In the end it has to look somewhat natural, and a truly popular site will always tend to have links from both nofollow and “dofollow” sites.

ken hill March 5, 2011 at 5:24 am

Great post – I wasn’t aware that nofollow sites were successful, but after reading your discussion I feel far more educated! Thanks for sharing!

Doc Sheldon July 27, 2012 at 9:20 pm

idietcola-
Google has long ago admitted that they use nofollow links for discovery, so the fact that a page gets indexed with nothing but a nofollow link to it is no surprise.

Leave a Comment

{ 2 trackbacks }

Previous post:

Next post: