Koozai > Blog > Should I Remove my Backlinks?

Should I Remove my Backlinks?

| 3 minutes to read

LinksInbound links are essential to any SEO strategy. They help refer traffic and build authority around your domain. However, is there such a thing as too many links? Or even the wrong type of links?

Businesses that live and die by search engines are completely obsessed with building backlinks. However, time and time again I have seen companies focus on quantity rather than quality. This leads to irrelevant links which in turn holds back the performance of their campaign.

The best way to ensure a valuable link profile is to build good quality links back to your domain. How do you know if a link is good quality? Review, and then review again. If you do spot inappropriate or potentially damaging links it is worth removing the link and monitoring the progress of your site within the rankings.

There are a range of tools to help review your backlinks. You can review key factors including PageRank, age of domain and even drill down to the IP address where the link is hosted. Below are a few areas I would suggest reviewing to ensure your link profile stays squeaky clean:

Irrelevant Links

Ideally you always want backlinks to be as relevant as possible. If you run an online bike shop you want to have links related to the bike industry. For example, an online bike shop won’t want links from a blog within the finance industry. If you do spot a pattern of unrelated sources I would encourage you to remove the link as soon as possible.

Links on the Same IP

Numerous links on the same IP address can look extremely suspicious to search engines. Whether the link is coming from a blog, directory, article or forum you need the links to be coming from separate IP addresses. Too many links to a domain with the same IP address is referred to as backlink bombing. The quickest way to review your IP referrers is to download your whole link profile and sort by IP address. If you have hundreds of domains linking to your website from the same IP address this will cause concern among search engines.

Bought Links

Buying links is a hotly debated topic and while it can gain quick results the outcome is rarely sustainable. If you have bought links in the past I would encourage you to test the value by removing them from your link profile. At first you may see a slight decline in rankings but the results will soon return once you take a more holistic approach to your linkbuilding.

Sitewide Links

Sitewide backlinks can often give signals that the link has been purchased. If you have attained a link from an appropriate domain it doesn’t need to be placed on every single page. Choose the most appropriate URL and implement appropriate anchor text. If you‘re doing some paid banner advertising with a specific domain then by all means have the link sitewide. However, if the sitewide link is a text link or a directory link you could get penalised.

Reciprocal Links

Reciprocal links are absolutely fine in moderation. If you have a business partner by all means give them a link from your website and ask for one in return. The problem occurs if you have extreme reciprocal linking patterns. For example, if a clothes shop has a reciprocal link directory which links to all their suppliers each of whom then link back.

I would also avoid reciprocal link farms. There are a number of third party providers which will manage the reciprocal linking process for you. This is a process where a third party will include you in a network of sites which will interlink to each other.

Anchor Text

You may have a superb variety of backlinks pointing to your domain but if the anchor text is spammy it will hold back your campaign. The best way to optimise links is to focus on what is called the ‘anchor text sweet spot’. This is where the anchor text will include 30% exact match, 40% brand name and 30% keyword variations. For example, if you were targeting the key phrase ‘surfboards for sale’ the anchor text for that specific page would be:

  • 30% ‘Surfboards For Sale’
  • 40% Brand Name
  • 30% variations i.e. ‘Surfboards for Beginners’

If you haven’t followed this approach it may be worth tweaking your anchor text to display a more natural variation.


In my experience when it comes to links, less can definitely be more. Time and time again I have seen search engines reward domains for cleaning up their backlink profiles. If it looks unnatural it will hold you back. Tidy up your backlinks today and reap the benefits of a sustainable link profile.

Image Source

Chrome chain with a red link via BigStock


  1. Michael avatar

    Hi to everyone on this great thread!

    I am a massive advocate of testing, or at least controlling as many vaiables as possible online. I prefer to base my opinions on direct experience -which agency work is fantastic for – but with so many variables to consider however, I think the fairest assessment of results is when first-hand knoweldge is considered alongside the results of other’s trusted opinions.

    Here’s some interesting ‘for and against’ links for futher reading:

    Says they are linkbuilding penalties:

    2 people argue and then trial if there are penalties from linkbuilding:

    Also, SearchEngineLand reported 2011 that Google Webmaster Tools will tell you if they have detected ‘unnatural links pointing to your website’ and to apply for reconsideration once these have been removed, although this does not prove an algorythmic penalty has been applied due to back linking:

  2. Laura Phillips avatar
    Laura Phillips

    Great article Oli, and perfectly timed for me. I have to echo Alec’s sentiments on this, and Jim, the idea of being too busy building good links is the way forward! :)

    I’ve received an email this morning from a company I used to work with because they have suddenly disappeared from rankings for three of their top keywords. I have reviewed the site for them twice in the past, and warned them about their SEO company’s dodgy link building activity.

    It would seem the latest changes have caught up with this practice and they have received a rather large slap on the wrist. They’re a great company and deserve to do exceptionally well, but I’m not convinced about the agency they’re using. I believe this sudden drop is due to high levels of spammy and poor quality link building, the consequences of which may take a long time to recover from.

  3. Jim Seward avatar

    I obviously agree you can’t give the name of the client and nor would I ask you to, I have worked in a “very competitive financial niche” and if it’s the one I think you’re talking about, I agree, the competition was notoriously unscrupulous/black hat.

    The concept of googlebowling is one that has interested me for a number of years and every time I see someone who claims to have proof, said proof is very quickly dismissed as circumstantial and I’ve yet to see a concrete case study. I do wonder if between us, we could do a study, choose a site, attempt to googlebowl it.

    If we could conclusively prove googlebowling does (or doesn’t) exist, that would be awesome conference material. Happy to sacrifice my blog to try and prove/disprove it.

    The J C Penny example was very high profile and they had been called on it publicly before Google introduced the penalty and I believe that was done manually, as are most of the other high profile examples (BMW using doorway pages for example).

    I believe in most cases, the algo will simply discount links it believes have been bought rather than “applying a penalty” People who primarily have paid links as their link source, have their sites devalued in the search and of course tumble down the rankings and scream penalty, but in reality, they are just back to where they were before they started buying the links in the first place.

    1. Mike avatar

      Thanks for the offer Jim.

      We are always looking for new ways to test theories and appreciate the offer of using your own blog to do this testing. As we try to keep all testing secret until results are ready, and as we don’t want to damage your blog we do have to refuse your very generous offer.

      If we do run a test such as this in the future you can be sure it will be right here, on the Koozai blog. Thanks for contributing to such an interesting debate.

      1. Jim Seward avatar

        Don’t worry, I’m fairly confident you wouldn’t damage my blog ;-)

        and no worries, I always enjoy a good debate and I look forward to any results should you guys decide to run such a test :-)

  4. Alec avatar

    Thanks for the follow Jim, I have followed you back :)

    Unfortunately I can’t give the name of the client because they are under NDA, but nonetheless they have a well established and successful website with a very strong link profile built up over many years, literally 10,000’s of links from a wide range of sources…

    Then one day their rankings dropped for their primary keyword plunging them into the deep fathoms of unread search results. We investigated the link profile using “Majestic” and were able to determine that all of a sudden they had received a massive number of links from a few websites which were very dodgy looking and from every page.

    The links were hidden in the code, all the same and all of the websites were very similar but not updated for months, aside from these links. The problem I suppose with proving something like this is that you can never eliminate all variables, but given that the site hadn’t changed, and these links were the only thing that stood out as unusual we determined them to be the cause of the problem. Once removed, the sites rankings reappeared very quickly.

    This is in a very competitive financial niche and so the competition are notoriously unscrupulous.

    A final note from me, would be this, buying links is a punishable offence by Google and there are strong examples of this like J C Penny, who have been Google Slapped publicly for partaking in these activities… This, I believe, demonstrates that bad links can result in a penalty from Google.. Furthermore I believe that this demonstrates that if Google thinks that you have bought links or the links look suspicious that they will (and have) penalised websites.

    What say you sir?

  5. Jim Seward avatar

    Hey Alec, started following you on Twitter this morning :-)

    I think for those sort of things to affect a site, the site in question would need to have a particularly poor incoming link profile anyway and it’s links made up primarily of poor quality incoming links.

    I would be interested however in an actual study showing evidence of a quality site being essentially “googlebowled” even under test circumstances

  6. Alec avatar

    Hi Jim,

    Thanks for your comments, always nice to see some debate springing up.

    I would contest the point you made that no incoming links can harm your rankings… Working agency side we see a lot of dodgy links, hidden links or sitewide links which can negatively effect your rankings. In a couple of cases we have determined that this has been done by a competitor of a client for the purpose of damaging their rankings.

    I think this kind of thing only becomes a problem when your rankings are badly affected, at which point you may need to stop building links and look to remove the bad ones causing your website to drop from the SERPs.

  7. Oliver avatar

    Hi Jim,

    Thanks for your comment.

    In my experience I have seen great results from removing pointless incoming links (in particular irrelevant sitewide links and links on the same IP).

    If you continue to test the effectiveness of your link profile then weeding out links can certainly be a worthwhile practice.

    1. Jim Seward avatar

      Hey Oliver

      Afraid we’re going to have to agree to disagree on this one, I work on the principal that no incoming link can harm your rankings, if this would be the case, it would be very easy to affect your competitors rankings.

      If you were to contact webmasters and suggest changing the site wide link to an in context link, I can see it improving your rankings but asking to remove it completely I can see having either a detrimental or no effect at all.

      Personally I’m too busy building quality links to worry about removing the pointless ones anyway ;-)

  8. Jim Seward avatar

    Interesting post, some I agree with

    Varied anchor text
    avoiding link farms etc

    However I would say that with regards to Irrelevent links/links on same IP, sitewide links, I couldn’t disagree more.

    Don’t get me wrong, I think that they won’t be as valuable as proper in copy links from related sites on different IPs surrounded by semantically linked text, however these types of links are not useless and they certainly couldn’t harm your rankings by having them.

    If I could harm my rankings by making the links you outline above, I would simply start making them for my competitors. There is no such thing as a bad incoming link, there is however such a thing as a pointless incoming link or a link that won’t carry as much value based upon the things you’ve outlined.

Leave a Reply

Your email address will not be published. Required fields are marked *

Gary Hainsworth

Senior Organic Data Specialist

Gary is our technical SEO specialist and boasts more than 10 years’ experience in the industry. With in-depth knowledge on site migrations and all aspects of technical SEO, he’s a valuable asset to our team. Gary’s worked with the likes of the V&A, Warburtons, the NHS and the Lake District National Park. He has a passion for guitars too, be that playing them, modifying them or even building them. Gary has appeared in Startups Magazine, Portsmouth News and Southampton.gov.uk.

Gary Hainsworth Read more about Gary Hainsworth

Digital Ideas Monthly

Sign up now and get our free monthly email. It’s filled with our favourite pieces of the news from the industry, SEO, PPC, Social Media and more. And, don’t forget – it’s free, so why haven’t you signed up already?