What Is Cannibalisation Of Content

Keyword cannibalization means that you have various blog posts or articles on your site that can rank for the same search query in Google. Either because the topic they cover is too similar or because you optimized them for the same keyphrase. If you optimize posts or articles for similar search queries, they’re eating away at each other’s chances to rank. Usually, Google will only show 1 or 2 results from the same domain in the search results for a specific query.

This is one of clients sites, for keywords that they have never ranked for before. It started with combining 15 location thin content articles into 1 good article. After a period of around 3 months it slow went up the rankings to page 1. Now at this point I decided to add some internal linking and a few external links with the keywords I found from GSC

Position 1 Achieved

So eventually the article ranks number 1 for the location and its holding steady. Two weeks later it ranks number 3 nationally, fantastic result. However 2 weeks later it drops out completely, after some digging I can see the term has gone back to the category URL. Further digging I discover that they have never ever previously ranked for this keyword and they should have. It is a service they have provided for the last 10 years but never ranked (they say it did years ago!).

I think it may too be all the sub categories on this service that is drowning it out so I have started to de-index some off them. Anyway the client obviously wants to rank nationally for that keyword again so I commission additional content (very specific niche). So after a couple of months one day I log in and the content I requested to add to the category has been published as an article. So ok lets run with it and see what happens

Cannibalisation Of Content Taking Place

Yes we are in times of covid and we have been having ups and downs with google. The last 2 articles that have been have done very well and didn’t take too long to rank ether. The one I specially commissioned at set out every detail that was needed is nationally number 1 & 0, it has also tripled the traffic to the site. So the article is published along with another one, and its taking an age to get indexed or even appear anywhere in the SERPS.

Then finally we get a few impression showing in google search console so I leave it a couple of weeks before looking again. When I look again it really hasn’t moved that much and when I dig deeper the keywords that are getting impressions are not what I expected. It’s all looking very weird at what google thinks the article is about ( I have another article that is strange as well.) So I look at the location article and its dropped two number two, OMG its been number 1 for months and months.

Deeper Digging in Google Search Console

So I decided to map out impressions, clicks and the amount of terms associated with the location article alongside the national article. At the very time the national article started to get impressions the location article started losing them. Also the location article started to loose keyword terms associated with that URL. This happened on the week of the 4th Sep to the 11th of sep. Also I did the same with category on the same topic and that was gaining more impression and actually had some CTR.

Looking at the data it has been number 1 since around March 2020 and never deviated at all. The impressions is for a weeks worth of data.

As you can see at exactly the same time we start to get impressions (no clicks) the location is declining. Also the impressions are for not what I expected or even wanted.

As I said they have never had any traffic, impressions for this category and it is a major service they provide. It only really starts to get noticed by google the week before the national article starts to get indexed properly. This could also be the effect of me adding more internal & external links to category after the keyword fell into the black whole after briefly ranking for 2 week.

Examine The Backlink Profile

So I do a little more digging into this keyword and decided to explore what backlinks were associated with this keyword. Also to where were they actually pointing to. I have been working on this site around 18 months. So looking at the backlink profile I can see this keyword with links from 2011, and lots of them. But wait they are all pointing at the home page, in fact looking deeper every keyword its pointing at the homepage. What a mess and can it be sorted, not sure yet.

So I ask if they have any records of work done in this period, or actually anything before I started. Absolutely nothing, barely any ranking reports just emails saying, oh we have done this and that here is the invoice. Shocking behaviour hence the tarnished image we have. I give the clients access to my google worksheets, my Active Colab project time tracking and tasks, they also have a link too the tracked keywords.

Reverse The Cannibalisation Hopefully

So of course as with all SEO its a theory and I break the bad news to the client that we have lost the location number 1 spot. At the same time I give hm the plan of what I should do (my theory). So eventually he agrees to it, so I dump the national article but use what content I can win the category. I also redirect the national article to the category because of the backlinks I have built to it. The next stage was to ensure that the internal links are pointing to the category.

The location article I have done nothing to it at all. I should mention that the article that has gone above our location article I did check their backlink profile. Nothing obvious that they have done any work on that URL, it just seems like we gave them the number 1 spot.

Cannibalisation Recovery

Well the drop started happening in August 2020 and it’s not until the middle of February that the article has got back to position 1. It still doesn’t have as much organic terms or traffic but eventually its back at the top. Thats 5 months to recover, haven’t actually spent much time focusing on that 1 single post as the traffic and search term isn’t massive anyway.

Watch this space to see if I can reverse it some more