Recovering from Penguin - Anchor Text Surgery

28 replies
  • SEO
  • |
So we all know Penguin targets over optimized anchor text, and a recent blog post put out by Market Samurai talks about the tools they offer to recover, and how their tests on dozens of niches and hundreds of keywords showed that they recommend now to diversify your anchor text thusly:

  • 40% Brand Related Anchors
    - Your domain
    - Your brand name
    - Your product name
    - Etc.
  • 40% Generic Anchors
    - click here
    - website
    - random words
    - check this out
    - etc
  • 20% Target Keyword Anchors
    - The keywords you want to rank for
I've been working on repairing one old site's link profile to meet Penguin's standards. I was able to target a few problem links which I pinpointed as a cause of significant SERP drops. I was able to have the webmasters at the linking site remove the links.


Here's My Problem:


In Google Webmaster Tools, the problem links still show up in the link profile, even though they were removed months ago, and I have confirmed this.


How can I get GWT to recognize that the links have been removed? I really appreciate anyone who's got some advice here.
#anchor #anchor text #link profile #overoptimization #penguin #recovering #surgery #text
  • Profile picture of the author Wayne Liew
    Currently, there is no way you can tell Google via Webmaster Tools that a link is fixed or removed. Google bot will have to take their own sweet time to recrawl the source of the link again to determine for themselves.

    But rest assured, it seems like Google is working on a feature in WMT for us to remove or notify the bots to ignore certain problematic links that they discover.

    Google To Build Disavow Link Feature In Webmaster Tools
    {{ DiscussionBoard.errors[6833297].message }}
    • Profile picture of the author kylemarvin
      Originally Posted by Wayne Liew View Post

      Currently, there is no way you can tell Google via Webmaster Tools that a link is fixed or removed. Google bot will have to take their own sweet time to recrawl the source of the link again to determine for themselves.

      But rest assured, it seems like Google is working on a feature in WMT for us to remove or notify the bots to ignore certain problematic links that they discover.

      Google To Build Disavow Link Feature In Webmaster Tools
      I've wondered why they wouldn't implement such a tool as soon as possible. This should be on the top of their list. Especially since they're all for 'human edited'. This would take lots of human interaction.
      Signature

      kylemarvin.com

      {{ DiscussionBoard.errors[6833601].message }}
  • Profile picture of the author bradstern
    I noticed this too. It might take some time to refresh after they recrawl. We don't know when but at least you are doing the job of cleaning it up. All we can just do now is wait.
    {{ DiscussionBoard.errors[6833698].message }}
  • Profile picture of the author MaryPabelate
    Banned
    According to my experience, you have to build links 50% for the main one keyword which is also in heading and 50% for other sub keywords including click here, website, view my sites, your domain name, brand name etc.

    It works.
    {{ DiscussionBoard.errors[6833779].message }}
  • Profile picture of the author BAC
    You need to remove few backlinks on over targeted anchors to be in quality websites list.

    VGM
    Signature

    Get Content Targeted Super Real Traffic to Boost your Adsense Revenue!

    {{ DiscussionBoard.errors[6833797].message }}
  • Profile picture of the author divinegrace
    You should also mix brand, target keywords, general words with one another. You might find GenerateAnchors.com: Generate Anchor Texts very useful for this.
    Signature
    Analyze Backlinks - Check backlinks of any website. Free yahoo site explorer alternative.
    {{ DiscussionBoard.errors[6836394].message }}
  • Profile picture of the author Mantasmo
    Originally Posted by kylemarvin View Post

    So we all know Penguin targets over optimized anchor text
    I have a suspicion that out of all the factors that Penguin looks at, anchor text isn't anywhere near as important as most people believe.

    In other words - it's a statistical distribution function that uses machine learning. Basically, your site is ****ed.
    {{ DiscussionBoard.errors[6836410].message }}
    • Profile picture of the author paulgl
      Originally Posted by bnetwork View Post

      I have a suspicion that out of all the factors that Penguin looks at, anchor text isn't anywhere near as important as most people believe.

      In other words - it's a statistical distribution function that uses machine learning. Basically, your site is ****ed.
      It has nothing to do with shinola. But it's a close rock. It's an easy rock.
      So, people keep looking under the same rock, saying, "Okay, I'll still spam
      and get low level links, I'll just vary the anchor text. That'll show
      that mean ol' google that I can still go in the back door!"

      Hardly. Anchor text is a moot point if your links are not on the same,
      tired, worn out, spammed to death, de-valued places.

      I have no idea why this is such a popular item. Just like unique content.
      They both have nothing to do with what google says when taken out
      of context.

      But, it beats patience, hard work, and actually doing what google has
      always wanted.

      Still trying to go through the back door is what is making google
      become militant, and enforce what is essentially the same webmaster
      guidelines that has always been around.

      Paul
      Signature

      If you were disappointed in your results today, lower your standards tomorrow.

      {{ DiscussionBoard.errors[6842253].message }}
  • Profile picture of the author kylemarvin
    You should also mix brand, target keywords, general words with one another. You might find GenerateAnchors.com: Generate Anchor Texts very useful for this.
    Thanks for the source. Interesting concept.

    I have a suspicion that out of all the factors that Penguin looks at, anchor text isn't anywhere near as important as most people believe.
    If the SEOs who have determined the anchor text issues were doing voo doo or just 'believing', you might have a point, but these have been determined via testing. In other words, quantifiable tests with various methods where the only differentiating factor was anchor diversity, in order to come up with a tested method of what works at that time. Google's algorithm is 100% mathematical, and certain bits and pieces can be 'cracked' via testing.

    Obviously there are many factors not just one, but there is no disputing the fact that sites with too many target keywords exact anchor text have been hit. Millions of them.
    Signature

    kylemarvin.com

    {{ DiscussionBoard.errors[6836484].message }}
    • Profile picture of the author Mantasmo
      Originally Posted by kylemarvin View Post

      If the SEOs who have determined the anchor text issues were doing voo doo or just 'believing', you might have a point, but these have been determined via testing. In other words, quantifiable tests with various methods where the only differentiating factor was anchor diversity, in order to come up with a tested method of what works at that time. Google's algorithm is 100% mathematical, and certain bits and pieces can be 'cracked' via testing.

      Obviously there are many factors not just one, but there is no disputing the fact that sites with too many target keywords exact anchor text have been hit. Millions of them.
      Bollocks. Correlation != causation. Your problems are link sources, on-page and probably some other stuff we don't know about. If anchor text was the culprit, simply diluting the %'s would fix it, but it doesn't. And thank you, I wrote a couple research papers on Machine Learning and advanced algorithms while at uni, lol. I don't claim to know 0.01% of what Google does, but scientific testing my ass.

      Link me to those scientific studies?
      {{ DiscussionBoard.errors[6836700].message }}
      • Profile picture of the author kylemarvin
        Originally Posted by bnetwork View Post

        Bollocks. Correlation != causation. Your problems are link sources, on-page and probably some other stuff we don't know about. If anchor text was the culprit, simply diluting the %'s would fix it, but it doesn't. And thank you, I wrote a couple research papers on Machine Learning and advanced algorithms while at uni, lol. I don't claim to know 0.01% of what Google does, but scientific testing my ass.

        Link me to those scientific studies?
        I never said anything about 'scientific studies'. Simple field testing my friend. I'm not going to try to dig up instances of testing, which if you do SEO I would hope you follow some of those, like SEOmoz, hubspot, noble samurai (*the guy who built market samurai), Michelle Macphearson, just to name a few. Obviously, we'll never crack the whole code but even Matt Cutts himself (The guy who lead Penguin), has told us about the penalties caused by too much exact keyword phrase anchors. Ever go right to the sources' mouth? If you don't read this, I suggest you carve out some time to do so. I'm not getting into an argument on WF, I'm just speaking matter of factly.

        Of course link sources and things like that matter too, there's no disputing that. There are thousands of such determining factors that affect SERPs. But if you dial in everything the same (set a control), and tweak one factor at a time, you can identifiably determine Google's response. Rand Fishkin (at SEOmoz) talks about their methodology all the time, and they've been doing it over a decade.
        Signature

        kylemarvin.com

        {{ DiscussionBoard.errors[6836809].message }}
  • Profile picture of the author 36burrows
    Do you have a link to that Market Samurai post?
    {{ DiscussionBoard.errors[6836648].message }}
    • Profile picture of the author kylemarvin
      Originally Posted by divinegrace View Post

      You should also mix brand, target keywords, general words with one another. You might find GenerateAnchors.com: Generate Anchor Texts very useful for this.
      Originally Posted by 36burrows View Post

      Do you have a link to that Market Samurai post?
      Sure, the video and links to their free tool (which analyzes anchor text diversity) is here.
      Signature

      kylemarvin.com

      {{ DiscussionBoard.errors[6836710].message }}
  • Profile picture of the author 36burrows
    Thanks, this is a great thread
    {{ DiscussionBoard.errors[6836745].message }}
  • Profile picture of the author Mantasmo
    ^^yeah, yeah. Look, I'm trying to help you out here - the site won't recover for a long time (if ever). I strongly suggest rebuilding on a different domain, you'd already be ranking by now. It is what it is.

    Penguin uses tons of factors that your quoted sources have never even heard of. You can start by reading through AND understanding Google patents (they are publicly available).

    Market Samurai as a source of legit info? Really?

    Matt Cutts never said this: "penalties caused by too much exact keyword phrase anchors" - people quote him out of context all the time.

    SEOMOZ never determined that anchor text was the cause of Penguin slaps.

    Best of luck!
    {{ DiscussionBoard.errors[6836879].message }}
    • Profile picture of the author 36burrows
      Originally Posted by bnetwork View Post

      ^^yeah, yeah. Look, I'm trying to help you out here - the site won't recover for a long time (if ever). I strongly suggest rebuilding on a different domain, you'd already be ranking by now. It is what it is.

      Penguin uses tons of factors that your quoted sources have never even heard of. You can start by reading through AND understanding Google patents (they are publicly available).

      Market Samurai as a source of legit info? Really?

      Matt Cutts never said this: "penalties caused by too much exact keyword phrase anchors" - people quote him out of context all the time.

      SEOMOZ never determined that anchor text was the cause of Penguin slaps.

      Best of luck!
      lol don't try and say you're "trying to help" while also talking in a negative demeanor. He's just trying to have a discussion and kindly sharing what he's finding. It may or may not all be correct, but we're all trying to learn more about this topic.

      Based on my OWN research and experience, linking the same anchor text over and over and over again is what I've found to be a part of these updates, definitely not all of it, but it makes sense to me.
      {{ DiscussionBoard.errors[6836929].message }}
      • Profile picture of the author Mantasmo
        Originally Posted by 36burrows View Post

        lol don't try and say you're "trying to help" while also talking in a negative demeanor. He's just trying to have a discussion and kindly sharing what he's finding. It may or may not all be correct, but we're all trying to learn more about this topic.

        Based on my OWN research and experience, linking the same anchor text over and over and over again is what I've found to be a part of these updates, definitely not all of it, but it makes sense to me.
        So you'd rather I throw a bunch of nonsense his way and tell him that everything's going to be alright as long as he follows some study that's based on nothing? I am trying to help. There's nothing worse than working on something for months on end without anything to show for it. The site won't recover.

        Read my first post in this thread - that is the truth about Penguin. I have sites that recovered briefly and tanked even harder. All of my 301'd sites that recovered got wiped out again. I don't know of a single successful recovery (even sites which were pulling in $50k/month - those guys would do anything to get their cash cows back).

        Here's the best result I got on a penalized site: page 2 recovery (from n/a) after getting about 300 strong PR3-PR6 links (blog posts, newspaper articles, blogrolls). It cost me way too much to get there and I regret not rebuilding it right away. Now I'm stuck contacting webmasters to switch the links over to my new site...
        {{ DiscussionBoard.errors[6837013].message }}
        • Profile picture of the author 36burrows
          Originally Posted by bnetwork View Post

          So you'd rather I throw a bunch of nonsense his way and tell him that everything's going to be alright as long as he follows some study that's based on nothing? I am trying to help. There's nothing worse than working on something for months on end without anything to show for it. The site won't recover.

          Read my first post in this thread - that is the truth about Penguin. I have sites that recovered briefly and tanked even harder. All of my 301'd sites that recovered got wiped out again. I don't know of a single successful recovery (even sites which were pulling in $50k/month - those guys would do anything to get their cash cows back).

          Here's the best result I got on a penalized site: page 2 recovery (from n/a) after getting about 300 strong PR3-PR6 links (blog posts, newspaper articles, blogrolls). It cost me way too much to get there and I regret not rebuilding it right away. Now I'm stuck contacting webmasters to switch the links over to my new site...
          No, I definitely don't want nonsense thrown anywhere, it's just your attitude that is discrediting you (in my eyes).

          I haven't gotten any of my sites to recover either, I do agree that starting over is the best option. But don't tell me that varying your anchor text on that new site isn't essential to not getting penalized again. I know there are many more factors than just that, but you even preach that on your blog.
          {{ DiscussionBoard.errors[6837080].message }}
          • Profile picture of the author Mantasmo
            Originally Posted by 36burrows View Post

            I know there are many more factors than just that, but you even preach that on your blog.
            I don't preach it. I suggest doing it just to be safe - why not? The main reasons for using varied anchor text are: rank for more long-tail keyword variations and protect yourself from any possible future penalties. That's it.

            Other factors are way more important.

            But this thread is about recovering from Penguin and anchor text is not the answer.
            {{ DiscussionBoard.errors[6837131].message }}
        • Profile picture of the author zecke
          Originally Posted by bnetwork View Post

          I don't know of a single successful recovery
          Seriously ? I got "couple" niche sites that tanked on 26.04.2012

          1. I changed the whole site structure to a silo structure and expanded the page amount from 4-6 pages per domain to 20-100 pages per domain. Thanks to this I think I put some kind of authority to those domains.

          2. Then I did 301`s - they all came back to rank as before 26.04

          3. The result is that about 70 % of my niche sites were recovered (so far). About 30 % of them got tanked again

          4. I also did some lazy tests - just put 301 to new domain without any changes - all domains tanked

          5. Oh and one more thing - to every niche site I had 4 high PR exact anchor sitewide links - I took them all before doing 301. I think sitewide today is not a good solution to non-authority sites

          I`m not saying that it`s the only way to recover, but for now it`s working for me. Instead of looking at the boring backlink thing I`m creating long tail pages to convert my niche sites into authority pages. So even when my 70% recovered domains will bounce back to hell I will have couple of big sites and just put them to new domains and put some quality links on them - they will be growing like my belly this month I need to practice some sport and stop changing and testing things after the penguin

          The penguin was a disaster for me, but now I think it turned out to be good, because now I finally learn real SEO not just backlinking sh*t
          {{ DiscussionBoard.errors[6838159].message }}
  • Profile picture of the author shaunD
    really good information.. thanks for sharing !
    {{ DiscussionBoard.errors[6838167].message }}
  • Profile picture of the author kylemarvin
    @bnetwork, most of what you're saying isn't based on 'anything', not what I've mentioned in the thread, I don't need your help if you're help is to tell me to dump my site, that's the easy thing to do. Actually the specific site I'm working on is an EDM that is 10 years old, and is still indexed, ranking for hundreds of keywords, but 3 or 4 major keywords lost SERPS. Not worth starting over in this case. It is a member-generated content site. Thousands of forum comments.

    Look, I've been doing internet marketing since 2003, almost 10 years, and have done over $4 million in revenues through my sites, I'm not a beginner, and I don't want to have a battle over who knows more. I know about the risks and volatility of this business... The point of this thread was to see what people have done to get GWMT to reindex its data faster. That's it.

    To anyone reading this thread down the road, don't think that sites can't recover from Penguin, they can, and even have. The update is still only a few months old, and as Matt Cutts himself says in this video:

    https://www.youtube.com/watch?v=ES01L4xjSXE

    Google has a sentence or term set on their penalties.. Chances are, the term is 6 months or more, which is why many affected sites are still not recovering. Also, the fact that, as @Wayne pointed out, they're coming out with: Google To Build Disavow Link Feature In Webmaster Tools shows that down the road there will be ways to recover. Why would they give such a tool if they planned on never letting sites recover from penalties?

    In short, if you're talking about a non-authority micro niche site, yeah it's probably worth it to start over, but in my particular case, it isn't.
    Signature

    kylemarvin.com

    {{ DiscussionBoard.errors[6838191].message }}
    • Profile picture of the author Mantasmo
      Originally Posted by kylemarvin View Post

      In short, if you're talking about a non-authority micro niche site, yeah it's probably worth it to start over, but in my particular case, it isn't.
      I really hope that you're right and good luck with the site! If you've done millions online, putting $20k or so into the site shouldn't be a big deal and it may just bring it back, more or less. That's the only thing that worked for me (not a full recovery though).

      A friend of mine with a 99% member-generated site got hit in April (not a huge site, doing roughly $6k/month through Adsense) and never recovered. Been ranking fine for nearly 6 years before this. He got greedy last year and built some questionable links - shame.

      @the guy with micro niche sites: expanding a 10-20 page site into a 100-200 page site using a marked domain... not sure about that. Just rebuild fresh or get a legit aged domain if you're going to put in all that work.

      I'm off for the day. Have a wonderful evening everyone!

      Hey just a thought - if you've removed all troublesome links, have you thought about applying for reconsideration? They do actually look over your link profile and stuff. Don't do it if you still have questionable links though.

      As for GWT not updating - the sites that I 301'd and then took off 301's (once penalties hit again) still show all the links in GWT (on new domains), but rankings are back as I built new (high quality) links to them. GWT seems to be lagging significantly behind actual SERP changes, so I'd be monitoring rankings/traffic and not GWT backlink data. I'll post here when my GWT data changes, thought it's been a while now and nothing.
      {{ DiscussionBoard.errors[6838331].message }}
  • Profile picture of the author unikbit
    I agree with bnetwork.

    I have 500 websites in travel industry, the ranking was done with link exchange, removed a part of the links but still no improvement.
    I have websites which stayed in first spot for 5 years and now are on the second page. Also even if a website remained on first page the traffic dropped to 30%.

    I agree to list here few websites and I will do whatever you suggest and to see if any improvements.
    {{ DiscussionBoard.errors[6841535].message }}

Trending Topics