- Joined
- Jun 15, 2014
- Messages
- 148
- Likes
- 474
- Degree
- 1
Today's Newsstand: Portfolio's & Canonical Gifting
The basic summary of the above post is quick to write, but the power behind the play is pretty interesting.
Background
Basically, this guy Nick LeRoy has a client that has two sites in the same niche for more tightly targeting users.
The sites are very similar and get very similar content posted to them. But, as Ryuzaki complains about, you never know which site Google will favor for whatever reason. Nick discovers that certain sub-niches of content perform better on Site A than they do on Site B, and vice versa, though they're pretty much the exact same.
He suspects this may have something to do with topical relevance and the amount of each content in each sub-niche posted per site. Or maybe the backlinks. He doesn't know.
Experiment
So Nick, with a slick bit of thinking, starts copying posts from Site A to Site B, and Site B to Site A, depending on the topic and if it will perform better on the other site. The key is that he's applying canonical tags to these posts. If a post moves from Site A to Site B where it will do better, he adds a canonical on Site A telling Google to give the link credit and traffic to Site B's version of the article.
Results
His post makes it sound like he didn't pay attention to the results of how well this was working until some update on the theme or a plugin caused the canonicals to break, which is what the image below shows:
You can see how once this article received the canonical tag from the other site it took off, then there was the problem that was then fixed. The site also got hit like nearly everyone did around August 2018, so while it looks like the post didn't recover after fixing the canonical, it did. They're just losing traffic in the same series of updates that's been discussed here on BuSo.
Implications
A lot of people talk about having several sites in a niche you're starting to dominate. The reason is, you have insight into all the keyword data and competition levels, what works and doesn't, can leverage social properties, whatever. It makes sense to get even more traffic out of the same SERPs even though it diminishes rapidly. Two or three sites in the same niche is probably the maximum I'd go based on SERP CTR's decaying so quickly from position #4 and down.
If you're already in this position, you can look at the keywords each site ranks for and try to find a pattern. For instance, if you have 3 sites about Tools that are all similar, but Site A performs better for Circular Saws, you can move all of you circular saw content from Site B and Site C over to Site A and give it canonicals.
I probably wouldn't recommend syndicating every post across every site and trying to canonicalize each one. That's getting into "content farmer" territory and SEO schemes. But if you're maintaining real brands and can get major gains out of doing this with perhaps 10-15% of your content, then I'd give it a shot, if and only if you have the ranking data that shows you'll knock it out of the park. Otherwise you're going to create an indexation nightmare that takes a while to get undone.
The basic summary of the above post is quick to write, but the power behind the play is pretty interesting.
Background
Basically, this guy Nick LeRoy has a client that has two sites in the same niche for more tightly targeting users.
The sites are very similar and get very similar content posted to them. But, as Ryuzaki complains about, you never know which site Google will favor for whatever reason. Nick discovers that certain sub-niches of content perform better on Site A than they do on Site B, and vice versa, though they're pretty much the exact same.
He suspects this may have something to do with topical relevance and the amount of each content in each sub-niche posted per site. Or maybe the backlinks. He doesn't know.
Experiment
So Nick, with a slick bit of thinking, starts copying posts from Site A to Site B, and Site B to Site A, depending on the topic and if it will perform better on the other site. The key is that he's applying canonical tags to these posts. If a post moves from Site A to Site B where it will do better, he adds a canonical on Site A telling Google to give the link credit and traffic to Site B's version of the article.
Results
His post makes it sound like he didn't pay attention to the results of how well this was working until some update on the theme or a plugin caused the canonicals to break, which is what the image below shows:
You can see how once this article received the canonical tag from the other site it took off, then there was the problem that was then fixed. The site also got hit like nearly everyone did around August 2018, so while it looks like the post didn't recover after fixing the canonical, it did. They're just losing traffic in the same series of updates that's been discussed here on BuSo.
Implications
A lot of people talk about having several sites in a niche you're starting to dominate. The reason is, you have insight into all the keyword data and competition levels, what works and doesn't, can leverage social properties, whatever. It makes sense to get even more traffic out of the same SERPs even though it diminishes rapidly. Two or three sites in the same niche is probably the maximum I'd go based on SERP CTR's decaying so quickly from position #4 and down.
If you're already in this position, you can look at the keywords each site ranks for and try to find a pattern. For instance, if you have 3 sites about Tools that are all similar, but Site A performs better for Circular Saws, you can move all of you circular saw content from Site B and Site C over to Site A and give it canonicals.
I probably wouldn't recommend syndicating every post across every site and trying to canonicalize each one. That's getting into "content farmer" territory and SEO schemes. But if you're maintaining real brands and can get major gains out of doing this with perhaps 10-15% of your content, then I'd give it a shot, if and only if you have the ranking data that shows you'll knock it out of the park. Otherwise you're going to create an indexation nightmare that takes a while to get undone.
