SEO: In The News
SEO is a moving target at best! And, myths whirl though this industry like no other! We at SEONitro don't believe what we hear, but only what we test. This is where we report and discuss the happenings in our industry as well as our test results. Please feel free to praise, argue, or just rattle on about YOUR point of view. It is all good. :-)
Panda/Penguin Case Study Analysis
I have been asked a lot about what I think about the state of SEO, what others are saying about it, and if I think blog networks have a future.
And of course I do! Now more than ever. Google proved that by the massive de-indexing they did to the public ones. And, I am not alone on that thought.
First, because it is still really early since the recent Panda and Penguin updates, no one really knows "exactly" what to do (except for Google of course) as so many things have changed but there has been some great analytical data done on it in addition to the case studies I am going to share with you below. My personal favorite is MicroSiteMasters Is SEO Dead write-up on the Penguin, deriving at some interesting conclusions from looking at the data they had. (its a must read!)
Let's start here..
DID SOMETHING JUST CLICK IN YOUR HEAD? It should of, 'cause the JC Penny thing happened LAST YEAR!
I wrote that post Feb 24th, 2011!!
And of course that was the Panda 1.0 update and the Article Networks were hit HARD! The SEO community divided on the use of posting articles as a SEO benefit.
One year later, Article Network pages still don't appear on page 1 results as they once did, but links are proving to be as beneficial as they ever were in proper link density (more on link density later.)
So lesson #1. Deja vu!
One year later, the SEO community is taking another beating almost to the day! This time it is blog networks taking the brunt of the hit as everyone turn themselves in when Google sent out those dreaded "Un-natually Linking" emails into webmasters inboxes. We need to remember that THIS is the nature of SEO and not to panic, blame or react until the dust settles.
Here is the rundown of the last couple of months as "I" saw it.
May 10, Google's Matt Cutt's makes a revealing admission...
Um, "not about bad link networks?"
So even though the blog networks took the brunt of the blame this time around, all those notices were not the result of the blog networks? Wow!!
But, it actually makes sense.
Blog networks traditionally, and if done right, can't take on a massive amount of customers. I had only a little more than 400 in both SeoNitro and SEOLinkvine Elite so I knew "my" networks weren't responsible for 700,000 notices!! I don't know about BMR, HPRS, or ALNs membership stats but I would be surprised if all of us put together totaled more then 5k customers.
I had put that 2 + 2 together a long time ago, but it didn't seem to matter, the damage had been done. We (the collective SEO industry) told on ourselves and the unravelment of the incredible blog network ranking machines were thrust into the cross hairs of Google's de-indexing committee, and we did NOT win.
But here is the kicker.
Not only did they de-index the sites, they demanded that webmasters have their links taken off of those de-indexed sites before any reconsideration would be done.
Now, I scratch my head. Why would they demand links to be taken down if these links were on de-indexed sites that were dead, gone, forgotten, virtually rendered useless in the eyes of Google?
Except for the tattle tale element, there was nothing really compelling about it, I mean, surely Google is smart enough to know when a link is on a de-indexed site and rendered useless? Right? But the drill continued over and over, until we all blindly followed the piper and we took down links and then finally entire networks.
Now I would never want to open the can of worms or even suggest that de-indexed sites can still have some sort of relevance in Google. That would be ludicrous, wouldn't it? And I certainly would not want to get in ANY debate over it, I am still incredibly sore from the forum bashing's and hate mail I have received in the last couple of months over the entire de-indexing debacle the moment we launched SeoLinkMonster! Ughhh.
But.. I will say, I do have people in my line up that want those links.. de-indexed or not.
Strange right? I thought so until they have showed me examples of sites ranking for amazingly hard gambling keywords with the majority of links coming in from BLOGROLL LINKS on DE-INDEXED SITES!! Yeah, that's right, blogroll links!! Forget that the site is de-indexed, if that isn't enough, they were BLOGROLL LINKS!! (I haven't given blogroll links any credence in years, but I saw it for myself.)
And I checked those links and sure enough, most of them are on de-indexed sites. (I can't publish them here to protect the site owner so you will have to trust me on that one.)
Then I thought to myself, well it is true.. whenever we suffered partial de-indexing in the past it NEVER correlated with a drop in rankings. Hmmm
So what if it wasn't the de-indexing of the networks that made a lot of these sites drop? Hmmm
(I was among everyone that thought the ranking drops of blog network users were because of the loss of links derived from the mass de-indexing. I mean, if the site was de-indexed, the link was as good as gone, right?)
Well, just what if..
..what if it was actually a loss of links from the networks taking down links at customer requests in combination with the "exact match" keyword dial down?
(the "exact match" keyword dial down had been coming and getting stronger (or should I say weaker) for a while. I had done an entire write-up on it for my seo mastermind back in Nov, 2011 and had been advising customers to make sure they had a large percentage of URL (brand) links to balance their linking profile for years)
I then wondered if Google still crawled de-indexed sites?
Well one way to find out was to go look at the logs of the network sites that I had that had been de-indexed.
And to my surprise... they DO!
This made me start digging a into money sites that had recently dropped in rankings.
And what I found was that in the sites that "I" looked at, every single one of them had a higher dead link percentage.
(IMPORTANT: a dead link is NOT a link from a de-indexed site, it is a link that is no longer there or the site is not available)
Average for dead links is running around 20% and anything with 40% or higher was NOT on the first page for their respective keywords and the ones over 70% were not in the first 100 for their respective keywords.
Here are three case studies I did for this article. Each of them in a different market and each had used blog networks, some more then others, but they all had dipped their toes into it.
You can see the affected sites on the left have a WAY larger % of dead links (in blue) and the two that dropped into the hundreds on the top right have a huge amount of dead links.
IMPORTANT: I want to say this again. Dead links are NOT a link from a de-indexed site, it is a link that is no longer there or the site is not available. The tool I used to pull this information does not decipher if the link is coming from a de-indexed site or not just if the link was THERE or not.
So what does all this mean?
I don't really know, but if we have a site ranking in the gambling industry that left their links up from de-index sites and sites that have dropped because they took those de-indexed links down, that there might be something here, and Google may have played us big time.
So, I am testing it.
I have created a case study on one of my own sites that ranks respectively in it's market and have just started a linking campaign towards it using ONLY de-indexed sites.
I am linking using industry phrases that I am currently NOT ranking for (top 1000) and if I show up for those keywords, then we know that de-indexed sites do play a role in ranking efforts. If I go down in my established rankings, then we know that links from de-indexed sites can be used to negatively seo a site. If nothing happens, then we know de-indexed sites do nothing. (which I am still inclined to think, but I will finish this out. lol.)
Lesson #2: Don't Panic!
I learned a long time ago to take what comes out of the mouths of Google's distinguished engineers with a discerning ear. Things may not be as they seem, and then again, they may be, but we won't know for sure until we test it. SEO changes all the time and a quick reaction may do you more harm than good.
Now with that said, I don't want to leave here with you thinking that the dead links are the ONLY contributing factor to the above sites loss in the rankings. That was something that was just glaring out at me and had to look into further.
But, as I studied theses particular sites and the competition that is currently on page one of Google for their respective keywords, in most cases I found the following post panda/penguin off-page disqualifies.
So lets go into the other 4 points a little deeper.
Let's face it, social factors have finally become a SEO factor.
For example, here is an example for Case Study #1 with the keyword being "model ships."
We see that even though we are not currently ranking for this keyword (and we were previously #2) our site has the most MozTrust and Linking Root Domains score and has the second best page authority, but is low on social aspects and especially low on "Times Shared on FaceBook" in comparison with the sites now ranking on page 1.
On Case Study #2 with the keywords of "moving companies" and "mover reviews" for the example on the left, again our two sites that have tanked are very low on social signals.
While this site out does everyone in Google+ in the group, our Facebook shares and likes along with tweets are the lowest.
In Case Study #3 it is important to note this site only dropped to the bottom of page two so the affects of Panda/Penguin were not as severe, but as we all know, going from the top of page one to the bottom of page two is like going to page 1000 traffic wise, so this drop is incredibly severe to the pocket book and worth analyzing the partial drop.
And as you can see, we have the best page authority and out match in almost every area and even have done a nice job in the social arena, so it may be a hint to why this site only was set back to page two.
Link Type Diversity
So, lets stay with this site and Case Study #3 for a moment as we look into reason number three why sites tanked, "Link Type Diversification." We can see that this site did a pretty good job as compared to the other two case studies.
And in every case, sites remaining on page one have a great deal of link type diversity. And I look at a lot of these links, and believe me, they are coming in from "low value sites!"
I hear a lot of SEO guru's telling people to try and get rid of these low value incoming links, but I say, the internet is mostly made up with low value links, de-indexed included, so chasing people down to have them stop linking to you may not be the best use of your time. Just sayin'. LOL
Lesson #3: Low Value Sites Make Up the Majority of the Internet
The internet is made up with low value links just by being the nature of the internet, so get links from them all, i.e. from article and web directories, forums, sidebar and blogroll (didn't think I would ever say that!! lol) blog networks, press releases networks, social, web 2.0, images, profiles, comments, etc.
And yes I want the authority links too, but might be funny if I only have PR5s pointed at me!
And now we get so anchor text and look at Disqualifiers #4 and #5 as they go hand-in-hand.
This is probably the biggest post Panda/Penguin disqualifier as in most sites we will have researched did not diversify their link anchor density and were hit hard this go around with the "exact match" dial down.
Sticking to Case Study #3, let's look at the exact match anchor text ratio for the term "logo design" in comparison to page 1 sites.
We find that it is indeed the highest, but not totally out of the ball bark considering that LogoYes.com which is number 1 for the term pretty high as well compared to the others.
But the affected site has only 6.60% brand of incoming links (i.e., anything with their name or URL in it) while the other page 1 sites have them into the 26-35%. That says a lot.
But also noteworthy is that the affected site IS ranking on page 1 for the two terms that I show that have a lower exact match anchor density.
What is really great about the affected site is that it has a lot of onesy and twosy links that are all combined in the "other" category, so they are getting a lot of the "family" terms which we are finding important.
Another interesting thing about this case study is LogoMaker has an incredibly low exact match density for the term "logo design" except in the alt tag, which is the image descriptor. Hmm.
As we go up to Case Study #2 we find that our affected sites have VERY LITTLE brand incoming links. Ouch! This seems so drastically low that it is probably affecting a couple of the other low-density terms that I point out under site 2.
And look at MoverReviews.com with that HIGH exact match percentage for "Blue Horizon Shipping" where it ranks number 2. hmmm
(I would have to look at the other sites ranking for that term to see if that was high or not, but with the nice brand linking they have done, (and yes they rank number 1 for "mover reviews") they are looking in decent shape even though all the other metrics on our comparison charts were low.)
Lesson #4: Market/Keyword Sensitivity
This just shows me that EACH market/keyword is going to be different and I will have to research each one to understand what it will take to out rank my competition in that arena.
Case Study #2 was a little more difficult to do since I had two sites and different keywords, but hopefully it provided value as we look at the case studies as a whole.
Which leads us back to Case Study #1 for "model ships" and again our affected site has virtually no brand links coming in.
And again we see a site ranking (modelexpo-online.com) for a term with a trace amount of exact matches for this keyword that it is ranking on page 1 for. But this site has a ton of brand linking coming in and a lot of onesy and twosy links making it a very rounded site.
Lesson #5: Link Brand it!
It is natural for branded links (URLs) to come into your site, just look how many times I have done it in this article. And from what I just saw in these case studies, it is now become critically important.
So, besides the lessons that I pointed out along the way in this article, what are some of the solutions that can be taken for each of these disqualifiers.
And this isn't to say that this takes the place of great marketing and creating great content that people want to link to and like! LOL, I know you are already doing that. So if you are suffering from any of our disqualifiers, here is what you can do..
One More Thing..
..worth mentioning is On-Page factors are back and looking like Latent Semantic Indexing has finally come into play. I have not tested this, but have "heard" about it from a couple different sources (from people whom I consider know what they are talking about) and that this is helping sites that suffered and also ranking new sites at great speed.
The idea that is being presented is putting NEW content on your site using nothing but family keywords and synonyms that are used in your market and for the keyword you want to rank for and then linking to these new pages using brand or onesy, twosy types of links, which of course, are family anchors.
I also watch a great webinar that Matt Carter did with a guy who has been testing on-page factors related to Panda Recovery and seems to have narrowed it down to..
Which seems to go hand-in-hand with what I am discovering with my off-page tests. It is a very good webinar and I recommend you watch it for more on-page guidance. He does sell a product at the end but don't let that keep you from getting the goods he is delivering on the call.
I hope this study was helpful for you. If you liked it, please pass it around. I could use some coveted shares, likes and tweets! :-)
And, if you would like to spend a weekend with me going over this type of stuff in addition to learning how to build your own quality link network, then head on over here and reserve a seat. I also have packages where I will build your private link network for you so you can have FULL control over your linking.
"A private blog network is still something very powerful if used correctly. Building one does take some time but once you have a few high pr blogs that you can use for link juice (and you are the only one who’s using it) chances are you will not get de-indexed as the site in your network won’t look any different from any other website on the internet."
p.s. don't forget to comment below and let me know if you know of anything working or not working post Panda/Penguin as we are ALL interested. :-)
p.s.s. get on the update list to find out if Google did indeed make a fool of us when the test results come back from my de-indexing site valuation.