LRT Link Research Tools

comparethemarket.com – Really another Penguin 2.0 victim?

19 Comments

This analysis was created using our Superhero Plan and the new Competitive Link Detox tool extensively. The Superhero plan allows you to perform professional SEO and backlink analysis for your own or your competitor’s sites, but if you only need Competitive Link Detox (CDTOX), that’s also included in all new Link Detox Pro plans and up.

Don’t miss our next FREE LRT Associate Training. Get one of the last spots if you’re up for 8 hours of free professional training with Christoph C. Cemper.

Don’t forget to signup for our notification list to get informed on other Case Studies on Google Penguin updates.

Are you ready for the Google Penguin 3.0 Update?

comparethemarket.com - Really another Penguin 2.0 victim?

Learn from a SEO and Backlink Audit.

I am very proud to present our 12th deep-dive case study into a sharp drop in traffic and rankings after the Google Penguin 2.0 update comparethemarket.com

This analysis is a very interesting one, and despite having 60 pages again, our Certified LRT Xpert Harry Tschuggnall who runs the first Certified LRT Agency found things that left questions open for you, our community.

This case is particularly interesting; as it could be a new Google filter pattern and maybe we need to devote similar patterns an even longer analysis… did someone say 100+ pages?

But decide for yourself. We look very much forward to your feedback.

I greatly appreciate you sharing this research work by Harry Tschuggnall

- Enjoy & Learn!

Christoph C. Cemper


How we will structure our SEO & Backlink Audit

Introduction

In this case study we are going to look at one of the biggest UK online insurance comparison websites that lost a lot of its SEO visibility after the second Penguin update.

SEO Visibility before and after Penguin 2.0

Having a look at the Searchmetrics visibility chart for comparethemarket.com, we see that the website suffered a big drop after Penguin 2.0:

searchmetrics seo visibilty drop

The obvious question is: Why got comparethemarket.com hit that hard? What have been its mistakes?

Loser Keywords

In order to find out the reasons for the drop, we first need to take a look at comparethemarket.com’s loser keywords of the week after Penguin 2.0.

Checking this issue, I noticed that the website lost lots of rankings for keywords that are more or less off-topic ones e.g. “isuzu trooper” or “audi a3″. As comparethemarket.com is specialized in comparing insurances, of course the website doesn’t deserve top rankings for those keywords. We have seen many times that websites lost their rankings for off-topic keywords after Penguin 2.0, so Google is able to judge pages better now.

But of course, these off-topic keywords are not the ones that we are going to look at in detail, we are going to check the top money keywords that lost positions.

searchmetrics keywords rankings

And:

searchmetrics keyword rankings
searchmetrics keyword rankings
searchmetrics keyword rankings
searchmetrics keyword rankings

Loser Directories

Here we see the visibility of the Top10 directories before the update:

searchmetrics directories rankings before the update

After the update:

searchmetrics directories rankings after the update

What we can see here is that comparethemarket.com had a big visibility drop of its three strongest directories. Looking at other directories, we notice that comparethemarket.com did not only lose visibility after Penguin, but also won some: some directories and keywords gained rankings! To me it seems that not a single directory caused the massive drop in comparethemarket.com’s visibility. Hence, we need to look at the whole website. Let’s check comparethemarket.com’s anchor text ratios first.

Anchor Text Ratios

First I start the QBL (Quick Backlinks Tool) to get a first overview:

quick backlinks start

  1. Simply enter the domain you want to look at
  2. Select if you want to look at the whole domain or only at a special page
  3. I also recommend to set the sitewide links filter to 5 in order to ensure to get accurate data

After starting the report, the “Report Information” section shows us all the details about the report we’ve made:

quick backlinks report header

As we see, the QBL found 12.862 links and removed 748 duplicate links. As the tool is using 22 different backlink sources, it is able to filter out all duplicate links automatically.

Next you see a quick overview that tells you how much Power and Trust the domain has got and also how many referring root-domains are linking to the website:

quick backlinks report summary

AH! I just noticed that we’ve made a mistake. Can you tell which one? Alright, alright, I am going to betray it to you ;-)

If we go back to our QBL report and look at the “Report Information” section, we see that the tool removed dropped links:

quick backlinks remove dropped links

2.5 months after the update it is very likely that comparethemarket.com has removed some links already. Hence, the data of our report are not significant. Let’s re-run this report to get accurate information. We can simply do this by clicking the “Re-Run” button on top of the “Report Information” section:

quick backlinks rerun report

So the settings are exactly the same. The only thing we have to do is to uncheck the “Remove Dropped Links” box:

quick backlinks remove dropped links

Now we can start the report again. These are our new results:

quick backlinks report header

As the first report removed the dropped links, only 12.862 links have been found. The second report found 14.255 links. This can just mean one thing: comparethemarket.com has removed 1.393 links already! We are going to check this presumption later on.

Now we go to the “anchor text” tab and sort all the anchor texts “by count”:

quick backlinks anchor text breakdown

Here we see an anchor text cloud where you can easily see what kind of anchor texts have been used the most. Further, we get a nice chart showing us the breakdown in detail. We can also download the links as “CSV” or “XLSX”.

Wow!

Almost 10% of comparethemarket.com’s links have the anchor text “compare life insurance”.

As far as I remember, this keyword hasn’t been listed in the “Loser” table we have looked at before. At this point, I don’t have an explanation for this issue, but we are going to look at this mystery later on.

Instead, the keyword “best saving rates” was one of the loser keywords. Looking at the anchor text ratio table we notice that “best saving rates” wasn’t used that many times. But don’t forget: the QBL report is made to give you a “quick” overview as it only shows you the 16.000 strongest links. To get a full report with 100% correct data, we need to start a BLP (Backlink Profiler) report:

backlink profiler start

  1. Simply enter the URL of the website you want to look at
  2. Decide if you want to analyze a “Domain” a “Page” a “Subfolder” or an own list of links. For our purpose, we need to tick “Domain”
  3. As we are no members of comparethemarket.com’s SEO team, we don’t have access to the website’s webmaster tools and also don’t have any own link lists to upload – this would be the recommended way for inhouse analysis, as there’s always some more links added then
  4. To get as many links as possible we set the Link Boost 5x

Now we need to tick the metrics we want to look at:

backlink profiler metrics

a) Link Check

What can these metrics tell us about? Well, we can use it
– to get additional information about comparethemarket.com’s anchor text ratios
– to see if comparethemarket.com has a lot of redirects
– to find out if comparethemarket.com has plenty of footer links

b) Basic SEO stats:

These metrics are useful to find out
– from what kind of websites comparethemarket.com’s links are coming from
– how many backlinks comparethemarekt.com has got in total

c) Trust & Power Details

The Trust & Power details help us to
– see if comparethemarket.com‘s links are mainly from weak, unnatural websites or from strong authentic ones
– see if the power and trust metrics are balanced

d) Domain Registration Details

For what can we use these metrics for?
– To see if a lot of domains are having the same IP (to identify possible link networks)
– To find same registration details and therefore to uncover possible link networks

e) Link Source Country & City

Here we get information about the country and city of origin of comparethemarket.com’s links

f) Authority Backlinks

With the help of these data we can check if comparethemarket.com is using sneaky link building tactics in order to get lots of high authority backlinks

g) Authority Metrics

Here we can check comparethemarket.com’s backlinks according to the MOZ authority metrics. Those tend to be washier than CEMPER Trust, but it’s still good to look at any picture from many different angles. After all that’s why we love LRT for their 100 or so metrics, don’t we?

h) Advanced Link Counts

We are going to use these metrics in order to see if comparethemarket.com has websites from same IP’s and/or ClassC’s.

After ticking the metrics we want to look at, we set the sitewide links filter up to 5 and “Run the Report”:

backlink profiler sitewide links-filter

Remember: the keyword “best saving rates” is one of the biggest loser after Penguin 2.0, it lost 80 positions:

searchmetrics keyword position history

Let’s check again if comparethemarket.com has got lots of anchor texts with the keyword “best saving rates”:

backlink profiler report filter by anchor text

I can’t find a single anchor text containing the keyword “best saving rates”. That’s strange! What I’ve noticed so far is that there are 1.901 links that can’t be found anymore. I can’t tell you for sure but to me it seems that comparethemarket.com also changed some other links.

So now let’s look at other keywords that lost their rankings.

The second loser keyword is “cashback credit cards”:

searchmetrics keyword position history

Mhm….here we see the same pattern as we’ve seen before: the keyword disappeared for 2 weeks, shortly after it came back. But in this case, the “cashback credit cards” retrieved its position back. To me it doesn’t look like a Penguin issue.

Let’s look at this in detail:

backlink profiler report filter by anchor text

What? Comparethemarket.com has only 2 links with the exact match anchor text “cashback credit cards”? Mhm…

Let’s try another loser keyword: “travellers insurance”

backlink profiler report filter by anchor text

Alright, no results for this keyword.

We don’t give up: keyword number 4 “cheap insurance quotes”:

backlink profiler report filter by anchor text

At least we found some matching links for this keyword: 36 pages are linking to comparethemarket.com and containing the keyword “cheap insurance quotes”. Alright, 36 pages out of 9.363 can’t influence the website’s ranking that heavily I suppose.

Let’s carry on! Keyword number 5: “halifax credit cards”

searchmetrics keyword position history

Also this keyword lost its rankings for about 2 weeks after Penguin. Now it slowly starts to recover. Again, I don’t see any sign that the website got hit by Penguin. But let’s check that issue in detail.

We open the BLP report we’ve made before and reset all filters:

backlink profiler reset all filters

Now we simply write the new keyword “halifax credit cards” into the “Anchor Text” search box:

backlink profiler report filter by anchor text

No results. Again, I think comparethemarket.com has removed some links already. I suggest, Google didn’t value the keyword as relevant because having a look at comparethemarket.com’s landing page for “halifax credit cards” (http://www.comparethemarket.com/credit-cards/providers/halifax/) I noticed that the page doesn’t provide any additional benefit:

landing page for halifax credit cards

Let’s carry on to the next keyword “pet insurance lifetime cover”:

searchmetrics keyword position history

Alright, again the same pattern. Very strange, right? Let’s look at one more keyword that dropped. If we won’t find anything suspicious, we need to continue our detective work by using different tools.

Last Keyword: “price comparison websites”

searchmetrics keyword position history

Even here we see the same pattern again.

What we’ve seen before is that the visibility of the most important keywords changed after Penguin 2.0.. To me the reasons for that are either because of some technical issues or maybe because of a website relaunch. Well, of course I can’t find out if this happened because of technical issues, but I can compare the website before and after May 22 to see if a website relaunch happened by checking the Wayback Machine Archive:

http://web.archive.org/web/20130513080913/http://www.comparethemarket.com/

wayback machine archive

The page looks the same before May 22 and after May 22 so no relaunch had caused the drops.

Let’s find it out if we can find some other strange pattern.

Links not Found

Looking again at the links status section in our BLP report, I noticed this:

backlink profiler link status breakdown

Wow! More than 20% of comparethemarket.com’s links can’t be found anymore! Let’s look at this in detail:

By clicking on the “LinkNotFound” tab we see all the links that can’t be found anymore. I further sorted them by Power*Trust to see which of them are the strongest:

backlink profiler links not found

Let’s check out some of these links. Maybe we are able to find out why comparethemarket.com removed them.

LinkNotFound Nr. 1

link not found

Here we have a Russian SEO article directory. Well, maybe not the best site to get a link from these days, so it wasn’t the worst idea to remove that link in my opinion ;-)

LinkNotFound Nr. 2

link not found

Maybe comparethemarket.com advertised on many websites using follow links (as we know that’s against Google’s guidelines) and now removed them in order to get back its rankings. As the box is empty – the link is not found – I can’t tell you for sure, but I suspect that’s the case here.

LinkNotFound Nr. 3

link not found

Here we have a Polish article directory. I found 50 more links from the same IP. All of them are article directories that look the same. All of these links are not found anymore.

Further, I found 35 pages from a different IP but from the same Polish network:

link not found

Al in all I’ve found more than 100 pages that can’t be found anymore. All of them look very spammy and have a .pl country level ending:

backlink profiler links not found

I also found some blogspot spam:

blogspot spam

This German website is saying that the links have been removed.

Link Networks

Look! Now I am showing you a very cool feature that I’ve just started to use by myself and I love it!

I am talking about “Filter Slices“. With it you are able to analyze a certain aspect of your reports or to analyze all your “bad links” in a separate report. With the filter slices you can save certain sections or pre-selected data areas of your reports in a separate (partial) report and analyze them individually. Each new report delivers 2 slices, which automatically selects your predefined “bad” and “good” links and saves them in a separate slice.

filter slices

Don’t forget to apply the right filter, name the slice – I named it “cheap insurance quotes” – and hit the “save slice” button. It is good to save the filter, it can be useful at some later point.

Now I must admit this feature has been in Link Research Tools for almost a year now, but is a great example how you don’t use all features and options of a software you use daily. Stepping back and looking at new features really can save your time big time.

I am going to show you in detail:

filter slices

In the section “Stored slices” we find our saved slice. By clicking on it, all the metrics are going to show us only the relevant data for this slice.

anchor text breakdown

For example: Here the report tells us that 33 keywords have the anchor text “compare insurance | cheap insurance quotes | compare the market” and 3 links have the anchor text “cheap insurance quotes”:

ip breakdown

Further, 66,7% are coming from only one IP.

backlink profiler details table

Let’s look at some of these pages in detail:

http://bestuk.co.uk/information/insurance.html

web directory

Here we see a link from an ordinary web directory, having the anchor text “insurance”. This link looks quite normal to me. But wait! Have a look at this:

web directory

Comparethemarket.com’s entry in this directory looks exactly the same. Also the banner looks quite familiar, do you agree? ;-)

Same here:

web directory

Although the text has some differences to the previous ones, the website has the same structure as the ones before as well as the same adds.

Let’s see if comparethemarket.com sets up an own link network or if the website used some other existing network to get links!

Toxic Links

Let’s start a DETOX report, maybe comparethemarket.com has lots of toxic links that caused the massive drop:

link detox start

  1. Enter the domain you want to analyze
  2. Select the theme of the domain. If you are not sure you can select “automatically detect”
  3. As we are not the owner of the site and we don’t have access to its Webmaster Tools, we don’t know if comparethemarket.com got an unnatural link warning. Hence, we select the option “Don’t know”. If you ARE the owner or have access you should really fill that out in most possible detail, as it helps the team of LRT tune their algorithms. Especially example links Google now gives for manual penalties are very helpful for Christoph and his team.
  4. We select classic mode in order to check comparethemarket.com’s existing backlinks
  5. As we don’t have any other backlinks, we don’t upload additional backlinks (If you do a detox for your own website I strongly recommend to upload the links from your Google Webmaster Tool or your own backlinks you might have).
  6. In this step you also have the possibility to upload your disavow file or some other backlinks that you want to be ignored.

link detox upload disavowed / ignored links

Finally run the report.

Here we see the progress of our report. The more backlinks a website has, the more time it takes to finish the report.

link detox report header

  1. 15% of the report are actually done (You can reload the page any time to get the current progressing status)
  2. Within the “Report Information” section we find useful data like the start date of the report, mode, domain, backlink sources, and so on. Further, we see a very interesting number of links that have been found! Look at this! Detox found 91.283 links! By clicking the number we get all details about the links that have been found: link detox found links
  3. Here we see a short overview of the CEMPER Power & Trust Metrics as well as of the website’s referring root-domains

Let’s grab a coffee and wait until the report is finished.

About 15 minutes later the report is processed already. Wow, that’s very fast for this high amount of links, don’t you think?

High Link Risk

Here are our results: comparethemarket.com has a high Link Detox Risk!

link detox average link detox risk

The tabs on top of the chart give us additional information:

a. Link Health Breakdown

link detox health breakdown

b. Link Detox Risk Breakdown

link detox risk breakdown

c. Detox Rules Breakdown

link detox rules breakdown

Below these charts we also find a clear summary:

link detox summary

6% of all links are toxic, 62% links are rated as suspicious and 32% are valuated as healthy. Below the summary we find a table with lots of columns and opportunities to filter our links:

link detox toxic links

Let’s look at some links in detail.

There are 2 types of toxic links: “TOX1“, which means that the Link is NOT indexed.

The reason for that can be because the link is coming from a very deep page that Google did not crawl till now or because the link is coming from a brand new website that isn’t indexed yet. But in most of the cases a link not being indexed means that the website got de-indexed in consequence of a penalty.

The second type of toxic links is “TOX2“. Links are classified as “TOX2″ if they contain malware or virus.

Alright, let’s apply the “TOX 1″ filter and take a deeper look at some of comparethemarket.com’s toxic links:

link detox toxic links

The first domain is an article directory. The link to comparethemarket.com is not indexed anymore:

article directory toxic link

Let’s double check if the site is really not indexed anymore:

not indexed website

Yes, the tool was right, nothing has been found!

Wildcard DNS Subdomain Spam – Negative SEO?

I also noticed something else: If we change to “subdomains” in our Link Detox report, we get hundreds of pages displaying the same content but having different URLs:

wildcard dns subdomain spam

wildcard dns subdomain spam

wildcard dns subdomain spam

This can be a track leading us to the causes of comparethemarket.com’s drop. But maybe this findings are just a result of a wildcard domain. A wildcard subdomain is a DNS setting that leads each request to the same subdomain. Let’s check that:

wildcard dns subdomain spam

Alright, it always shows the same content so yes, it is just a wildcard domain.

Hint: Do not use wildcard domains!! Why? Simply because wildcard domains are an open door for negative SEO attacks! How does this work? O.k. I will show you but only to help you to protect yourself from negative SEO attacks and not to help you to attack someone’s website!!!

Actually it’s very easy to attack a website that is using hundreds of different URLs for the same content: Someone could build links for this websites by using hundreds of different URLS like I showed in the sample on top. You can basically write anything in front of the URL. This practice would generate massive duplicate content. Duplicate content will for sure drop your rankings sooner or later (Latest when the next Panda update will be rolled out).

What can you do to protect yourself? First, do not use wildcard domains! Second, implement a canonical tag that protects you! OK enough from the dark site of the planet…

It’s hard to find out if the website used wildcard domains as part of a SEO tactic or if the page was just expediently linked to
wildcard dns subdomain spam

Let’s go back to our Link Detox report and apply the “From URL Filter”:

toxic links grouped by domain

I also pressed the “Grouped by Domain” button just right on top. The results: 74 links are coming from these domains. To find out if the links from all these subdomains have been generated by accident or if it is a matter of a sneaky SEO tactic, we need to check if some of the wildcard subdomains have got own backlinks. Well, you probably think that this can’t be the case because all of these domains are “TOX1″ which means they are not indexed. But keep in mind that they might have been indexed already.

To find out if these subdomains have any backlinks, simply hover your mouse above a certain domain and the sign “a +” is going to appear:

start quick backlinks for toxic link

Simply press “Start QBL Domain” in order to check if this subdomain has got own links:

a.) 33.londovor.com: 0 links

quick backlinks found links

b.) aa.londovor.com: 0 links

quick backlinks found links

c.) w.londovor.com: 0 links

quick backlinks found links

As the subdomains don’t have any own links, using a wildcard domain seems like a sneaky link building tactic to me! But there are other ways to get a page indexed, even without links – like pasting the URL in your Google+ status will have it indexed within minutes …

Links from Infected and Malware Sites

O.k. let’s go back to our Link Detox report and have a look at comparethemarket.com’s “TOX2″ links.

link detox toxic links

The first entry directs us to the following page:

toxic malware site

To me, this website doesn’t look toxic as it even has the “MacAfee SECURE” seal and it runs under https.

So let’s check why this website is classified as “Tox2″. Therefore we visit the website http://www.urlvoid.com/ and enter the URL we want to get additional information for. That’s what the website is telling us:

blacklisted website

By scrolling down we can see where the website is listed as detected:

blacklisted website

By clicking the “Info” button we can see exactly why the page got listed:

blacklisted website

Here we see that some people rated this website as “Phishing or other Spam”. That’s why it is listed.

Let’s click another link from the Link Detox link table:

toxic malware site

It’s all German but I think you know what it means ;-)
Alright, so we covered the part about toxic infected and virus sites.

Link Detox Screener – Link Review on Steroids

Now let’s use the new “Link Detox Screener” to quickly screen comparethemarket.com’s suspicious looking links. The easiest way to start is to go back to the summary chart and to click the “Show me the links” button next to the suspicious links.

link detox summary

Now we get comparethemarket.com’s suspicious looking links:

start link detox screener

Next we hit the “Link Detox Screener” button and get to a nice dashboard:

link detox screener

  1. The “Link Screener” dashboard gives us all details of a website: “From URL”, “Anchor Text”, “Detox” the rating, the link risk based on all the rules, “To URL”, “Link Status” and number of the site’s backlinks.
  2. The “next” button takes us to the next suspicious looking website, the “previous” button obviously takes you to the previous one.
  3. This chart shows us the link risk for the actual link.
  4. You can also add one or multiple tags to each link. You could add a tag like “check later” so you are able to filter all the links by using this tag.
  5. Here you have the possibility to disavow a link. The tool automatically puts the link into a disavow file that will be generated at the end. You can also disavow the whole domain by selecting the “Domain” tab.
  6. The rating button gives you the opportunity to give feedback to the tool developer. If you think the tool rated the link right, simply give your thumb up, if not, just give a thumb down. Hint: Voting can bring you some free extra credits!

Let’s click the “next” button and have a look at the next suspicious looking link:

link detox screener

For our analysis, we don’t need the Link Detox Screener as we don’t want to generate a disavow file for comparethemarket.com. Hence, we close the screener and go back to the Link Detox report table. Now we reset all filters and look if we can find some unnatural link patterns because also the Link Detox tool is great for this task.

First we add some more columns to the table. This makes it much easier for us to find unnatural patterns. I have already shown you how to find patterns by using the BLP and by applying the IP filter. Now I want to apply the “Registration Detail” filter. Maybe we can identify some more bad links ;-). I select “WhoisGuard, Inc.”. Although there is nothing wrong with “WhoisGuard, Inc.” as it protects domain owners from spam, it can still be a pattern.

link detox filter by registration detail

Let’s look at some of the “WhoisGuard, Inc.” websites:

First one:

webpage screenshot

Second one:

webpage screenshot

Third Page:

webpage screenshot

And so on …

Most of the websites are blog posts, having hard exact match anchor texts!

Maybe comparethemarket.com wants to hide its link profile from its competitors? Well, let’s think about it: How can a website hide its real link profile?

  1. Build lots of spammy links and disavow them straight away. I don’t suggest to do so because always remember that the disavow file is just a recommendation, it doesn’t necessary mean that Google follows it.
  2. A more clever way is to build links for a different domain and then redirect the domain via 301 to your main domain. I have to say that this practice isn’t as easy anymore as it was before because Google changed its position regarding redirects but it still works.

We can’t answer this question as we don’t have access to comparethemarket.com’s disavow file. But to be honest, I don’t think that anyone is using these tactics.

What we can do is to make a CDTOX (the new Competitive Link Detox tool) in order to find out if comparethemarket.com’s link risk is higher than the link risk of its competitors.

Alright, let’s get started:

competitive link detox start

  1. We enter the URL we want to analyze
  2. Select the right theme
  3. We tick “Find Competing Pages”. Now we enter the keyword “cheap insurance quotes”, pick the right search engine, language and hit “Find Pages”.
  4. Now the tool automatically shows us comparethemarket.com’s competitors for the keyword “cheap insurance quotes”:

cdtox competitors' domains

Finally, we hit the “Run Report” button.

By using the CDTOX we can find out if a high link risk is common in this industry.

cdtox average link detox risk comparison

As you can see comparethemarket.com has a little higher risk then their competitors but they all have a moderate link risk scores. This is the typical use case of “know your battlefield”. If you think your moderate link risk is bad, then first look at your competitors, too – theirs might be worse and Google has to take what they get in every industry, every language and every country as Christoph always says.

So we have to check this off and try to find the secret of the keyword drops somewhere else ;-)

Redirects

Now we’d like to find out more about comparethemarket.com’s redirects. If there are many other domains forwarded using a 301 redirect to comparethemarket.com, maybe the website redirected expired domains that have lots of backlinks. This could be comparethemarket.com’s link building tactic: to hide some parts of its link profile by using 301 redirects and to get link power from the redirected.

To find out, we go back to our BLP report and select the “Link Status” Tab:

link status breakdown

60 links (0,6%) are redirects. O.k. that sounds a lot but let’s stay calm ;-)

What we are going to now is to try to find out what kind of redirects comparethemarket.com has – either 301 ones that are transmitting link juice or 302 ones that do not pass any link juice. Further we want to check how many and what kind of backlinks these redirected domains have. By clicking the “REDIRECT” button, the tool filters all the 60 redirects automatically:

blp redirect juice filter

Alright, as we just want to see the redirects that pass link juice to comparethemarket.com, we simply apply the link juice filter:

blp redirect juice filter

As you can see there are 15 redirects that pass link juice. Let’s have a deeper look on them! We are just going to look at the root domains:

a) http://compare-themarket.com/
We want to see if this domain has got lots of backlinks. Hence, we open the “URL Actions” and start a “Start QBL Domain” report:

start qbl domain from details

I am getting a bit excited now! I wonder how many links will come up!

quick backlinks found links

Mhm… only 2 links from 2 root domains. To make sure this page does not have redirects as well we check their links status too:

qbl link status breakdown

Nothing!

b) http://comparethe-market.net/

quick backlinks found links

This domain has 11 links and also no redirects. That’s not enough to influence the link profile of the main domain comparethemarket.com.

c) http://comparingthemarket.co.uk/

quick backlinks found links

14 links. We are looking for a redirected domain with 100s or 1000s of backlinks! 14 links can’t influence comparethemarket.com’s link profile that much.

d) http://www.comparethemarketholidays.org/

quick backlinks anchor text cloud

64! We are getting better ;-) 64 links could have impact on the main domains link profile but only if almost all of the 64 links have hard exact match anchor texts. Let’s look at the breakdown:

anchor text ratio

As we can see, there are money keywords but the majority of the keywords are brand ones. This cant’ have a great impact in my opinion.

Ready for one more? (It’s the last one, I promise you :-))

e) http://www.comparethemarketholidays.org/

quick backlinks link status breakdown

63 more links. Let’s look at the anchor text breakdown:

anchor text breakdown

Again, there are some exact match anchor texts but most of the anchor texts are brand or compound ones I would say. Classifying keywords for exact match domains can be tough, as it is hard to say if the keyword is a brand one or a money one.

All in all we found 153 links that are redirected to comparethemarket.com. In the case of comparethemarket.com this amount of links is very small compared to the total amount of backlinks of more than 90.000. But for websites with far less backlinks this amount of redirected links could have a big impact on the main domains link profile.

So if you want to check your competitors link profiles watch out for 301 redirects! Further, also keep an eye on your own 301 redirects! Check all redirects to your website with which you are not familiar. This redirects can be a consequence of a negative SEO attack and could have a high amount of links with anchor texts you don’t want to be associate with!

Link profile

Alright, until now we couldn’t find a lot bad signals. Let’s compare comparethemarket.com’s link profile with the ones of its competitors. Maybe an unnatural link profile is the key answer to our question why some keywords dropped rankings and why the page lost visibility. Let’s start the CLA (Competitive Landscape Analyzer) Tool:

1. Keyword “cheap insurance quote”

competitive landscape analyzer start

a) This time we use the “Quick Analysis” mode
b) We use the “Find Competing Pages” function again

competitive landscape analyzer start

c) The tool shows us comparethemarket.com’s competitors
d) As we want to analyze the domain backlinks we pick the first options
e) To get as many links as possible we set the Link Boost 5x.

We also set the sitewide links filter up to 5 and hit the “Run Report” button:

a.) Power, Trust & Power*Trust

competitive landscape analyzer summary

Here we see that most of comparethemarket.com’s competitors have more power, trust and links. Let’s look at the Power*Trust distribution:

CEMPER Power*Trust™ metric comparison

78 % of comparethemarket.com’s links have Power*Trust 0 while the total average is 53%. That’s far too much and means that the website has lots of very low quality links! After classifying all keywords we get the results for the keyword distribution:

b.) Keyword Ratio

metric comparison by keyword

Here we see that comparethemarket.com has more money keywords than the total average. This is of course a very bad signal.

c.) Link Status

link status metric comparison

Also comparethemarket.com’s link status distribution seems quite unnatural. Having lots of nofollow links can be a signal for having lots of spammy blog comment links as they are usually marked as nofollow. I also checked the other metrics. Comparethemarket.com’s link profile fits quite well into the link profile of its competitors. So I couldn’t find anything else suspicious.

This was just the competitive situation for one keyword. To check the link profiles of other competitors for other keywords that lost rankings we need to run a CLA for every keyword.

2. Keyword: “price comparison websites”

a.) Power*Trust:

CEMPER Power*Trust™ metric comparison

Also for this keyword comparethemarket.com has many more weak links than its competitors.

b.) Keyword Ratio

metric comparison by keyword

Wow! This time comparethemarket.com’s anchor text ratio is far higher than the ratio of its competitors! Almost 3 times higher! All other metrics are o.k.

I also checked 4 more loser keywords. Also in these cases it turned out that comparethemarket.com always has many more weak links and a much higher money keyword ratio.

Link Velocity Trends

Another important part of a natural link profile is the websites link growth. If it is natural or not can only be answered by comparing it with the website’s competitors.

Let’s start a CLV (Competitive Link Velocity) to check this issue. To get a full report, we simply need to enter the domain we’d like to analyze as well as its competitor websites. The tool shows us a clear heat map:

a) Link Velocity for the last 30 days / keyword “cheap insurance quotes”

To get accurate numbers without sitewide links, we need to select the “Domain Popularity” tab. The like velocity looks quite balanced I would say.

competitive link velocity domain popularity

Now let’s check if the website maybe deleted a massive number of links in order to get its rankings back. If that’s the case, we would see a high number of deleted links in the end of May / in the beginning of June. Let’s check the 24 months chart:

competitive link velocity deleted links

Comparethemarket.com removed lots of links in November/December 2011, but not in May / June or July. On the contrary. In the chart below we can see that the website has been building lots of new links:

competitive link velocity domain popularity

Conclusion

The case of comparethemarket.com’s visibility drop and visibility return remains a bit of a mystery for me until now. It is really hard to put the spot on why comparethemarket.com lost part of its visibility after the roll out of Penguin 2.0.

Further, there are very strange pattern for the loser keywords. All of them dropped out of the Top100 rankings. Some of them got back on track after 2 weeks, some didn’t recover yet.

So this is not a normal pattern for a website that just got hit by Penguin! We have to keep in mind: Just because the drop happened around the time when Penguin 2.0 has been rolled out doesn’t necessary mean that the drop can’t have any other reasons.

In my opinion one reason for the fact that the website lost its rankings and gained them back so quickly, could have been some technical issues. Maybe the webmaster had to set the pages to “noindex” or “nofollow” or maybe he implemented the canonical tag wrong. I have never seen a pattern like this.

Maybe anybody has further suggestion on how to analyze this case? I would be happy to read your proposals in the comments box below and potentially update or follow up with the findings.

Another possible reason could be simply an anchor text filter. As we all know, building too many links with exact match anchor texts can have negative effects on a website’s or their pages ranking. That has been the case since 2008 or earlier.

As we have seen, comparethemarket.com removed many links. Maybe lots of them have been links with exact match anchor texts. This would be worth diving deeper in historical archives, but that would have added another 60 pages to the 60 for this case-study I think.

Maybe removing those exact anchor links was the key to success and the reason why comparethemarket.com got its rankings back so quick. As we could see in the link velocity trend chart as well as in den BLP table, comparethemarket.com did remove lots of links – maybe also all links with the anchor text “best saving rates” – but not after Penguin 2.0. So maybe comparethemarket.com gained it rankings back because it removed lots of unnatural exact match anchor text links.

We also found 153 links that are redirected. That’s a very small number compared to comparethemarket.com’s total amount of backlinks. 153 links couldn’t affect its link profile that much.

You also see that comparethemarket.com was using link networks like web directory networks or business directories (e.g. xxx.lodover.com) in combination with other unnatural pattern like lots of weak links. Also, the website’s link growth was much faster than the link growth of its competitors before the rollout of Penguin 2.0, so an unnatural link profile could have caused the drop.

Also comparethemarket.com’s link risk is very high, only 32% of its links are healthy, but then, according to the CDTOX analysis, they blend in!

But my best bet after this research is to blame technical issues or keyword over optimization filters, because these issues can be fixed very fast so that can be an explanation why the website got back its rankings quickly.

What do you think?

If you have any ideas do not hesitate to leave a comment or to contact me!

Thx
Harry

This case study was written by Harry Tschuggnall, CEO at Mediabase, and long-time user of Link Research Tools and Link Detox.


A word from Christoph C. Cemper

Certified LRT Professional
This analysis was conducted and post written by Harry Tschuggnall, now Certified LRT Xpert.
Harry showed proficiency in doing a backlink profile and SEO audit using the LinkResearchTools and pointed out quite some critical issues to fix. He even further brought up more issues that make this one of the most complex audit cases, and I definitely look forward to community expert feedback and follow-ups.
As usual with Harry, I am thrilled by the detail he went into in this case study. Therefore, I’m very happy to certify Harry as the first Certfied LRT Xpert by approving and publishing his research on our site.

 

 

Certified Xpert Certificate

 

Certified LRT Professional

Harry now also achieved the first Certified LRT Agency certification for his company Mediabase GmbH. This now allow Mediabase to market their brand and service with this highly sought after certification and receives consulting leads from us free of charge.

Goal of our certification program is to provide our user community and clients a high quality service, and our certified experts are key to that.

I look forward to future work and personally recommend Harry Tschuggnall and Mediabase GmbH to work with you whenever you get a chance!

 

Harry Tschuggnall Certified Professional

This analysis was created using our Superhero Plan and the new Competitive Link Detox tool extensively. The Superhero plan allows you to perform professional SEO and backlink analysis for your own or your competitor’s sites, but if you only need Competitive Link Detox (CDTOX), that’s also included in all new Link Detox Pro plans and up.

Don’t miss our next FREE LRT Associate Training. Get one of the last spots if you’re up for 8 hours of free professional training with Christoph C. Cemper.

Don’t forget to signup for our notification list to get informed on other Case Studies on Google Penguin updates.

Are you ready for the next Google Penguin Update?

What do you think about this Case Study?

Let us know!

Harry Tschuggnall

Harald Tschuggnall is the Founder of Mediabase, an Austria based SEO and Link Building Agency. He’s holding a Certified LRT Xpert accreditation and runs the first ever Certified LRT Agency thanks to his extensive knowledge and case-studies published.

Harry Tschuggnall

Harry Tschuggnall
Google Penguin 3.0 is here any day.

Hire one of our 29 Certified LRT Professionals, Certified LRT Xperts and Certified LRT Agencies to conduct your Penguin 3.0 link audit – TODAY.

19 Comments

  1. Hendrik August 28, 2013

    Hallo Harry,

    interesting case study!

    As with a lot of things lately it is very hard to tell precisely which update/link had the effect to make or break it.

    I have examples of sites that ranked top 40. Then all of a sudden it went up to top 10. No links having been placed. Onpage everything remained the same. A few weeks later it goes back to normal.

    And as most know already: The SERPs change on a daily and even hourly basis. This makes it harder and harder to tell which links are good/bad.

    So a year back you could tell 1 – 4 weeks later whether your measures were good. Nowadays this only works with high class backlinks that make sudden big difference.

    In this case the competitors are just better and Google finally realized it ;)

    Thanks for the good read

    Hendrik

    reply
    • Harry Tschuggnall August 28, 2013

      Hi Hendrik,
      thx for your command!

      Yes you are right its getting harder and harder to get the right links ;-)

      see you
      harry

      reply
  2. Kunle Campbell August 28, 2013

    Interesting and very detailed case study Harry.
    Would you happen to have an idea of how many months Penguin 2 Penalty recoveries take after a link clean up has occurred?
    Are we looking at over or under 6 months?

    reply
    • Harry Tschuggnall September 1, 2013

      Hi Kunle,
      first penguin was in April 2012 second in Mai 2013 yes I think that Google will get better and faster so I would guess around February 2014.

      Thx.
      Harry

      reply
      • Kunle Campbell September 3, 2013

        Harry – does this mean that after a link clean up, I would have to wait up until the next update in approx. February 2014 to know if the Penguin Penalty was lifted?

        reply
        • Harry Tschuggnall September 4, 2013

          Hi Kunle,
          yes if it was a penguin – you would have to wait! (or 301 redirect to a other domain / if you cleaned up already)

          ;-) harry

          reply
  3. Ingrid Cliff August 29, 2013

    Brilliantly thorough analysis Harry – I felt like I was reading a detective story trying to follow the clues. If the Link Detox system could add in a crystal ball in the next enhancement I suspect that would help!

    reply
  4. Tobias Sasse August 29, 2013

    Harry,

    thank you for sharing the exhaustive analysis with all those screenshots and your profound knowledge!

    A brief look at the visibility for me clearly shows that it’s for sure (!) no Penguin. The times/durations and the amount of visibility loss don’t match it. The site seems to have been released from another penalty or filter (maybe a soft Panda influence) between March and June. Temporary effects like this are quite common nowadays. The site fell back to the previous (normal) level, afterwards.

    It was delighting to follow your investigation approaches and ideas, anyway!

    Tobias

    reply
    • Harry Tschuggnall September 1, 2013

      Hi Tobias,
      thx for your command!

      Yes your are right there are very strange patterns if you look at the keywords.
      Loosing most of the important keywords for 2 weeks is definitely strange.

      thx
      and have a nice day!

      Harry

      reply
  5. GODOVERYOU August 30, 2013

    You’ve definitely done a lot of work here!

    Initially I went in this direction for my analysis but quickly decided that inbound link metrics were only part of the problem and that they were being used in comparison to content freshness – which is possible per Google’s use of the BigTable management system.

    Anyways, if you want to get a feeling for how I tackled the problem (minus cool screen shots) it’s at http://www.godoveryou.com/matt-cutts-fumbles-penguin-2-leak

    reply
  6. Tom August 30, 2013

    Hi Harry,

    Absolutely loved reading the analysis you have done for this post!

    The tools you are using to give you deep insight into the effects of historical link building techniques seem to really help.

    It would be really interesting to see the clean up work the other insurance comparison sites have also done not to be effected, as response time is key when competing for organic ranking in an eCommerce remit.

    Tom

    reply
    • Harry Tschuggnall September 1, 2013

      Hi Tom,
      thx!

      I hope some of the company’s are realizing this problems and starting to protect themselves. In the offline world thy protecting themselves against everything online nobody is taking care.

      Very risky if you do your business online!!

      harry

      reply
  7. Gaurav Jain August 31, 2013

    How we can check
    SEO Visibility before and after Penguin 2.0 ?? which tool do this

    reply
    • Harry Tschuggnall September 1, 2013

      Hi Gaurav,
      you could use the tools: searchmetrics or sistrix.

      thx
      hary

      reply
  8. webwatcherreviewed.blogspot.com September 12, 2013

    Internet has become all prevalent in the modern day world and one cannot do without the same even for a day.
    If your husband is really cheating on you for the second time
    you will be able to find out for sure with computer monitoring software.
    system, you can prioritize certain keywords or phrases that you think are suspect,
    and our software will specifically filter for them.

    reply
  9. Cara Donofrio September 17, 2013

    Computers have become such a requirement, such
    a precious reference and acknowledged way to work, play and communicate, that they are not going away.
    Mainframe computers have lots of hard disk drives having more than 30 TB
    of storage capacity. You will have seen how the
    practice moves in the starting and how light seems when
    either of the figures moves their wands.

    reply
  10. Adrian Land September 19, 2013

    Thanks Harry for the very long and detailed analysis. It was a test of patience to take all the screengrabs as much as review and piece together the data.

    I enjoyed your analysis and LRT analysis – I have recently become a customer because of these tools.

    From a holistic analysis too, there were many company changes. They had many microsites for different campaigns they had run that got blended [thrown together IMHO] with strange redirects and lack of canonicals and other tech bits and pieces etcetera, you get my point.
    The site was [from being a customer] was crippling slow for a while and many landing pages that [you eluded to] that didn’t/don’t add any value based on their title e.g. Halifax CC’s.

    I really enjoyed your analysis and definitely not criticizing, actually far from it. But, a general observations as an SEO, is that we all love data and we jump straight in. I have found on a good few cases that the cause of a traffic drop can be a) a competitor is just better, b) a dodgy release/downtime/500 not 503 etc c) someone has changed robots.txt/url structure etc d) an SEO was doing ‘dodgy’ stuff.

    I guess this why SEOs are corporate chameleons and have an inquisitive mind who can explore all the known issues that therefore leave the possible causes on the table.

    Thanks for taking the time to write this up, it is appreciated.

    reply

Leave a Comment

 

Compare Plans and Pricing

Check out our other tools

 
LinkResearchTools is trusted by the world’s most respected brands.

LinkResearchTools trusted brands

Link Research Tools is an important weapon in any serious Internet marketer's arsenal.

Chris Winfield, CMO & Managing Partner at Blueglass,

Finding the 'right' link sources with little effort and monitoring and optimizing the own link building activities. The Link Research Tools of CEMPER.COM offer the tools for it. You can analyze several SEO data sources at one time together with CEMPER.COM's own Juice parameter. Combining the individual tools offers a new, previously unutilized link building potential.

Marc Schmitz, goFeminin.de GmbH