Make sure that your website is always ready for the real-time Penguin update. You can learn more in this FREE WEBINAR about what Google Penguin 4.0 (real-time Penguin) means for you.
The real-time Penguin update makes Link Risk Management - the practice of proactive link audits and disavows - more important than ever for your business. You can manage the risk of your backlink profile using the LRT Superhero Plan. This plan allows you to perform link risk management (link audit, link cleanup and link disavow boost), competitive research, professional SEO and backlink analysis for your own or your competitor's sites.
Penguin 2.0 - Learn from Icelolly.Com's Penguin Filter
I am very excited and proud to present our 10th SEO case study on Penguin 2.0 drops in 2 months - A deep dive into the Icelolly.com Google Penguin 2.0 derankings.
This research was done and written by our latest Certified LinkResearchTools Professional Derek Devlin.
I think this is one of the most-detailed, extensive and best structured case-studies ever, so congrats to Derek, you really deserve your certification.
We look forward to your feedback and help spreading the word on this!
Enjoy & Learn!
Christoph C. Cemper
Table Of Contents
- Aftermath of Penguin 2.0 - Icelolly.Com Case Study
1.1 Aim of the Case Study
- 2.0 Impact of Penguin 2.0 on Organic Search Visibility
2.1 Indexation - Site: Search
2.2 Sistrix VisibilityIndex
2.3 Significant Keyword Drops
2.4 Visibility Summary
- 3.0 Topline Metric Comparison using QDC
3.1 Quick Domain Compare Summary
- 4.0 Identifying Hypotheses for Ranking Drops
4.1 Anchor Text Diversity
4.2 Followed / NoFollowed Links
4.3 Link Type
4.3.1 What Redirects are in place and why?
4.3.2 Are Redirects Harming Icelolly.Com?
4.4 Deep Link Ratio
4.5 Geographic Location of Links
4.6 IP Address Link Diversity
4.6.1 High-Risk Links
4.7 Links from De-indexed Sites
4.8 Links from Malware Sites
4.9 Link Networks Being Used?
4.10 Link Directory & Article Directories Used to inflate Pagerank
4.11 Overall Link Risk Profile Versus Competitors.
4.12 Link Velocity Trend
- 5.0 Findings & Hypotheses
- 6.0 Conclusion
- What became of holidaygems.co.uk? I hear you ask!
On June 11th 2013, Econsultancy published an indicative list of Penguin 2.0 “losers”.
The study tracked the search visibility of websites competing for 500 major terms across many popular verticals, including retail, finance, entertainment, property and gaming.
The methodology of the study was to compare ranking performance just before the Penguin 2.0 update (21st May) versus ranking performance just after Penguin 2.o update (24th May).
The estimated change in click volume per domain was based on typical position-by-position click-through-rates, with losing domains defined as those that lost at least 3000 of their tracked click volume and had greater than 20% decline in click change.
Top of the list of losers, was the UK travel price comparison website - Icelolly.Com.
Econsultancy speculated that Icelolly lost 105,712 clicks, equating to a decrease of 75% search visibility across the terms surveyed. This is a staggering loss of search visibility for a well-known brand, which would indicate that Icelolly have been using tactics outside the acceptable bounds of the Google Webmaster Guidelines.
A drop of this magnitude will surely be hurting Icelolly’s commercial success at what is a critical time of year for all travel websites.
Penguin “Winners & Losers” - Image Source:
The aim of this case study is to better understand the reasons why Icelolly.com suffered such a steep decline in search visibility; thereby helping us gain a deeper understanding of the factors that most likely triggered the Penguin 2.0 filter.
My goal is to identify a set of hypotheses by comparing and contrasting www.icelolly.com with high-ranking competition in their core verticals.
Let’s first get an up-to-date snap shot of Icelolly.com’s search visibility as of the 1st of July 2013.
Using the site colon operator, we can first check to make sure that Icelolly.com is still indexed by Google and that the site has so far not received a blanket ban and de-indexation of all pages.
Google still has well over 400,000 webpages indexed for the site, including the homepage so it’s clear we have not been completely de-indexed at this stage.
Next we can use Sistrix’s VisibilityIndex to assess the extent to which Penguin 2.0 has negatively impacted Icelolly.Com..
Straight away you can see that there has been a substantial loss of organic search visibility coinciding with Google’s Penguin 2.0 Update.
There has been a sharp decline in the SISTRIX Visibility Index score for the domain, which peaked at 0,7028 towards the end of May and dropped to a low of 0,0668 as of the 1st of July.
This represents a 90.4% drop – Wow!
Clear evidence that Icelolly.com have been hit hard by Penguin 2.0.
By looking at the “Negative Changes Report” and the “Dropped keywords Report” in Sistrix, we can see a lot of keywords that have declined in rankings, as well as the many keywords that have been completely dropped.
Sample of the NEGATIVE Changes Report
Sample of the DROPPED Keywords Report
We have a large number of possible keyword verticals to analyze so it makes sense to prioritize the keywords into some form of order ready for more in-depth investigation.
We could sort the keywords based on the highest number of places dropped or alternatively the search volume of each keyword but in isolation I feel that these approaches are sub-optimal because they aren’t fully representative of the keywords overall asset value to Icelolly.
Instead I want to take a multi-faceted approach to uncover the keywords that not only lost a lot of visibility but also most likely yielded the highest value to Icelolly.com prior to the drops. I aim to identify the highest value keyword vertical to be the focus for my attention once we move into the competitive link analysis.
To do this I come up with some rules upon which to score each of the keywords.
My first step is to export and combine the Sistrix ranking data with the Google keyword tool search volume and CPC data in Excel.
I’m most concerned with finding a reasonable starting point to get going so I decide to set some quick and dirty thresholds to get us on our way.
I use my own judgments based on the overall data set, as well some gut feel to determine what a high value keyword looks like and I settle upon the following criteria to determine what the highest priority keywords are:
- Had a starting rank lower than or equal to 10 – i.e. were on the first page.
- Dropped more than 30 places.
- Competitive in the SERPS - higher than 0.80 according to the Google Keyword Tool.
- Over 1000 Exact Monthly Searches for the term in the UK market.
- Approx. CPC higher than £0.70.
Using the ‘conditional formatting’ feature in Excel, I highlight the cell values that exceed each of my pre-determined thresholds.
By color coding the cells that meet all of the criteria, I can see that 3 keywords meet all of my thresholds.
“Cheap Holidays” – Dropped from 1 to 65
“Cheap Holidays” – Dropped from 1 to 51
“Cheap Holiday Deals” – Dropped from 1 to 37
These keywords high commercial intent, significant search volume and all were high performing keywords prior to Penguin 2.0.
Lowering the thresholds would provide a wider data set to analyze, which would be prudent in a real life scenario but for the purposes of this case study I’m going to focus on the highest value keyword.
Given that “Cheap Holidays” has the highest search volume and by a significant margin, I consider this to be the most important vertical.
“Cheap holidays” will be therefore be the focus of my analysis for the purposes of this case study.
- Icelolly dropped 90.4% in the Sistrix Visibility Index since Penguin 2.0
- Icelolly.com has NOT been completely de-indexed by Google.
- The most commercially damaging keyword drops occurred in the “Cheap Holidays”, “Cheep Holidays” and “Cheap Holiday Deals” keyword verticals.
Now that we have ascertained that Icelolly.com lost significant search visibility it’s time to find out what’s going on here.
Let’s dive in to Link Research Tools.
My approach when conducting a backlink audit is to first use the quick analysis tools to build my understanding of the site in question and try to identify link-building tactics being used.
I’m sure some SEO’s like to start with the Quick Backlinks Tool. However, I prefer to start my backlink audits from the Quick Domain Compare tool because I find that it’s more valuable to get a snapshot of the topline metrics for the site versus competitors before delving into the sites own backlinks.
QDC is a great way to get a bird’s eye view of what is going on with the site and uncover some initial clues that can help you decide where to focus during the more in-depth stages of analysis.
I begin by using the “find competing pages” function to pull in the top ranking competitors from our highest value keyword vertical “cheap holidays”.
The tool returns the top ranking sites on Google.co.uk:
Handily the tool returns the top ranking sites for “cheap holidays” and gives me a comparison table of the 4 domains side-by-side with Icelolly.com
What can we learn from this initial topline analysis?
- Trust is significantly lower than power.
As we have seen with many of the Penguin 2 victims featured in the previous LinkResearchTools case studies, there is a disparity between Cemper Power and Cemper Trust. This is symptomatic of overly engineered links that are passing higher Pagerank relative to the trustworthiness of the site the link is being placed on.
This suggests that that there is likely to be a high proportion of low quality links to be found pointing at Icelolly.com. Links with Power but likely very little trust – this is something we will need to explore as we progress through our analysis.
Travelsupermarket and Thomson both have an even balance of Power and Trust. This is more what you would expect from a premium travel brand, likely a mixture of high quality links combined with a spread of links from highly trusted sources on the web.
The anomaly that is somewhat perplexing is that holidaygems.co.uk, which is currently ranking well in position 4 for “cheap holidays”. This site has a clear disparity between power and trust and I would suggest that it is only a matter of time before that site suffers the same fate as icelolly.com.
Interesting though how holidaysgems.co.uk has so far dodged the Penguin bullet but for how long? I’m sure in time we will see!
- Lowest Number of keywords Ranking
Icelolly clearly has the least number of keywords ranking in comparison to the other leading sites in the “Cheap holidays” vertical.
A possible indicator that Icelolly.com’s focus has been too narrow, with little attention being given to building long-tail content.
- Lower ratio of Site wide links than competition
At a glance, it looks like Icelolly have not overdone it with the NUMBER of site wide links.
Both TravelSupermarket and Thomson have total backlinks in the millions coming from less than 10,000 unique domains. This is clearly a far more significant ratio of total links/unique domains than that of Icelolly.com.
To confirm this I hop over to the link juice tool and run a “detailed” analysis using only the site wide link metrics.
This is one of my favorite ways to analyze site wide link ratios.
The link juice tool confirms that Icelolly do not have a significantly high site wide ratio relative to competitors.
Although Site wide links appear to have not been used in as high proportion as the main competitors that does not mean that site wide links are not harming Icelolly.com.
The number of site wide links in itself most likely did not trigger the Penguin filter but the techniques used for placement of the site wide links could still have a negative impact on the overall link graph.
I am interested to study the impact of including as well as removing, site wide links from the link profile of Icelolly.com. This is something I will endeavor to look at as we move into the deeper analysis tool.
- Lower .Edu and .Gov links than main competitors
Jumping back across to the QDC report.
Icelolly have 0 .edu and .gov links, this is substantially less than the main big brand competition. However, on a par with holidaygems.co.uk who also have 0.
It is unlikely that the lack of .Edu and .Gov backlinks has been the cause of a penalty from Penguin 2.0 but it could be a contributing factor towards the lack of CEMPER Trust attributed to the site, inadvertently causing Icelolly.com to lose out when the filter was implemented due a lack of trust and a over abundance of power.
- Competitor Anomaly - Holidaygems.co.uk
At this point I’m tempted to go off on a tangent.
Something is clearly suspicious about holidaygems.co.uk, this instinctively feels like an outlier and a site that shouldn’t be ranking 4th for a competitive term such as “cheap holidays”.
A domain comparison in Sistrix shows that holidaygems.co.uk has come from no-where and benefited from a sudden boost coinciding with Icelolly.com’s dimise at the same time as the Penguin 2.0 Update.
I am interested to see how holidaygems.co.uk progresses over the coming weeks as I work through the Icelolly.com analysis.
The Quick Domain Comparison has already provided a great starting block from which to start our deeper analysis.
I am now on the look out for:
- Weak links
- Too Narrow Focus – anchor text?
- Site wide link strategy
Rather than having a preconceived idea about why Icelolly have dropped in the rankings, I want to base my hypotheses on observations made from first studying the backlink profile and then cross examining it against high performing competitors, this provides context and helps identify anomalies.
I’ll start with the Quick Backlinks Tool.
I first run the report WITHOUT the site wide link filter ON so as to understand the impact of the links as a whole.
Wow! The first thing that jumps out at me is anchor text.
This is a good place to start our analysis since other Link Research Tools Case studies have confirmed that this is likely to be a contributing factor in the Penguin 2.0.
Let’s investigate the results of the QBL reports further…
48.3% of links found have exact match anchor text “cheap holidays”!
48.3% for one single keyword is very, very high.
I want to see if this is being overly skewed by site wide links so this time I set the site wide links filter to 5, so as to get a more accurate picture of how the site wide links are impacting the overall link graph.
“cheap holiday deals” is sitting at 0.8%.
We are now viewing the link profile of Icelolly.com with SITE WIDE links restricted to 5 per site.
Instantly, we can see how the anchor distribution has changed!
Site wide links are clearly causing a massive proportion of the exact match anchor text, particularly for money anchor texts.
“Cheap Holidays” is no longer the dominant anchor text – instead of 48.3% it’s now only 3.3%. This may still be too high, especially when combined with the other exact match money anchors, but at least the tag cloud distribution is starting to look more natural.
“Cheap Holiday Deals” has increased relative to the proportion of overall inks, going from 0.8% to 1.9%.
Brand anchors with variations of the URL are more prevalent than before but are there enough?
- Icelolly.com - 3.8%
- http://www.icelolly.com/ - 3.7%
- www.icelolly.com - 3.2%
- http://www.icelolly.com - 1.7%
- icelolly – 1.5%
I intuitively think that this is a negative signal but I will need to confirm that this is out of the ordinary for this keyword vertical by comparing with Icelolly.com’s competition.
Time to hop over to the Competitive Landscape Analyzer. CLA is my favorite tool in the LRT arsenal because of the powerful insights that can be gleamed from the visual output the tool provides.
First I classify 93% of the top 10 sites keywords:
Then we can compare the keyword ratios against the top performing competitors in the “cheap holidays” vertical.
Here are the results with site wide links INCLUDED:
Wow! There are some startling anomalies here.
Not enough ‘brand’ OR ‘other’ keywords used – instead far too many ‘money’ keywords have been used!
Let’s check this with site wide links set to 5, which should take site wide links out of the equation.
Still far too many money keywords used!
The percentage of brand keywords has increased slightly by filtering out the site wide’s so site wide links are playing a part in increasing the money anchor text.
However, that doesn’t hide the fact that even without site wide links; Icelolly.com has massively overdone it with money keyword anchor text links.
We can see just how much they have over done it when we look at the same data illustrated as the ‘absolute’ number of links.
Icelolly.com has more than double the amount of keyword links compared to the average for the whole of the top 10 ranking sites in this vertical.
I think we can say with near certainty that Icelolly.com have been far too aggressive towards building money keyword anchor texts.
This must of contributed to being hit by Penguin 2.0.
With site wide links included – Followed links are 83.6%.
Restricting site wide links to 5 - 82% of links are followed:
Site wide links are pushing the percentage of followed links up slightly but I don’t see anything overly concerning about this at this stage.
Let’s benchmark this with competitors, which we can do again using CLA.
Looks like Icelolly.com have not been hit because of too high a percentage of Followed links.
98.7% of links are text links and 19 links are from redirects:
Let’s see how this compares to competitors:
Text link percentage is inline with the competition, however redirects are 4 times higher. These will need to be examined more closely.
Redirects can be potentially devastating to a site if not handled appropriately and can lead to inadvertently inheriting a spammy link profile. Alternatively, some companies try to harness the power of old expired domains by redirecting their link juice to their own site.
To look at redirects we jump across to the Backlink Profiler because we want to get the most accurate measure of actual live redirects. The BLP has found 178 redirects in total.
First we remove those links that don’t pass link juice, i.e. use the 302 HTTP response status code.
We are left with 159 individual redirects that are passing link juice. This reduces to 45 unique domains once I filter out the duplicate URL’s and URL’s with obscure parameters.
It appears that most of the redirects come from variations of Icelolly brand name, including other TLD extensions as well as possible misspellings.
45 unique root domains looks a bit excessive but given that Icelolly have being going for over a decade, it’s not unusual for a company to of changed domain name a number of times during the course of the business.
Icelolly has no doubt also looked to secure other .tld extensions to stop competitors from cyber squatting.
By checking these domains with the Wayback Machine we can see that some of these domains are older versions of the site before they moved to the .COM.
I have found an interesting clue from this looking at the Wayback Machine snapshots of one of the old domains, the footer has been tagged with a link that says “Ice-lolly.com Network” perhaps the reason for all of these old sites being in existence is that Icelolly was at one point in a time a network of sites?
Interestingly all of the redirects, with the exception of one have been set up to point at the homepage!
This is not current best practice. To transition from an old domain you should map all the pages of the old site to the most relevant pages of the new site.
I have managed to find one instance of them implementing a 301 redirect as it is supposed to be done:
Old site URL:
Redirects to New Site URL:
My priority now is to see if any of the 301 redirects are causing Icelolly.com harm.
Specifically I’m interested in:
- Redirected Domains with a high number of links
- Redirected Domains with high Power & Trust
Most of the redirected domains only have between 10 and 20 links, so relative to the total links for Icelolly.com these are a very small portion – their influence is most likely not that significant.
Redirects with more than 10 links:
From the table above, two redirects stand out because they are the ones with the most links pointing at them.
The top one appears to be a cached / scrapped page from a web service, it has been found to be linking via a rel=canonical. I would contact ntopic.org to get this page removed if possible.
The bottom one looks like a redirect created by a wordpress plugin on an affiliate site of some sort. The page itself has 0 links, so I’m not overly concerned with this but in an ideal world this would be removed also because it’s not relevant to our business.
Although many redirects are in place, from our initial assessment I would say that redirects don’t look like a major contributing factor to the drop in search visibility.
In a real life recovery plan, I would be more likely to go a level deeper and look at the links pointing at all of these redirected domains but for now, let’s move on to the exciting findings!
QBL shows that with site wide links included 67.8% of links point at the homepage:
When more links point at the homepage than the inner pages, I’m always concerned. This is usually a sign of a heavily engineered link profile.
Let’s see how that looks with site wide links restricted to 5:
Here is another example of how site wide links are skewing the link graph.
Homepage links have gone from 67.8% down to 39% when restricted to 5.
We can infer from this finding that the site wide links have mainly been pointed at the homepage.
It’s possible that this is a negative signal, which we will now verify with CLA.
With site wide links set to 5 the deep link ratio looks fine:
But with site wide links activated, Icelolly.com looks like an anomaly again!
Based on this analysis, I believe too high a percentage of site wide links are targeting the homepage relative to the overall link graph.
69.9% of links are from sites hosted in the USA, only 21.7% are from the UK:
Given that the primary target market for Icelolly.com is the UK. This most likely is NOT a positive signal.
We can confirm this with CLA:
Even with site wide links restricted, we can see that Icelolly.com have overdone it with US links.
QBL shows us that there are many links coming from single IP locations. As many as 17.1 % of links come from the same IP address, as do 12.6% of links:
This may not be a true representation because let’s remember that we have initially included site wide links.
So let’s restrict the site wide links to 5 and re-run the report.
Even though the site filter is restricted to 5, there are still a significant number of links coming from single IP addresses. At a glance it looks suspicious because if we divide the number of links in the table above by 5 we can see that there are still single IP addresses with a high ratio of links/domains.
This could be down to the fact that there are a lot of redirects in place being held on the same server, which wouldn’t in itself be an issue.
Alternatively, this could be an indicator of network links?
The IP address with the most links is:
Bingo! This is the same IP address as Icelolly.com.
Drilling a little deeper in QBL shows that this is the case. The IP duplication is primarily being caused by the redirects that we previously discovered.
This explains one of the IP addresses, however the other IP’s do not belong to icelolly.com and could still be link networks.
The best way to assess whether link networks are involved is to fire up link detox. Link Detox is also great for assessing weak and high-risk links, so this leads nicely into our next section of analysis.
To analyze low quality and high-risk links I am going to use Link Detox – DTOX, which I will then compare with the new Competitive Link Detox – CDTOX in order to look for anomalies.
So onto Link Detox…
We can clearly see that Icelolly.com’s average link risk has been flagged as “Very High Risk” – this is a cause for concern but then again, we already suspected that there are a lot of poor quality links in play here.
Only 24% of links are marked as “healthy”!
18% are marked as toxic and need immediate attention and 58% need manually checked because they look suspicious!
Let’s take a look at the toxic and suspicious links and try to understand what rules have been triggered…
Weak & Low Quality Links
More than half of the unhealthy links (60.3%) Triggered SUSP 1 & SUSP 4.
The majority of the unhealthy links are coming from the rules that are triggered because the domain linking to Icelolly.com is very low quality.
SUSP 1 - 784 links from domains with no links – these are virtually worthless links as they have 0 CEMPER Power and 0 CEMPER Trust.
SUSP4 – 630 links from domains that are weak because they have little power and trust and they also don’t rank for their title. These links need manually screened.
This would explain the lack of earned CEMPER Trust for Icelolly.
We can confirm this again by looking back at the CLA report.
126.96.36.199 Relative Link Power
Looking at link POWER pointing back to Icelolly.com:
We can see far too high a proportion of CEMPER Power 0 links; not enough CEMPER Power 2, 3’s, 4’s or 5’s as a proportion of total links.
188.8.131.52 Relative Link Trust
Looking at link TRUST on it’s own:
Too high a proportion of CEMPER Trust 0 links; not enough CEMPER Trust 1, 2, 3’s, 4’s or 5’s as a proportion of total links.
184.108.40.206 Absolute Link Power*Trust
Looking at CEMPER Power*Trust combined:
Same story, as before too high a proportion of CEMPER Trust 0 links; not enough CEMPER Trust 2, 3-4, 5-7 or 8-12’s as a proportion of total links.
220.127.116.11 Absolute Link Power*Trust
It’s very clear that Icelolly.com has way too many low quality links and not enough emphasis has been put on earning high quality links.
23.2% of the unhealthy links triggered the TOX 1 Rule.
The second biggest threat after weak domains are links, which come from sites that have been de-indexed. These are most likely sites that have themselves been penalized for Blackhat SEO techniques.
A massive 23.2% of links are from sites de-indexed by Google – 543 domains – this is big!
This instinctively feels like a crazy high number, which I want to check. How does this stack up against the top performers in our niche?
Competitive Link Detox shows me…
Wow! That is significant, just look at that spike!
Way too many TOX1 domains pointing at Icelolly.com – this had of course contributed to the domains demise and these need dealt with as a priority.
Interestingly, you can see that if these were reduced to a like for like proportion then the link profile wouldn’t look too bad from a risk perspective. It would just require some more pruning of the SUSP1 links to come into line.
A closer look at we can see why these sites have been de-indexed.
All of these sites are shady sites using sub-domains as an old school tactic for building links from content farms!
0.5% of the unhealthy links triggered the TOX 2 Rule.
Although not such a large number as the TOX 1 domains, there are 12 links are coming from sites containing Malware and/or Viruses – these need removed immediately and are never good news.
Let’s take a look at one of these sites:
Bingo! We can see this is not a neighborhood we want associated with our site.
Interestingly, two of these are .pl domains.
Let’s use use the Link Detox filters to assess whether we have an issue with malicious foreign links.
I first group the links by domain and then filter using the .pl extension.
After manually inspecting these sites, I’m happy with link detox’s assessment that the majority of these links are healthy links that are relevant to Icelolly.com, only the “deadly risk” links need to be removed.
I repeat this process for .ch, .cz and .ru domains uncovering a similar outcome, a few nasty links needing removed but in the main the foreign links do not appear to be a threat.
I don’t believe that malicious foreign domains have been instrumental in getting Icelolly.com hit by Penguin 2.0.
4.7% of the unhealthy links triggered rules SUSP7, SUSP8 & SUSP9.
- SUSP7 - This domain has the same IP as 17 other linking domains - possible Link Network.
- SUSP8 - This domain has the same Class-C as 51 other linking domains - possible Link Network
- SUSP9 - This domain has the same DNS as 22 other linking domains - possible Link Network.
Upon closer inspection I find that blog networks are not in use because SUSP7, SUSP8 & SUSP9 have been triggered by shared hosting platforms.
For example, we can see that SUSP9 rule has been triggered by sites are all hosted by blogger, hence the reason they have been flagged as using the same DNS.
Upon manual inspection all these sites look legitimate and I would not pursue link removal requests or disavow for them.
This is why its important to always do a manual review of sites marked as ‘suspicious’ by Link Detox because the rules can only tell you so much, it’s up to us humans to ultimately decide whether a link is legitimate or not.
2.5% of unhealthy links triggered SUSP15 & SUSP16
The longtime favorite of Blackhat SEO’s was using mass-submission software that automated the process of uploading articles and links to directories.
2.5% is a small proportion of our toxic links so it’s not overly concerning.
We can confirm whether this percentage is a threat or not by running a ”detailed” CLA and selecting only the link metrics we are interested in:
4.10.1 Site Type
Let’s take a look at how the type of links Icelolly have been building stack up against the top ranking sites in the cheap holidays vertical.
We can see that link directory links are actually 3 times less as a proportion of total links than competitors, which is not necessarily a bad thing. In contrast, they have clearly over cooked the number of blog links at the expense of ‘general’ sites.
The high proportion of blog links has played a role in increasing the proportion of in-content links, which now stands out against the competition also.
4.10.2 Site Theme
Thankfully, site theme does not appear to be such an issue:
The themes of the links pointing to Icelolly.com I would say are in the acceptable ranges.
Competitive Link Detox is the new kid on the block at Link Research Tools and I’m really enjoying the visualization data it provides. Great for mining the healthy links of competitors and of course, for assessing how your toxic links tie up against the competition.
Here’s a look at the Average Link Detox Risk…
What an eye opener!
Icelolly.com’s average link risk is “Very High” when they should clearly be aiming to fall into the moderate risk bucket.
Competitive Link Detox shows us how this is broken down:
You can see that suspicious links aren’t abnormally high as a proportion of the overall link profile.
The key problem area is that there are far too many toxic links and not enough healthy links.
CDTOX shows us another level deeper still and similar to what CLA told us about the lack of really high quality links, CDTOX is confirming this. The “very low” risk links aren’t high enough, whereas the “high” and “Very high” risk links are over-powering the link graph.
We can therefore conclude that Icelolly.com’s link risk profile as a whole has most likely contributed to the drop in rankings, just too many poor quality link signals from bad neighborhoods’ and weak sites, created for the sole purpose of passing Pagerank.
Link Velocity refers to the rate at which a site gains or loses links.
Essentially, we are trying to ascertain whether links are being built at the same rate as competitors or whether there has been some kind of surge in activity either positive or negative that could make the site stand out from the crowd and be targeted by a Google algorithm.
Competitive Link Velocity is great for visualizing spikes in link growth that may stand out as an anomaly versus competitors, signified by the dark areas.
Have Icelolly.com been too aggressive in their link building efforts and was this a contributing factor in tripping the Penguin filter?
Let’s look at CLV. First, here’s the link growth of all links for the last two years.
The link velocity trends don’t jump out as being anything to be worried about but what I find interesting here is that link building peaked in March of 2013; this is when icelolly.com were at their most aggressive in terms of link outreach, this isn’t that significant because competitors were also highly active at this time.
April cooled a little and then May dropped sharply against the activity of competitors and previous months. Perhaps amid news of an imminent Penguin update?
The real tell-tale sign is June. Post Penguin 2.0, half of all sites in the top 10 have cut their link building efforts significantly whilst the others have kept on the gas at a similar pace as before.
Presumably the top sites have been buoyed by their success and given renewed confidence in their tactics.
Looking at the most recent timeframe tells another story.
As we thought, link building in June ground to a stop. No doubt mass panic ensued!
However, activity has ramped up again for July… but I wonder what kind of links are increasing?
CLV indicates that icelolly.com have decided to build more site wide links. To be sure that this is in fact the case we should check that this data is not skewed by links getting deleted because after all, we are looking at a “ratio” here.
Ah. No. It would appear that Icelolly.com has not been removing links - they are in fact trying to build site wide links, probably in an effort to recover.
As we have seen, the site had room to increase this ratio but only if the strategy was controlled. Increasing the ratio in itself is not a threat.
That said, I sincerely hope that Icelolly aren’t targeting exact match money anchor text as before! And… surely they are not pointing them at the homepage!
This case study set out to understand why Icelolly.com has suffered rankings drops after Penguin 2.0.
Based on this analysis, I have formed the following hypotheses as to why Icelolly.com suffered a large drop in organic search visibility after the Penguin 2.0 update.
My hypotheses for factors that contributed to the drop are:
- Site wide links have negatively influenced Icelolly.com’s link graph and disproportionately increased the signals for exact match money anchor texts and homepage link ratio.
- Too many site wide links target exact match money keywords.
- Too many site wide links target the homepage.
- Even without site wide links, too many links still target money anchor text.
- Not enough ‘brand’ OR ‘other’ keywords used.
- Icelolly.com is not a trusted domain because they have too many weak links coming from sites that pass CEMPER Power but very little CEMPER Trust.
- Even with site wide links restricted, we can see that Icelolly.com have overdone it with US links.
- Link profile is out of proportion with the competition, far too many links from weak and low quality domains with Cemper Trust 0. Not enough quality links as a percentage of their overall links.
- Too many blog links at the expense of ‘general’ sites.
- Too many links from old school content farm sites and domains that have been De-Indexed.
- As a whole, link profile is far more toxic than competitors.
Factors that did NOT contribute to the drop:
- Actual number of site links and site wide ratio is not an issue.
- Although many 301 Redirects are in place and not following current best practice, they are not responsible for lack the drop in rankings.
- Although some foreign links such as .ch / .pl and .cz are toxic, in the main they are legitimate links and they should be taken on their individual merits and not blanket disavowed.
- The theme of the sites pointing at Icelolly falls within the acceptable range.
- Link directory links are actually 3 times lower than competitors and so have not caused the drop.
- Link velocity has not been a contributing factor in the lost rankings.
- Proportion of followed vs. no followed links is acceptable and not a contributing factor.
- Link networks have not been used excessively and are not a contributing factor.
In terms of “cheap holidays” on Google UK, they have dropped back 7 places since I started the case study and currently as of the 18th of July are sitting 11th at the top of page 2.
Looking at their Sistrix Visibility Index they are still holding on as one of the big winners from the Penguin 2.0 update.
I will watch with interest over the coming months to see how they progress – keep an eye on my blog as I will monitor their progress and perhaps do some further analysis on them.
As for www.Icelolly.com. As of the 18th of July 2013, they have stabilized their search visibility but have made no solid signs of recovery yet.
They are currently languishing on page 4 for “cheap holidays” in position 32. Resorting to Adwords to try and get the crucial page 1 exposure…
This case study was written by Derek Devlin, CLRTP and was reviewed and approved by Christoph C. Cemper for publishing as Certification work for the Certified LRT Professional level.
A word from Christoph C. Cemper
Derek showed proficiency in doing a backlink profile audit using the LinkResearchTools and pointed out critical issues to fix. I am thrilled by the detail and structure we find in his contribution. Therefore I'm very happy to certify Derek as the latest Certfied LRT Professional by approving and publishing his research on our site.
This is Derek's next step towards the Certified LRT Xpert level which is pre-requisite for the Certified LRT Agency certification. Both will qualify him to receive consulting leads from us. Our goal is to provide our community and clients a high quality service, and our certified experts are key to that.
I very much look forward to future work by Derek. As you can figure, I can recommend Derek Devlin to work with you whenever you get a chance!
What do you think about this Backlink Audit and SEO Analysis by Derek?
Let us know!
With the Superhero Plan you can perform link audits, link cleanup, link disavow boost, competitive research, professional SEO and backlink analysis for your own or your competitor's sites.
You can avoid a Google Penguin Penalty! Learn all about the Real-Time Google Penguin Update in this free webinar.
Join our free 21 Day Link Strategy Training below
Signup to receive snack sized bits of knowledge about Link Research Strategies, SEO Tactics & features of LinkResearchTools platform and technologies.
Bonus: signup you will also receive a free copy of the eBook "7 Golden Rules of Link Building" for immediate download.