How to prove a site is suffering from conflicting content
Written by: Alexander van Aken
After reading many interesting articles, including ‘Identifying keyword cannibalization for better SEO’ and ‘Solving suspicious flux with a canonical strategy‘, by Pi Datametrics, I decided to investigate whether some of the websites I manage have internal conflict issues.
MANUALLY IDENTIFYING INTERNAL CONFLICT:
Results for Hostnet.nl
For the Dutch domain of Hostnet.nl, multiple landing pages existed that competed with one another for the theme of build your own website (eigen website maken)
On the 26th of January, three of those landing pages were 301-redirected to https://www.hostnet.nl/eigen-website-maken for the purpose of increasing organic traffic and keyword rankings in Google.nl.
Results were monitored during the months of February, March and April. During this period no changes to on-page SEO were made, nor were any backlinks directed to this page.
For this case study, data was analyzed within Google Analytics, Google Search Console and a keyword rank-tracking tool.
Google analytics data
Straight after the implementation of the 301-redirects there is a clear increase in daily organic traffic to https://www.hostnet.nl/eigen-website-maken
Is this uplift temporary?
It’s clear from the plot of weekly traffic that the effects of the 301 redirects are long term.
Is this uplift seasonal?
You could assume that January is always a slow month for this landing page. Unfortunately, there is no data from January 2015 to compare YOY performance and validate this theory. The data only goes as far back as 21st of July 2015.
This chart demonstrates the notable uplift between the periods preceding and following the 301 direct implementation on January 26th. The orange line represents weekly organic traffic from 5th Nov – 25th Jan and the blue line displays significantly higher traffic thereafter.
The number of organic visitors, rose from 501 to 3,429 between both time periods. This constitutes a substantial increase of 584%.
Google Search Console data
In GSC it is clear that the number of clicks also increased structurally after the implementation of the 301 redirects. Unfortunately, data is retained for 90 days only so we could not look at any data prior to the 12th January 2016.
Did keyword rankings associated with this landing page increase?
Rankings for 10 keywords associated with the landing page were tracked prior to and following the implementation of the 301 redirects
In almost all cases the keyword rankings increased structurally.
In total, there was a net increase of 20 positions, seven of which occurred on page one in the SERPs.
This is important to note, as it’s harder to move the needle onto the top page of Google from beyond page one.
Keyword ranking success over time
The following graphs demonstrate the change in rankings over time. Each chart represents a keyword grouping based on similarity and relevance. A plot combining all data would be too hard to interpret.
Out of 10 keywords, eight increased in organic rankings. The phrase online website maken (build website online) decreased in rankings.
A reason for this may be that it’s the only phrase containing the word online, which Google might consider more relevant for another landing page within the hostnet.nl domain.(i.e. the landing page https://www.hostnet.nl/website/website-maker ranks well for terms like website online).
Arguably, we can further optimize on-page elements so that Google will start to associate our chosen landing page with online website maken.
Manually identifying internal conflict: It can be done, but only on a small scale, over a long duration
Hostnet experienced internal conflict for ‘Build your own website’ related phrases.
This problem was resolved by 301 redirecting three conflicting landing pages, which shared similar content, to https://www.hostnet.nl/eigen-website-maken.
On the basis of our data, we can conclude that the 301 redirects led to a structural increase in organic traffic, clicks and keyword rankings.
The results were satisfactory, but in total we used six tools to analyze keyword rankings, traffic and internal competition between landing pages. It took us over a month to get the results presented in this case study.
Also, we know that we are only scratching the surface here. Our method is not scalable and inherently limited, whereas with daily data from Pi Datametrics, internal conflict could have been detected sooner, and this work could have been done in a matter of days, as opposed to months.
What’s more, the majority of internal conflict issues would be solved entirely using Pi, whereas we currently estimate that we have only addressed about 10% of all issues affecting our domain.
As ever, we’d love to hear your thoughts on this number one search issue, so feel free to leave your comments below.
What’s more, if you think your site has been affected by internal conflict, book a demo to find out how Pi can help.