As almost anyone reading this post already knows, April 24, 2012 marked a big day in the search industry. Once the initial Penguin update was rolled out (please believe me this is only the beginning and there is much more to come) the SEO industry, as we know it exploded in a flurry of fear and satisfaction.
For those of us that had sites get hit (I'll admit I had several sites dinged by this update) what started out as anger quickly turned to fear and curiosity. Many industry publications jumped the gun and, in my opinion, began publishing tips and processes on how to 'recover from Penguin,' when the truth is, as mentioned by Ian Howells in his recent SEO Podcast, it's really too soon to tell the full effects of these algorithm updates and anyone out there preaching is really just speculating. The best information I have seen thus far is from The HOTH, and that is DON'T PANIC, and BE PATIENT.
I did, however, have a real life experience where one of my sites, my own personal blog, got hit for what I am now almost sure was over optimization, and I was able to recover. What's really interesting to me is that my site, nickeubanks.com, got hit at all... let me explain. My personal site is low traffic, low importance. I do not build links to it, I do not monetize it, it really just exists to serve as my digital resume and a place for me to openly ramble or rant when I feel like it.
The Penguin Smack
Here is a screen shot of Google on Friday May 18, 2012. After a friend of mine reached out to me to ask if I had taken my site down (of course he didn't just go and check the domain :P) I asked him what he was searching for. He mentioned he had typed in some words in the title from what he could remember and my name - which should be more than sufficient to generate a SERP with my post(s). It did not. Instead this is what he was seeing:
With Inbound.org starting off the list, it was page after page of places that linked to my post - but not the post itself. My immediate reaction was fear that somehow my site was sandboxed. So to check I did a quick search for the full post title in quotes and there it was... what does this mean? That my site was penalized... but for what? As I mentioned before I don't do any active linking, advertising, and the site has slim to no traffic. My first thought was that I might have been a victim of Negative SEO. I logged into Webmaster Tools and pulled down my indexed Google back-link profile, which I have put into a public Google Doc here so you can see it. Upon review you'll see this is a pretty natural back-link profile, even with some links from some pretty authoritative websites... at this point I am scrambling for answers...
What The Hell is Going On!?
I was racking my brain to think of what it could possibly be that was causing my site to be buried in the SERP's, especially for posts that have a lot of natural links, social signals, and are full of unique, well written content (note: I didn't write most of the content in these posts).
I reached out to my buddy Mark Kennedy, as among the Philly SEO crowd he is certainly one of the most passionate SEO's I have ever met. He had the same line of thinking that I did and immediately hit up ahref's looking for spam-links or clues. Nothing. His next suggestion was to pour over any recent changes I made to the website. I reviewed some of the CSS changes and couldn't find any messy code or mistakes that may have warranted the site to be dinged (Did I mess up my headers? Did I botch a declarations statement?) Nothing.
The only thing I could think of was to really take a closer look at my links, so I started inspecting each of the sites that was linking to me. During this process I stumbled across my old blog from college, 23Run.com. Here are the Google indexed links from 23Run. As you can see there are 77 of them, which out of my total indexed link profile, is roughly 11%.
I went to 23Run.com to take a closer look at how my site was linked:
I Had to Change This
And there it was... right in line with the Pengiun post from Microsite Masters showing sample data from their analysis, I had over 10% of my links over-optimized for anchor text. So I made this quick change:
How Long Will I Have to Wait?
And then needed to gauge about how long it would take for Google to crawl my site and index these changes, so I took a quick peak at my average crawl rate in Webmaster Tools:
and seeing that my average crawl rate was 59 pages, but my low was 24, I decided to give it the weekend and check back on Monday May 21, 2012. When Monday's production activity calmed down, sometime in the early afternoon, I decided to run the query again and alas;
Resurrection!
It is still ranking underneath Inbound.org, which is a bit strange, but it's back!
Furthermore, the post is back to ranking for more broad terms, such as 'fresh insights nick eubanks' as you can see below:
Conclusion & Takeaways
Plain and simple, over-optimized anchor text can be dangerous. What was once the holy grail of SEO, getting links with your target keywords in the anchor text, is now something that requires careful planning and attention.
My advice is to develop your link profile to not look natural, but to be natural. If your anchor text is 'over optimized', you run the risk of being penalized, so make the effort and put in the time to naturalize your links. Try to replace anchor text links with naked URL's or at the very least more natural anchor text - try to think about these links in the same sense of someone who doesn't know you, finding your page or post and creating a link organically; most likely it won't be your target keywords but your name, page/post title, or a more generic link text such as read more, learn more, etc.
I hope my real-life example proves useful and helps, in any small way, to dispel some of the speculation out there. Thanks for reading.
UPDATE: Per an idea by Bill Sebald I changed it back to see if I can get myself re-bounced and really prove this theory out. So yes, I have changed it back (yesdterday afternoon) in hopes that Penguin will hit my site again and we have more concrete data on the threshold and means for activating Penguin.
lol actually a good idea....
Brilliant!
a masochist getting hit by penguin on purpose :)Thanks for taking the hit in order to inform us!
What better way to test right? Use riskless assets. Thanks Julien!
Thank-you again for taking one for the team.
I have a few sites in the same situation as you have described above - I'll now try and get the backlink profile looking more natural and drop you a note on whether it helps & how quickly.
My problem has been the anchor text is skewed because I have links coming back to a www.exact-match-domain.com - it is difficult to look natural when your natural links look spammy - if you know what I mean :-)
Thanks Robin.
I know what you mean, I also have a handful of EMD's where I was worried about the same thing, what I've done in an attempt to naturalize these links, for example if the domain is BlueWidgets.com, instead of building links as 'Blue Widget' I've switched them to the naked URL, so either 'BlueWidget.com' or 'www.bluewidget.com' as these are more natural.
Cool. I'll try mixing up the domain links like you suggest.
Please let us know as soon as you know something. This is one of the best empirical investigations I have come across.
I just started a different approach, making links nofollow instead of editing all the anchoe text (I know, I'm lazy but if it doesn't work it's easier to change back.).
Did you change the anchor text of that link back? When I go to 23run.com, I'm still seeing the anchor text for that link to your site as "Web strategist Philadelphia".
Also, it seems strange that Google wouldn't put some sort of minimum threshold on the number of links a domain has pointed at it before leveraging a penalty for over optimized anchor text. It seems crazy to me that one sidewide link can knock a domain out of the SERPs like that!
Hey John, I did. This test was great but is really only half of the data case - if I can get bounced off again for my brand term (my name) then it's proven 100%.
Please update this post as soon as you find out, that'd be very helpful to know. Thanks for putting in all the effort to share it with us. Good luck!
Well, based on this link provided by @joshbachynski https://www.google.com/search?q=%22Web+Strategist+%2B+Audience+Developer+%2B+SEO+Philadelphia+|+Nick+Eubanks%22&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US it seems the change may have already taken effect and re-bounced me... which I must say, is pretty freaking cool.
I now see the anchor text as 'NickEubanks.com' so testing is still on?
It's interesting that you lost traffic 3 days after the official roll-out date of Penguin. I've seen others recovering in a similar manner simply by switching off/amending exact match site-wide anchor text links. However, if that was Penguin it would mean that Penguin doesn't really run periodically as Danny Sullivan's post suggests. Which makes me think that such overoptimisation penalties aren't caused by the Penguin update.
Thanks Modesto. I believe this wasn't in fact Penguin, which based on my conversations with some much smarter SEO's than myself, but an over optimization penalty which may act more like an algorithmic filter, which simply had it's sensitivity index increased along with the Panda/Penguin updates end of April.
That's interesting - if the over-optimization thing is separate from Penguin, what is Penguin about?
Bad and Unnatural Links.
Totally agree with you, John. Afterall, while he does have a percentage of keyword anchors, it's not as if they're coming from a bunch of spammy pagess.
The content on the site is natural and well-written; yet, I see Google turning a blind eye to some big corporations with some of the worst backlink profiles I have ever seen. A current client of mine has over 900,000 links and over 50% of them are from free domains with gibberish content, or are from terrible free web directories. I think Google needs to pay more attention to where a link is coming from, the context of the link, and the signs that is is definitely spam, rather that penalising people for what could, possibly, maybe, perhaps be a bought link.
It can't
This is a great article. I think over optimization of anchor text is a huge thing that can go oveooked. I also found that there is quite the SEO community in Philadelphia, if you can give me more info on any Philly meetups in the area please let me know. I attended the Search Church event, and found it very enlightening.
Thanks
Hey Zachary - thank you for the kind words. The best and most frequent meet-ups in Philly are definately SEO Grail, events by SEER, and of course when the Moz team makes there way out here. Truly an amazing community of SEO people in Philly though; lots of talent and a general willingness to help out.
Love these story-driven case studies. Thank you, Nick for being real and transparent with the community.
Agreed - loving this content. Thanks Nick! Very interesting to see the double verified case. I need to start doing that more often with results I discover. Do you typically see the kind of "proof" you mentioned above when duplicating an SEO experiment to test results?
Honestly not typically, it's pretty hard to hold all variables constant. Even in this case the test is not completely controlled, as other linking/ranking factors surely played their role. With that said I just wanted to provide some real data, as there is so much speculation right now...
Thanks for the comment!
You can't get this sort of instant recovery with Penguin. You'd have to wait until they run the filter again.
What you're seeing probably confirms that the site wasn't hit by penguin in the first place.
Hey Kevin - Thanks for the comment. Yes I believe you are right, hence the title change.
Based on the adjustment and timetable this is much more likely the effects of an over-optimization penalty that I believe is run as a filter, however, I do think with the latest Panda and Penguin updates rolled out toward the end of April the sensitivity of the threshold was significantly increased.
Great case study.
at the moment I have a new site which I have only natural links for and I'm ranking for a large amount of the keywords I would normally target.
I think the days of targeted anchor text are over. I think if you just use natural links and have your keywords around the anchor text google will reward you for it. They have already mentions this with images etc and I think they are doing this with link to.
Also have social media is another massive part, even just using twitter and facebook as a rss.
again great post.
Hey Isaac, thanks for the comment.
I agree - I think a natural link profile in combination with best practice on-page optimization is going to go farther now than it has in the past. My belief based on this experiment is that there is an over-optimization filter, beyond the Panda and Penguin updates, that had it's threshold/sensitivity increased and is algorithmically driven, hence the changes to links, crawl time, and net effect to the SERPs happening within just a few days.
UPDATE 2: I believe I was mistaken on classifying this as Penguin originally, and also for the record I will be changing back the 'over optimized anchor text' to a more natural anchor to see if this causes my rankings to recover a second time. Thanks to everyone for all of the feedback!
Thumb up for admitting your mistake! I hate it when people talk rubbish on the internet, and think this attitude should always be rewarded because it means less misinformation gets spread
Edit: make mistakes, not "talk rubbish"!
thanks for pot, but i think just apart from anchor text links there are few other things that will make impact on neww updates:
1. Meta tags using too much of keywords and do not look natural.
2. Alt text used in images as exact keywords with repeatations.
3. Reaapeating trxt in footer links again.
4. Inter-linking of pages with same anchor text from differnt pages.
make different combinations of keywords and use those in natural way and make it easy for visitors not for engines.
Once you have thse changes done, your website will start again coming up to original position.
Thanks for good post.
Good post, and something that we have been discussing for a while. It's amazing to think that what used to be the 'holy grail' of a link would now potentially damage your site.
Definiately try and mix up anchor text, target keywords, URL's and company/brand names for sure.
If it looks natural, it probably is.
For those trying to draw conclusions from this post...I'd take it all with a grain of salt.
1) Traffic dropped on April 27th. That makes it probably a Panda hit, not Penguin.
(https://www.seroundtable.com/google-panda-36-15107.html if you're skeptical )
2) The recovery for this site started on May 20th. That doesn't line up with the Penguin refresh.
(https://www.seroundtable.com/google-penguin-11-15206.html)
3) If it was a Panda drop, it isn't related to backlinks at all.
So, the drop and recovery could have been just about anything. A semi-educated guess would be that it was Panda related, given that a Panda refresh happened on the 27th, when the traffic dropped. Panda is all on-site factors...doesn't factor links at all. As for the recovery that started on May 20th, Panda is rumoured to be more integrated now, such that a recovery can happen in-between refreshes.
Hey Argh, please see my responses above, by using the links below:
Please reply if there is anything new, I'd be happy to continue tohe discussion as it has helped me to gain more perspective and consider more signaling factors. Cheers!
Yes, I see your responses.
I also see a url that says "recovering-from-the-penguin", a penguin image, and an H2 header that says "The Penguin Smack".
It's not just me...blogs are still linking back to this post as an example of penguin recovery.
Interesting perspective.
Thanks for the post Nick. I have had one of my sites hit by this as well. I looked into the link profile and found 2 links I had aquired unfortunately they were sitewide links (blogroll) and they were on large sites, so they accounted for 20,000 links all in the same anchor text. I was able to contact the website owners and completely remove the links, as I don't think they'll help now. But it's been several weeks now and nothing has changed in the SERPs for me. And the links still show up in Google Webmaster Tools.
Any idea how I can get Google to see sooner? Or any idea what's holding back the changes from showing up?
I'd go for natural anyway. I mean, why someone should use "web strategist philadelphia" as anchor text to your site? I mean, I was the user, I would never have clicked there to be sure: what is that? an agency, a personal site? a company?
If I have to put your site on my blogroll, I will put Nick Eubanks and maybe after "web strategist in philadelphia", so to tell the user "you are going to a personal website of an amazing SEO called Nick based in Philly".
In these cases, don't think as a search engine. think for the user. and you're gonna be good.
Alessio - Thanks a lot for the comment. I agree with you completely.
This link was placed a while back and was never meant for users, but in hopes of creating keyword relevance for the destination URL (my blog)... which now more than ever is potentially dangerous.
I think Google is going to contiue to move closer and closer to a semantic web built for users. Thanks again!
Really great study - especially with the re-bounce. People need to take over-optimisation far more seriously than they are currently doing so - I'm still seeing a lot of agencies seemingly choosing to ignore their client's historic backlink portfolios.
It would be interesting to see what happens if you remove the link completely. Clearly it's passing you some value when it's 'naturalised' so I'd hope to see a position somewhere in the middle - would help to prove that the link was causing you a penalty rather than just being devalued.
Great post, Nick. Thanks for sharing your experiences.
A client of mine had a similar problem, with a few site-wide links that were heavy on the keyword/exact match anchor text. These were links he'd set up himself, without thinking to consult with us first. I politely tried to tell him that he needed to vary some of them if he could, as he risked over-optimisation on the anchor text front (and this was all pre-Penguin, by the way). The problem? He didn't believe me.
I can see why though. His thinking was: "Hang on, you're telling me that if I change the anchor text to something that isn't the keyword I want to rank for, I stand more chance of ranking for it?! Sorry, but I don't buy it." There is a weird reverse logic to it, but to me it was common sense: of course Google's going to think it's fishy to have too much keyword anchor text in a few select areas.
I didn't win the fight unfortuantely, but I still gave him all my thoughts, suggestions and recommendations anyway, in case he changed his mind. Fortunately he survived Penguin V1, probably because we'd achieved natural, varied anchor text elsewhere. But it was still a cause for concern.
What I'd find useful is people's experiences in educating clients in varying anchor text and/or reducing over-optimised anchor text. Obviously the majority of clients will trust what the agency says, but there'll always be those one or two who refuse and become concerned that varying the anchor text would make their rankings even worse. If anyone has any advice or resources to share then go for it. Thanks!
One recent speculation I've heard (haven't tested yet) that I believe may have some merit is the idea that Google is now attributing more value to links that send traffic, so instead of a link is a link is a link, and beyond link juice being a derivative of DA/PR, links that are used may now be worth more.
Hey Nick,
Please let us know if you do any research relative to this train of thought. We have recently some experiences with new clients which leads us to believe you may be on to something here.
#hattip on the article too, at least you can say you sparked healthy debate right or wrong.
I have always been very skeptical of this possibility, since Google has no way to get click-through metrics for the whole web. They could use Google Analytics, but they say they don't, and that would only work for some sites. They could use Google Toolbar, but they seem to be phasing that out. They could use Google Chrome, but I don't think Chrome sends that kind of data to Google, does it?
I would be more inclined to believe that they are using algorithms that place higher weight on links that are positioned places that users are more likely to see and click on.
Interesting enough one of the more realistic ideas I've heard of late is that Google is now attributing more value to links that actually sends traffic, which is interesting unto itself as it separates link value beyond the associated authority signals of the linking website, going deeper to look at 'is this link actually sending traffic and potentially, what is this traffic doing upon arrival'
While I share your skepticism, one way that Google could collect click-through metrics is through Google+ buttons since every page that has a G+ button sends a request to Google with the URL of the page being visited.
It's not perfect data since any intervening pages without G+ buttons won't be seen, but it definitely gives them another way to track visitors on other sites.
Penguin sucks. What if a site has a huge amount of spammy links from 3rd party sites that can't be removed? (e.g. https://webbactivemedia.net/marketing-blog/285-google-penguin-for-negative-seo.html) Site owners have to do a huge amount of link building to even things out? If this is the case, then Penguin is encouraging unnatural link building! It's madness, and Penguin would seem to be an invitation for negative SEO.
Good news is there's a lot of services out there now to help you identify and remove these links. I recently started using Link Detox and have been happy with the results thus far.
Here's something to think about. If my main keyword that I target is Blue Widgets, and I have a massive amount of sitewide links that all target Buy Cool Blue Widgets, how badly would Blue Widgets be affected by that. I got hammered by Penguin on the 24th for my main kw phrase. It was about 60% of my anchor profile. I've changed a ton of links that were using the main 2 keyword anchor, but haven't seen any thing good from it. I do have a ton of sitewide links over a few domains that are using longer tail versions, but still have the main 2 kw anchor phrase included in the 4 kw anchors. Could those still be hurting the anchor balance for my main kw phrase? Would changing those sitewides to mywebsite.com possibly be beneficial. Just curious if anyone has done this on a large scale, like changing 10K links or more.
Sounds like you need to do some link removal on top of link diversification, toxic links are still toxic...
I'm more than a little dismayed that nobody at SEOMoz noticed that the traffic drop off was on April 27th and put two and two together here.
A penguin-related loss would have shown a drop on the 24th or 25th, and perhaps the 26th.
A drop on the 27th is almost certainly the Panda (not Penguin) refresh.
Hey Argh - If you read through the comments, and take notice of the title change, I realized quickly that this was not the result of Penguin but of an over-optimization penalty that seems to have had it's sensitivity / intensity increased, I believe, in line with Panda (to your point).
People are still linking to this blog post with the impression that it's a Penguin recovery story.
Perhaps since:
- There's a nice Penguin Picture near the top.- The url has 'recovering-from-the-penguin' in it.- There's a great big H2 header that says "The Penguin Smack", despite the evidence showing that it wasn't a Penguin Smack at all.
So, yeah, I get that you made a few edits, but this blog post is still spreading misinformation since you don't really get the point that's it's not about Penguin unless you wade through all the comments.
Cheers for putting this one in public Nick!
It will be interesting to see if there are any temporal penalties for Penguin too, for sites that don't clean up but continue to do the right things.
How do you make other sites that are linking back with over optimized anhor text. Any tips?
Hi Guys,
I have had my own major headache with the recent updates, however my clients site wasn't hit on the penguin update on the 24th which hit 3% of sites. However on May 26th when Penguin update 1.1 came out. His rankings for for pretty low competiition local terms disappeared. We were ranking for many local areas for specific keyword terms.
However, and this is where fixing an issue is much about blind guess work, patience and common sense comes in.
- on the day before my client removed links on his site to areas where optimised pages were set up
- Google also crawled the site the same day
- I had optimised the site a little further the month before to chase longer tail keywords
We were hit back in Jan, however this was my clients fault. He had various domains (I didn't know about all pointing back to the same site no 301 redirect etcs) and we were dumped because of duplicate content issues. Turning off the domains fixed this after three months.
Things I have done so far:-
- Put the links back on that the client removed
- undone what I feel may be over optimization.
I don't feel its the links as why weren't we hit with the release of Penguin which hit 3% of sites, and why would we be hit by the Penguin update 1.1 which effected 0.01% of sites. How long do I give it to see if this has worked?. Looking at the above I feel also the over optimisation penalty is not the penguin update but a filter.
Kind Regards
Neil
I have been trying for the last month to either get links taken down or modified that are optimized anchor text links but I have not received one response from any of the webmasters. I know this is what is hurting my site and was affected on April 25th. My question is, should I just starting building more links that are not anchor text links? Since I can't get the others removed would adding more links that don't use acnhor text dilute all the ones that do?. would this be the best strategy for my site at this point? Any help or insite would be appreciated. My site is https://www.philipshidkits.com, if anyone has any other helpfull hints or ideas let me know as I probably won't be able to survive in the long term on the traffic I am now getting with this penguin or over optimization penalty
Glad to see that you were able to recover. As you stated in the beginning of your blog post, we must be patient and see what sites Google is penalizing and which ones they are not. The more we know, the more we can plan our SEO efforts moving forward.
We are finding that using images as part of our linking building has been a safer method. No alt tags either. Plus you get the benefit of brand awareness and a banner ad, of sorts.
Hey Shell - I have begun using this technique as well (sparsely) and it has been working nicely. Cheers!
You will not believe, it really is the approach that i had adopted. Used long tail natural anchors, URL as an anchor, directional anchors such as click here, visit here etc. And what i got, it worked for me too.
Somehow it is one of the ultimate recovery task we people can do, although removing bad neighbourhood links, spammy links from linkfarms etc. fixing over optimised anchors and looking for appropriate keyword density on the page, content revision etc. can recover the lost ranks to an extent.
Rest is based upon the Length of Spam you had spread.
I am glad this turned out fine for you but I am not sold on over optimization causing any issues. I have one example of overoptimization being fine for every example out there saying it hurt the site.
Has anyone seen the algorithm details? Did Google publish a document indicating the overoptimized sites are going to be dooooomed? I honestly searched for the mathematical algorithm behind the penguin update but I do not think it's available.
WikiMouse
Search Engine Noob
Google is never going to publish details on their algorithm, at best you can hope to infer some information from Matt Cutts videos, but honestly there is so much information out on specific cases where over-optimization caused the hit that I don't think it can be said that this is 'just speculation' at this point.
I really got smacked on the 25th so I'm assuming it has to do with 'penguin.' The problem is I don't have many backlinks, at least I've never solicited them, my content is good, not wide ranging but from me. I do have quite a few pages/posts which typically link back to several main pages.
Would that be considered over-optimized anchor text? Here is the url: www.johnconner.com. Thanks for any comments about the site.
John
zinvine, your issue is maybe the repeated header user of term "Medicare"...
Google doesn't care much about raw links count but links from different domains. So I don't think sitewide link on just 1 domain can do that. (Imagine how easy the negative seo will be if it could).
My guess is that your site was hit by mistake and Penguin 1.1 fixed it.
Nice idea, i also check it on my sites, if it work properly, i also implement that idea.
Thanks for sharing,
I haven't found many experiences like this out there... very helpful.
Regards.
HI, Two days ago, My client's Website is badly hit by google.I was creating links on Edu sites, Social Sites, Content submission in various forms like Presentation Sharing, Web 2.0, Video.I can't understand why my client's website is badly hit. Take a look at
https://www.nissan24auto.com. This is on Top 10 on many competitive keywords. Will you please help me, what I do now?
oohh ...can't believe I'm saying this but I love this post! I've just screwed up my site ranking by getting myself a whole load of backlinks with the same anchor text. ...thanks for the tips. If anyone can point me in the direction of more stuff it'd be great. It's girlie stuff but my aim is to get to number 1 for "handbags" in the UK ..and I won't stop until I've done it. I've managed 23 for UK Handbags but just lost it with all my backlinks. Any help or advice on how to recover or next actions would be greatfully received. :o))) Thanks guys! It's an amazon webstore - minervacollection.com
Jackie
I have a site just four months old and was fortunate to get back links from a government health body which is related to my sites content. However as the links come from a large organisation they are now included in all the websites that come under the main site resulting in approx 500 back links all stating the name of my site which has a keyword in the name. They are all real and natural links and am concerned they might do the same damage as you mentioned. The Links came shortly after site launched so I've no way to tell if affected and asbits so new there's very little traffic. Your post is causing me to raise an eyebrow though with my own site. Thanks for the post
Thanks for this post. I enjoyed reading it and I like your ways on writing this... sounded like I'm watching at you while you speak :D.. Kidding aside, I'll sure try this advice on some of my sites ;)
*Gratefully! :)
I know this post is old but I’m wondering if anyone has any thoughts on the situation with one of my sites. It was hit by the original Penguin last April from what I determined to be anchor over optimization. I had about 75-80% of my backlinks with my main 2 keyword phrase. Over the course of 6 months I was able to get that down to about 3%. Seemed like I was on the right track, however I’m still being penalized for that keyword phrase and a couple other versions, which were way over optimized at the original Penguin. I now have a nice mix of anchors, raw urls, mydomain.com, generic anchors etc. I still have some older low quality directory links in there and a couple sitewides that are actually natural links on related sites (not purchased). The sitewides are still using anchor text, but a long tail version of my main two keyword phrase I don’t know if I should have those sitewides removed or changed and/or try to clean up the directory links or what at this point. I would have thought I’d be back by now. I still rank for some medium and long tails, just not my couple main keywords that used to generate 80% of the traffic. I never received an unnatural links notice and it’s not manually penalized…I checked. If anyone has any suggestions I'd love to hear them. Thanks.
Some of the recent clients I picked up are still hurting from this update years later. They did not realize that their previous SEO company simply went out and blasted their website with the same anchor text over and over, a super unnatural flag. Even a disavow has not seemed to help them recovery much of their rankings.
While going through your detailed post i notice that google webmaster tools play a very important role in checking the progress of the website .Its very compulsory to check the progress of the website during the SEO so that everything remains under control and best optimized.
Id like to know how frequently google punishes over-optimizing the internal links on your own website. I just had a huge drop in traffic, and I have no idea why. No shady links or anything. My content is top notch, only thing I can imagine is the internal linking because I did kinda overdo it. But I changed that, and I still dont rank.
Awesome! Thanks for sharing your experience!
Awesome! Thanks for sharing your experience!
I have just analyzed onlineschools.org ..... they are still on first page against the keyword online schools and most of there links are site wide, with an anchor text "online schools" ...
I really find this story interesting, although I think you can't ever be 100% sure about the reasons why you are penalized.
Today I performed a search for "convert dvd to ipad". I have no idea what Google is doing, but it seems really strange to see a site like convertdvd2ipad.com in SERPs on 2 position. According to ahrefs.com, about 35% of the links pointing to this site are anchored with my search term. Not to mention the spammy content you can find on the page.
How come this site was not penalized for over optimization? How come it managed to pass the google filters?
If any of you has any idea on this, I would really like to hear some opinions. I thought Google is smart enough to smell the incredibly low quality of this site.
I think in this case is an exact match: "convert dvd to ipad" for google is similar to "convert dvd 2 ipad" son in theory is their brand name and more difficult to detect.
On-Page i don't see that much spam. It's true that is an ugly website but the content is unique at Copyscape's eyes (https://copyscape.com/?q=http%3A%2F%2Fwww.convertdvd2ipad.com%2F) and in theory you can find what you need. There is no reason to Panda to penalize this.
I think in the nearby future, Google will decrease the value of the exact match domains detecting easilly if this a brand name or a KW name. The problem comes when new companies are created sometimes they use Keywords has their name, i think that's why Google hasn't penalize it hard for now.
Barbie, thank you for your response.
I am not a SEO engineer, but I think in this case Google should know that "convert dvd to ipad" is not a brand name based on the history of searches that include this keyword. But maybe the algo doesn't aproach this issue in a serious manner. As you said, probably in the near future this will happen.
Spammy content can be unique. I find the text hard to read (too many repetitions of the keyword). I called it spammy because of this. Definitely not written for the reader, but the search engines - and this is NOT what Google wants, right?
I agree with you, in theory you can find what you need. And this applies to many other sites trying to rank for the same keyword. But it doesn't deserves the ranking, and I am 100% sure Google would share the same opinion.
Anyway, the main idea of my post was regarding the link over optimization. As I said, 35% of the links are anchored with the keyword... no filter for this?
Regards
It's Barbio not Barbie. :)
I totally agree with you, but i was only trying to do an analysis has a machine having a look from a link perspective. I know that is a bad site and shouldn't be number #2.
This are some factors that i think it helps this bad website to be #2.
1. Website PR1 and KW is medium competition (from adwords KW tool)
2 . Has links to it. Where are links coming from? Maybe some of this links are good? 1 or 2 can be sufficient to rank well, has this KW doesn't have that much competition.
3. Has no internal linking. That exact KW only appears twice. I understand that can be difficult to the algo to be detected. i know it has loads of (convert, and ipad).
Has a conclusion, the algo has to improve and i think it will. Google Plus will also help to get rid of this set and forget websites.
So Long Love to Google Plus!! :D
Cheers
Very interesting facts, thank you.
Looking forward to more stories, with or without a happy ending, always interesting to see other SEO´s in action, how they think an manage in the daily work.
I don't think this blog is correct
The Penguin works by looking at the now of anchor texts but is related to how many c blocks share that anchor
Or else we would all be nuking competitor sites with just one site wide ecommerce link
I also did the same as you removing links such as these but my site has not recovered
Ok, then why a site I am working on has anchor text of 40% exact match, and many pages with similar titles linking to that homepage with the same anchor text (internal linking spamming), and it is #1 and #2 on the SERPs? oh wait, it has high PA.
A site with higher trust and authority can get away with more spam than one with less trust and Authority. There are so many factors that is hard to make a conclusion without doing a black box testing (for those programmers out there)
Just read this post and ALL the comments, some really interesting points. Buried somewhere in the comments was a little snippet about anchor text around links or something like that. I've been thinking that maybe google are now looking at, say a paragraph and figuring out what it's about and any link in that paragraph is ranking for the content or gist of the paragraph and not necessarily the actual anchor text. Just a thought, but maybe worth thinking about in an overall link building strategy.
Great Post....Google Penguin had really a drastic effect on the SEO world. Realizing its diversity and some analysis of its after effects I also made a blog post on how Google has actually led a way for Negative SEO. Check out my article on the link for your understanding.
www.marketified.com/blog/search-engine-optimization-seo/has-google-penguin-update-incited-negative-seo/
I don't think the issue is that simple, nor that the anchor text was hurting the overall positioning of the domain. What i think is happening is that if there is an over-optimization of anchor text links probably all overoptimized links will be deducted from Linkjuice. This must have a threshold depending on he Domain Authorithy and also links authority. But each website is a different so it might work in your case, and not on any other website. But it's allways good to share our case studies, it allways help to open our minds.
Thanks
Thanks for the informatove post, but was just wondering, was it just one link from one website that affected you?
I am not sure if that is the main reason for the fall. Let me know in case I am missing out on anything.
- Sajeet
So far, purely working on my own client base, Google appears to be getting the over optimisation penalty spot on. The only sites I have seen rank + visits drop considerably are those who have used aggressive on page optimisation techniques.
For example too many keywords in URL + page names, too many sub pages with similar content, etc, whilst the better quality sites, have held position well, long may it continue :-)
Hey FIMS, That has also been my experience both on my sites and my friends anc colleagues, especially content that is similar enough that it could be seen as duplicate. I think in general a good rule of thumb is to think about each page on your site from the perspective of top-level purpose, if each individual page does not serve a unique purpose then they probably shouldnt be individual pages and are better situated to be combined to beef up the content, relevance, and experience of just one page. Thanks for the comment.
Great post.
Don't forget that domain diversity itself is a factor. You didn't just have 10% of your link profile with optimised anchor text: you had 10% of your link profile from a single domain with optimised anchor text.
How these two factors interplay is harder to establish, but I'd certainly say people seem to be forgetting the basics and worrying almost entirely about anchor text variety.
I mention this because I know of sites with HEAVILY optimised anchor text -- including site-wides -- but from a vast array of unique domains. They've not been touched.
Hey Martin - you are absolutely right, and this is actually a point I should have made more prominent in the post; the diversity factor, specifically inbound link diversity ratio, is a big contributing factor here in my opinion. Thanks!
Great article!
What about over-optimized anchor text with IN-SITE links? Would Google see this as negative also?
On my site I have numerous exact match anchor texts pointing to the homepage and other various pages.
Cheers!
Hey Arnaud - In my opinion on-site 'over optimized' anchor text can be just as dangerous, I'm not sure if the the ranking factor is going to be as high signal, but I also think this has the potential to be seen as very spammy by Google. Even in the sense that G doesn't want you building over optimized inbound links to your site, they probably dont want you trying to game rankings by creating a series of links that add no value to the user experience. Thanks for the comment.
Thanks for your answer Nick.
Internal linking might be why my sites have been hit by Penguin. I need to change a whole bunch of pages! When I used OSE and MajesticSEO to analyse my backlinks, I didn't see internal links included in the reports which is why I hadn't paid attention to them.
Cheers!
Arnaud - The most reliable way I've personally found for exploring internal links, content, and keywords that are internal to your site is to make use of the data reporting within Google's Webmaster Tools, what's more is this is sort of 'right from the horses mouth' in terms of how they are seeing your site and your internal link structure.
Thanks for the tips Nick! All the best!
From personal experience, over-optimised in-site links also pose a problem.
For one of my pages, around 1,400 of its total 1,450 IBLs were internal links with the same keyword. That page used to be #4 in Google for said keyword. Now it's not on the first 10 pages!
Interestingly, the same page still ranks for another high-volume keyword that I haven't optimised for. That'll learn me...
It looks to me like your site got exactly 0 organic visitors during the period it was "hit". Was your site completely banned from Google (i.e. not indexed). Even sites with a bad penalty usually get at least a few visitors. Just wondering...
Hey Adam - I found it wasnt completely banned, but at the same time I was buried (10+ pages) for almost every single search term beyond my name, and for some reason (and this could be a bit of an anomoly) it seems I didnt have anyone searching my name and coming to my site over that same time period.
Interesting. And no visitors from Bing, Yahoo, or any other search engine for the same time period?
Oh no I still had some visitors from other sources, the screencap is my organic traffic from Google.
I can see how this would work. Shouldn't the majority of your anchor text links be branded? I'd say it's okay to get a few / small percentage of exact match- maybe. But I think Google is better off looking at site authority rather than anchor text. If you get branded links from major authority sites you should be able to rank for a wide variety of keywords in your niche.
I have noticed this a few times since the update, and I really have tried to be more carefull and now i do a
40% exact
20% LSI
20% Variation
20% exact URL
and now not receiving any penalization, hope this helps folks :)
Thanks
Joe
Thanks Joe, I've seen a similar link profile cocktail that has remained relatively sustainable since the latest update. Thanks for the contribution.
Completely agree - I'm advising clients at the moment to forget going after anchor text and to concentrate on getting brand links back to sites which are far more natural.
Really interesting case study.
Thank you for sharing.
Cornel
Great post. I got hit pretty bad and trying to fix everything i am seeing. I might have been going after the wrong thing though. I been getting rid of links that i thought appeared to be bad.
Thanks for sharing. I took a look at OSE and looks like there may be some more "web strategist" anchors. Anyway, interesting case study.
.... yeah it says Web strategist Philadelphia for me too
Hi Nick,
Thanks for sharing your excellent approved experiment with us.
Really it will helpful to me.
Thanks,
From the looks of things your site wasn't hit by Penguin, just an over-optimisation penalty, but you're spot on regarding the dangers of exact match anchor text - great post and one that I imagine will be very useful to a lot of people dealing with the same issue.
Very Well Researched and Explained with us.
This information can be very helpful to recover from this highly tight update of Google,but litle bit chaotic.
Thanks for sharing your study Nick.
I am not so sure about this though. If this is true, adding overoptimised links to competition could get them off the SERPs , which would a very dangerous thing.
Hey Saijo - That is a good consideration to keep in mind, here are a few points to consider:
So I'm not sure how well this work in the case of a competitor, but that sounds like it would be a good test :)
Good point about how overoptimization could be used against us by competitors. I could easily create a bogus website and spam it with your link using the same anchor text and have you penalized by Google.
Google wouldn't be silly enough to allow that to happen, I hope.
If I put myself in Google's shoes, the ideal algorithm for high quality results would be as follow:
1. Is the website providing useful and unique information or services (bounce rates, duplicate content, established/registered business, social media sharing etc)
2. Is the website presenting what it wants to say in the best manner (keywords, navi, speed, codes etc)
3. The CTR of the inbound links, which is the hardest of all for you or competitors to manipulate. If your link is in a blog post with relevant content, naturally it will have a higher CTR, not as high in the footer, and lowest if it's on a useless profile page or unreadable spinned article.
For ages the fundamentals of SEO and speculations on Google's algorithm have been based heavily on the technical side of things: keywords, anchor text, links and whatnot. And if I were Google, all these manipulative methods would be history.
What makes sense the most to me now is that Google will no longer care as much about how many links you have or how repetitive your keywords are, but how useful they are to the viewers. In other words, CTR plays an important role here.
It is not that your overoptimized keywords are getting you penalized, it's just that they are not giving you as much credit as before that's all.
These are all based on my personal experience and what I imagine, so take it with a pinch of salt.
Absolutely. This is exactly why I think G is making attempts to attribute more value to links that send traffic, although I realize this implies that this would only work for links that are indexed and traffic behavior that can be tracked.
I dont think only that causes your site penalized by google penguin. The reason is that you got most of the back links from blogs. Google likes mixture of natural back links. Try getting that. Also avoid strategy of building a blog and sending backlinks to your main site..
I took a look at your site (www.nickeubanks.com) and there are tons of spammy footer links.
Take a look any you will see.
Hey Olaguma -
I wanted to get some feedback on your comment.
First, what do you find spammy about my pages, posts, categories, and single most recent tweet?
Second, If you are talking about the long list of links that appear underneath my post/page content, those are the sidebars, take a look at the site's code and you will see.
Thanks,Nick
My mistake, the links that are in <!-- #primary .sidebar -->... https://i46.tinypic.com/2say1e1.jpg
WHOA!!! What page are you seeing that on? I am not seeing that injection code on any of the pages?
You know what this is more likely; My site was the victim of an injection attack a little while back - are you using IE? Clear your cache and your cookies and Shift+F5 and that should dissapear - I just checked the codebase for the sidebar files and those links are no where to be found.
I see it in Firefox, Chrome and IE??? (deleted all my browsing history, refreshed multiple times). Its there for sure (at least for me)!
You can add me on skype (markostrbac) and we can try figure this out...
Crazy! I've had 3 other people in my office pull it up in different browsers and we cant get it to render!! What OS are you using Windows/Mac? I'll hit you up soon, thanks very much for the head up!
Anyone else seeing these?
Windows 7, I am seeing this in Slovenia (EU). Try asking someone to double check from EU.
I tryed visiting your site via a couple of free proxys and I see a clean (OK) site. Strange stuff (hack)... make sure to check out your website code via mulitple sources. Regards
Thanks man!
This is a good one and even I need to work on mnay things for my Penguin effected blogs. Will refer your ideas for sure.
I wounder what link text varriations I could use now?Click here, Look here, www.domain.coom, Check this out, Brand name, site owners name....And what about this: Shall I still put in exact match keyword (e.g.20%) or not at all?Shall I choose better (juicier) Sises to link to me with exact match?Is it unnatural if somebody links to me from his main page?
Thanks for the post by the way. It is reasuring that I do now the right thing.
My guess is that Google analyzes link graphs per niche/industry and looks at the statistics based on those link graphs. If your competitors have a backlink profile of 30% anchor text links, and your website has 90%, then its clear you've been artificially building links. The key's to create a nice natural balance!
Also keep in mind diversity when it comes to which unoptimized anchor texts you are using AND the URLs in which you are backlinking.
Hey Nick, This is a great example to use. I like that you changed the anchor text back to make sure it was the problem. Defintely keep us updated to what happens with it.
I think Penguin is going to drive the spammer SEOs nuts. Haha, I'm glad I'm not one of them.
Sorry if you already said this.
In the example, Google saw one link that looks spammy. You changed the anchor text of the link and it did the trick, right?
So it's just the penguin saying "this link is spam" and taking away the ranking that the gave to your blog.
Then, your Final Blog wasn't over optimized, nor penalized...
So, was the linking Blog penalized and then de-penalized?
Hey Daniel, I believe it was more so related to the low inbound link diversity ratio (number of links divided by number of unique linking domains) and the fact that one keyword phrase comprised over 10% of my link profile (which were all also from the same website).
So it was the destination site, not the linking blog, that was penalized and de-penalized.
I am thinking that perhaps anchor text within copy is still going to work well, as this is the aurthor explaining a term by linking to the information. Maybe google is going to penalise those over-optimized, site wide, anchor text that is in footers and side bars.
Although our company's site gets a lot of anchored links, pretty much all of our links, this way, we haven't been hit, yet.
And so, the quest for the truth continues...
Hey Creode - thanks for the comment. I completely agree, anchor links within content is a pretty safe bet, as these 'in content' links will still be viewed as 'editorial' given they arent coming from automated content blogs or link networks.
In this specific example it IS an over optimized, sitewide (sidebar) link that was used to test both the threshold of the LRD diversity and percentage of backlinks within the profile optimized for the same anchor text.
Great information, well presented!
Nice tips for recovering from penquin but google is getting ready with next version any suggestion regarding it
Hey Shaam - Actually I believe Penguin 1.1 roled out on 5/25, and to date (by far) the best informaion I have come across is the SEOmoz post from Ross Hudgens.
Thanks for sharing - it has brought on an interesting debate on exact match anchor text and over optimization.
With this specific example, I wouldn't go for either "Web Strategist Philadelphia" or "NickEubanks.com". The first option is obviously over optimized which isn't good, but the second option is not very user friendly either. Peiople who don't know Nick Eubanks will not have a clue as to what the site is about. So, with user experience in mind, I'd prefer something that tells users (as well as search engines) what to expect when clicking that link.
Thanks for the comment Thea. I think you bring up a good point that I didn't adddress in the post, which is the anchor text should also provide a driect benefit for the user, in this case to provide some clues around what the website it's linking to is about. In my personal experience using naked URL's is one of the more natural anchor text variations (which I admit is not 100% what I did as I left out the "https://www") that is still considered to be safe.
But with the user in mind I have changed the anchor text to [name] [keyword] in [location], I think this is much better. Do you have any thoughts?
One of our clients has been demoted to page two of the SERP's for a specific KW despite having a natural backlink profile. Further analyis showed 23% exact match anchor text was used to link back to the page, which to me isn't spammy! Based on this post however I may have to change my stance.
Very Happy to see a post on over optimization penalty, this post is very helpful me.i liked the analysis on this post, very well!!!
Certainly interesting article and learned quite a few things from the article.
Thanks for sharing great information.
Same old story. Is the anchor text the only problem? Don't think so, at least changing anchor texts didn't help me at all.
Hey Risky Jj - if you read through the comments the story opens up a lot more; it wasn't just the anchor text, it was the volume (over 10%), low diversity ratio from LRD's, and the fact this was a sitewide.
Fantastic article.
Hay Nick, Over Optimization Penalty V/S Nick and NIck is beat Penguin Update, Win Against the Penguin
Sorry this is likely not Penguin, and doesn't prove anything about links being tied to Penguin: 1) The dates are off. Penguin came out Apr 24-26. 2) You did not remove all the "web strategist philedelphia" links from 23run - changing a couple links is not going to make a difference. 3) You are not ranking for "web strategist philedelphia" - this is most likely a "paid link" penalty (if anything), given that it is a footer / side-nav link which could look like a sitewide link purchase 4) FYI the anchor text nickeubanks.com is no more natural than web strategist philedelphia given that YOU MADE THE LINK You may indeed have also been hit by penguin - you do not even rank for your index page title tag in quotes: https://www.google.com/search?q=%22Web+Strategist+%2B+Audience+Developer+%2B+SEO+Philadelphia+|+Nick+Eubanks%22&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a Top page that ranks for that is on page 3 or so and is one of your subpages: www.nickeubanks.com/seo/fresh-insights-from-12-of-phillys-top-seos/ This is a classic symptom of a spam flag, which penguin provides when it finds pages with spam content - your index page is a good example as it does not have much in the way of textual content and has many footer links.
Josh,
First of all thank you, in all seriousness, thank you.
Actually I did remove all of the 'web strategist philadelphia' links from 23Run.com, If you read my response comments above I put the link BACK on 23Run.com to test this theory, so your link and the fact that my rankings have been degradated again is FREAKING AWESOME!
There is no way this is a paid link penalty as the link has a class set on it, as afforded by wordpress.com's platform, to say 'this is one of my sites.'
One other key point to note, on my site, nickeubanks.com - those are in fact my sidebars, not my footer, so this point is moot. Just to reinforce, the placement of elements on a web page does not dictate what it is or isnt, this is done via code, i.e. sidebar vs. footer.
In all seriousness, I appreciate the feedback as no doubt these sentiments are shared by others. Thanks man and Cheers to the long weekend.
Can anyone give me any ideas on this site? https://www. san-antonio-texas-homes-for-sale. com/
I had alot of links that i have already removed but google hasnt noticed I removed them yet. I took a 61 page drop for my main term and for my second term i dropped 28 pages.
Try 5 less hyphens in your domain name, that might help...
little late for that but it hasnt hurt my other sites
Just taking a cursory look at your site, your page and H1 titles combined with your and URIs could be the problem.
Example:
Title tag: San Antonio Luxury Real Estate & Luxury Homes For Sale in San Antonio Tx
On-page H1 title: Luxury Home for Sale in the San Antonio area
URI: https://www.san-antonio-texas-homes-for-sale.com/san-antonio-luxury-real-estate-homes-for-sale.php
Over-optimized by current standards. You might try playing with one page and see what happens.
I am trying to change this page now. What would you change?
I made some changes to the home page and to this page you taked about above. Let see what happens