7 More Penguin Findings: An Update From the Over Optimization Front Lines

More Penguin 1.0 Findings

After Penguin first hit on April 24th, I started performing a lot of analysis on websites that were affected.  I ended up quickly publishing my findings in two blog posts here on the Internet Marketing Driver.  The first explained how exact match domains could be susceptible to penalty based on how those domains were being used.  A few days later, I wrote a second post that included initial findings based on analyzing a number of websites hit by Penguin.  In that post, I explained how the initial rollout of Penguin seemed extremely inbound link-heavy.  I simply wasn’t seeing other webspam tactics getting penalized.  Every website that I analyzed that had gotten nuked ended up having serious inbound link issues.

Based on writing those two initial posts, I’ve had numerous businesses reach out to me that have been hit by Penguin.  They range from businesses running one single website to owners of hundreds of websites.  It’s been absolutely fascinating to hear what’s happened to various websites (and networks), and then be able to analyze those sites.  In total, I’ve analyzed approximately 60-70 websites since Penguin hit.  As a result, I have a lot of data.  My goal with this post today is to share some of my findings, explain what I’m seeing, and shed some light on the situation.

Real People, Real Problems
The first thing I wanted to mention before getting into my Penguin findings has little to do with SEO.  Since April 24th, I’ve had the opportunity to speak with a lot of business owners that have gotten hit by Penguin.  I’ve been amazed at how open everyone has been with me.  Some people emailed me, while others simply called me directly.  In almost all cases, you can feel their despair through emails or in their voices.

I just want to emphasize that no matter what you think about webmasters using grey hat or black hat tactics, it’s important to know that there are still people on the other side of those websites.  Real people, that now have real problems.  Yes, many broke the rules.  I get it.  But it’s still hard to hear some of the stories…   Some won’t be able to pay their medical bills now, while others are going to find it hard to pay their mortgages.  I think that gets lost when speaking about Penguin, and it shouldn’t.  That’s why I titled this article, “An Update from the Front Lines”.  It’s like digital combat.  And I’ve spoken directly with the wounded.

With that out of the way, let’s dig into my findings:

1. High Threshold of Exact Match Anchor Text Got Hammered
As I mentioned in my last post, Penguin 1.0 was extremely inbound link-heavy.  I haven’t seen other webspam tactics get hit like spammy inbound links.  Based on my analysis, websites with a high percentage of exact match anchor text were hammered.  For example, a site with 80%+ of its inbound links using exact match anchor text got smoked.  Performing an inbound link analysis on many of the sites I reviewed revealed unnatural links.  That included links on low quality sites, article marketing sites, etc. I have yet to come across a website with a truly diversified link profile get hammered by Penguin.

High Percentage of Exact Match Anchor Text

2. Other Spam Tactics Not Hit Yet
As part of my analysis, I came across several sites that were keyword stuffing, had over-optimized title tags, using doorway pages, etc., but didn’t get hit.  As I said in my previous post, I believe future releases of Penguin could hit those tactics.  I don’t believe sites using those methods are safe.  It is probably just a matter of time.  Beware.

3. Very Low False Positive Rate
I have seen a very low false positive rate.  Actually, I haven’t come across one site that was a clear false positive.  Also, Danny Sullivan interviewed Matt Cutts about Penguin this week, and Matt explained that Google is happy with the results of Penguin.  He said the false positive rate is very low.  I have to agree with his assessment.  I’ve analyzed many websites and almost every one of them had serious inbound links issues.

By the way, it also sounds like Penguin will be rolled out periodically (like Panda).  If that’s the case, then you won’t notice any changes until the next version of Penguin rolls out (no matter what you change in the meantime).

4. Panda + Penguin = Confusion for Marketers
This one is really confusing webmasters.  Panda rolled out on 4/19, and then Google rolled out a Panda refresh on 4/27.  In between, Google rolled out Penguin.  As you can imagine, this is extremely confusing for marketers.  Many don’t know how to even determine what they were hit by.  And the last thing you want to do is to take action thinking you were hit by Penguin, when you were actually hit by Panda (or vice versa).  I’ve had several companies contact me saying they were hit by Penguin, when in reality, they were hit by Panda.  My advice is to make sure you know which update hit you, and then form a plan of attack.  If you are confused about this, contact a professional SEO.

Panda or Penguin Update

5. Private Networks
I’ve had several owners of private networks contact me about Penguin.  To clarify, I’m calling a network of websites owned and operated by one entity as a private network.  The sites all leverage each other for links, and as you can guess, exact match anchor text links are heavily used.  Many of you reading this post would probably assume that all of the sites that were part of private networks would have gotten nuked equally.  But that’s not the case.  I analyzed dozens websites that were part of private networks and what I found is going to surprise you.  Only some of the network sites were hit, while other remained untouched.  Then there were some that dropped a few spots in the SERPs, but only marginally.

This was fascinating for me to analyze.  I would drill into a site that got hammered, and find all of the unnatural links.  Then I would analyze a site within the same network, with the same types of links, and it was untouched.  Then another site part of the network, with the same links, only lost marginal rankings.  Why?  Did Google really miss those sites?  That’s hard to believe.  They all linked to one another using exact match anchor text.

I did start seeing a trend with certain categories of websites.  For example, categories A and B were getting hit, while category C was untouched (across websites).  This led to me to believe that Google might be targeting certain categories with Penguin 1.0.  I can’t say for sure if that’s the case, but I saw this several times during my analysis of multiple private networks.

Private Linke Network

6. Google Isn’t Done With Public Link Networks
If private networks consist of websites owned and operated by one company, then public networks are large networks of websites where many different companies participate.  It’s more of a typical link network where sites unrelated to one another all link to each other.  As you can imagine, Google hated this tactic, and hammered many with Penguin 1.0.  But I noticed something really strange.  One of the largest public link networks I came across was untouched.  Literally, not one site that I checked seemed to have gotten hit by Penguin.  Again, how in the world did Google miss a network this large?  It just didn’t make sense.

Since these sites were all unrelated, the category issue I mentioned earlier couldn’t be what saved them.  So why would Google let such a large network of websites survive?  Good question, and it seems like it’s just a matter of time before that network gets hit.  I think this just proves that Penguin wasn’t perfect.  So, although I didn’t see any true false positives during my analysis, I did see some false negatives.  I expect future releases of Penguin to address those issues.

7. Penalty vs. Devalued Links
I mentioned earlier that I saw some sites get nuked while others were only marginally hurt.  I mentioned this earlier when referring to private link networks.  But, I also saw some standalone websites with unnatural link profiles drop in rankings, but not get annihilated.  So, why would some sites get hammered while other simply drop slightly in rankings?

This probably means that Google simply devalued the links pointing to the sites that were marginally hit, versus applying a serious penalty to the site.  If a site fell off the Google map, it was penalized.  If it dropped a few spots in search, or jumped to page 2, it’s possible that some of the website’s inbound links were devalued.  It’s worth noting, and you should analyze your rankings to determine if you were hit hard by Penguin or if you are experiencing devalued links.

Back to the Front Lines. More Updates Soon
That’s what I have for now.  It’s definitely ugly out there.  My recommendation is to analyze your current situation the best you can.  Determine what you were hit by, Panda or Penguin, and then form a plan of attack.  During my analysis, I found there are some sites with a clear problem, while others involve a deeper analysis.  Good luck.




13 thoughts on “7 More Penguin Findings: An Update From the Over Optimization Front Lines”

  1. Thanks for another great post on your latest findings. Like many I’m having the issue of determining whether it was Panda or Penguin that hit my site. Looking at my traffic data there is a clear dip on the 27th April and then for the next few days I watched my top keywords fluctuate from 1st to 2nd-4th positions. They then settled on the 3rd May in the lower positions. Despite this I’m still wondering if the fluctuations were actually due to link devaluations as you’ve mentioned above?! When an Google update hits are the effects always immediate? i.e. if Google say they rolled out Penguin on the 24th do these changes happen within that day?

    • A clear dip on the 27th would definitely signal Panda. Also, I have a post going live on Search Engine Journal today that explains how to determine which algo update hit your site.  You should check it out. It should be live by this afternoon.

    • Julia, it’s important to know which update hit your site because each algo update targets different issues. Penguin 1.0 was very inbound link heavy, while Panda targets low quality content.  You want to know which one hit your site so you can address the right issues. I hope that helps.

  2. my site is also hit by penguin update but i’m not sure what caused it.
    my domain name is nepallica.com did google penalized it for exact match domain for domain nepalica.com (not owned by me) or metallica.com ?

  3. I have a question where me and my business partner are struggling with for weeks…

    I have a VPS server with one big main site (500+ articles) plus something like 10 exact match domains on some high converting keywords.

    I put one great article on those exact match domains and they rank very well (Google top 3). Below the article is a link to one of our optins or salespages.

    I know Google doesn’t really like the idea of buying a lot of exact match domains.

    So my question is: can the tactic of buying exact match domains hurt my big main site in anyway?

    And when I would buy more exact match domains, for example 50 or 100, will that hurt it?

    • Oscar, thanks for your comment.  It’s how you use the exact match domains that matters, and not that they are just EMD’s. I would only build additional domains when there is a good reason to.  For example, building a site that will contain high quality content that can stand on its own. I would not use exact match domains to house just a few pieces of content with the hope it will rank. 

      I would think long-term with your domains.  That should work out better for you down the line.  I hope that helps.

  4. Spectacular article with actual data instead of customary speculation.

    I know you say you’ve seen very few false positives. I may have one for you. Could you look at my site because it may be a Penguin victim of a whole different type? You will find this fascinating. My site is youthchg.com. My content has been widely stolen by spam sites that now outrank my site. I used to be #1 for words like “classroom behavior change posters.” Now in #1 is azyouthchgaz.az.com/ a spam site composed of the text version of my menu. Even more interesting, I filed a successful DMCA complaint against the same site two weeks ago when it was called http://youthchg.az.com/.

    There are literally hundreds of sites featuring my content, and they frequently outrank me where I used to rank fantastic for the words people normally use to find my site. From doing time in Google’s Webmaster Forum, I know my site is far from perfect, but if your premise is correct that bad inbound links are the main Penguin sin, every spam site that stole my content is inadvertantly loading me up with crappy backlinks. My content includes what were internal site links but because the spammers repost those links every time they steal my content, they automatically send me lousy inbound links. I even have links from hemmerroid sites! Some spam sites have posted a single article of mine dozens and dozens of times creating endless bad inbound links.

    Please tell me what you think. I was in the business of helping teachers save troubled children. Now, I am in the business of trying to save my business. Traffic and sales are down more than 85%. Please comment. Thanks for your rare insight, and valuable, useful articles.

  5. good post,i have a question that how google can judge what link is the paid links? because i have get a ‘no natural link’ message from google webmaster tool. so what do you think? the signal of “ADS text link” or business link etc? so …

    • Great question, and unnatural links don’t have to be paid. For example, there are many spammy directories that have been set up for the sole purpose of providing exact match anchor text links. Those are unnatural too, although not paid for.

      Google has several mechanisms for determining what’s paid and what’s not. And that includes the bigger picture which is what is “unnatural” – which again, doesn’t always have to be paid. Some mechanisms are algorithmic and others manual. And of course, there’s always user feedback via spam reporting which could trigger further investigation. I hope that helps.

Comments are closed.