I am always running the logs on my websites seeing where the traffic is coming from. One of my websites has a blog and I post to that blog almost every day. My initial set-up for that blog had seven posts on the homepage and the archiving was done weekly. After I had been blogging for a few months I felt that the archive menu was getting pretty long so I changed the archive setting to monthly. This reduced the number of archive pages on my site but surprise... blog traffic went up almost immediately. After doing some study I could clearly attribute the increase in traffic to keyword combinations between two different posts. For example, I might have blogged about "Garlic Baloney" on the 5th and "Peanut Butter Sandwiches" on the 15th. To my surprise I was getting traffic on terms such as "Baloney Sandwiches" - where the SEs found the parts of a single querry in two different blog posts. After deep snooping I found that almost 1/3 of my traffic is from these... let's call them... "cross post keywords" (CPKs). This is bonus traffic since I would never have shown in the SERPs for any of these keywords while running the weekly archive.
Google seems to be much more effective at bringing me CPK traffic. Yahoo and MSN bring very few visitors from it. So traffic went up, but was that traffic happy to enter my site since they didn't really find what they were searching for? I checked the "average visitor time on site" and it remained fairly constant. So either these CPK visitors read the blog like everyone else or spent their time on site searching for a topic that didn't exist. I'll assume that a little serendipity occurred.
To take this a step farther, I changed the homepage of the blog from a seven post length to a 20 post length. Average traffic went up there too and again the CPKs were hauling in the bonus.
I feel that the 20 post homepage is the best way to go since it makes my blog look like it has a lot more information and a wider variety of topics... and serendipity is a good thing. What do you think?
CPKs and Blog Serendipity
Keyword Research
The author's views are entirely his or her own (excluding the unlikely event of hypnosis) and may not always reflect the views of Moz.
Way to go BigDoug. You are ahead of us all!
Very nice EGOL! My blogs as you know have been 20 - 30 long per/page for the last year and G is happy with it.
D
Nice article, EGOL - BTW, where've you been hanging out lately? Don't see many EGOL posts in the circles I'm traveling in these days - maybe need to expand those circles...
To the point, this is great information, and I believe is applicable to our forum site. We run a portal-type page that pulls a wee bit of content from the forums - but I believe the place it pulls from and the amount could both be changed to improve our traffic.
Thanks for the tip.
Thanks Robert, I am glad that you liked the article. I am mainly posting at https://www.cre8asiteforums.com,and commenting here at SEOmoz. But have been very busy with an educational software contract - on top of my normal work.
And here I read somewhere that it is best to break up your content to different pages if it is not closely related to avoid content overlapping (ie pages that are not FOCUSED on topic)... And now this tells me the more content the merrier whether it is on topit or not! So what is it then!? lol
Sorvoja: If you get a few visitors a month from Google but rank as #879, you in my experience could get a whole more traffic when you get among the top fifty.
Good idea. I just went into a few of the old posts that are drawing the traffic to optimize a little. (The CPK terms are closely enough related that I can easily get the exact phrase into the post in natural language.) Am also writing that Baloney Sandwich special page.
Travis: How many items were in each archive, and how many archives did you have ?
20 - 30 in each archive after the switch. 4 - 6 before
I've had this same experience on one of my blogging sites. After changing the amount of frontpage posts from 4 to 10 SE-traffic to the frontpage went up by 40 percent. Adsense conversions went up too and that was the intention :)
Now that you have the traffic data your could tailor some new content to target "Baloney Sandwiches" directly.
I never visited performancing.com before. I need to read some of their articles. Thanks for pointing to their site.
I see that they have 20 articles on the homepage. I wonder what would happen if they increased the number of words that shows on the homepage page for each article? My bet is that they would pull more SE traffic but I wonder if that would cut into the number of pageviews done by devoted readers? Don't know until you try.
I think that this was successful on my blog because it has a lot of technical terms that are typed in by searchers in an enormous diversity of ways. The same approach might be less successful if your posts are common language or if the number of keyphrase variants are few.
Bottom line is that folks should experiment with different formats to see what happens. I think that I might extend this by making archives by date and by category (right now I am only archiving by date). My guess is that such an approach would nearly double my CPK traffic. What do you think? Or, would this put too much dupe content on the blog since the same post could appear simultaneously... on homepage, on category page, on monthly archive page and on single post page. Too many?
Sorvoja: Now that you have the traffic data your could tailor some new content to target "Baloney Sandwiches" directly.
That's a great idea. I can do that for the CPKs that pull the most traffic. Are you searching for baloney sandwiches?
Baloney Sandwiches are a american thing I only got a wage memory of their taste. When targeting CPK it is an idea not to look at the traffic alone, but also of how much the traffic could increase. If you get a few visitors a month from Google but rank as #879, you in my experience could get a whole more traffic when you get among the top fifty.
For long tail keywords I look atthe keyword in the following order: -relevant for the site. -search engine rank. -brings in a lot of traffic already.
Sounds like a load of baloney. Just Jokes.
Very impressive stuff. How many items were in each archive, and how many archives did you have ?
We have a series of news items, which are un-archived, but soon, for the purpose of clarity, we need to start categorising them.
Perhaps the categorisation of your blog stories with the introduction of a node above them helped with the traffic as well.
We only write about 20 news items a year, so I will use an annual archive.
Thanks for the insight.
I'm very impressed, EGOL. You're basically saying you used aggregated combinations of posts to bring long tail searchers and succeeded in converting them to visitors who find, if not exactly what they want, something that they want.
It's a brilliant tactical use of visitor analytics and creativity - thanks for sharing! I bet the Performancing crew will be jealous :)