Update: Although PBNs still work, they now have a history of being targeted by Google and therefore may not be the safest option. This is why we now focus on creating online businesses that are independent of SEO traffic.
This post was actually written a bit over a month ago, but posted privately for interns.
Now that the internship is over I will post it publicly. Note that the EMD update changed a few things, I have indicated those with an update at the bottom of this post. I have painstakingly analyzed a boatload of SERPs and have come to a conclusion: Micro-niche sites are dead (or dying ridiculously fast). Before you hang yourself with your power cord, let me give you a glimmer of hope – SEO is even easier than ever. In a nutshell, content is King yet again. In fact lots of people are ranking now with content alone and zero backlinks. My theory is that there’s a threshold of how many articles you need to be included in a given SERP, and this threshold varies SERP to SERP (likely based on query intent and niche). This has kicked out a ton of people who were doing MNS, despite the fact that their backlink profile was far superior. I analyzed a bunch of SERPs that I used to have a ranking 1 page MNS on that was earning >$50/mo.
These sites got hit back in June/ July, and I being otherwise preoccupied with this blog and the interns, didn’t put up a site to replace them. Well I’m glad I didn’t because these changes happened in July, and those sites wouldn’t have ranked anyway! Who knows how long this lasts, but based on my analysis it seems that there is even higher ROI in SEO because sites with 15 pieces of content are having their subpages rank with 0 backlinks. That means instead of having 15 Micro Niche Sites with 1 piece of content, you have 1 Niche Site with 15 pieces of content. The sites that are doing this are ranking across the board, and none of the subpages have any links, while the homepages have minimal backlinking. Personally, this is huge for me as it makes backlinking my sites so much simpler. No longer do I need 20 links in an article, now articles on my PBN can have 3 links to them, which will be backlinking what was the equivalent of 45 Micro Niche Sites before!
The only caveat is this only applies in certain niches, and keyword research needs to be extended a bit to find them. So how to find them? I still check APA but now add a 2nd filter, I check the # of pages indexed in Google for each of the sites in the SERP. Actually I just care about the site with the lowest # of pages. To do this I use SEOQuake, and I look for niches that have a site in their top 10 with under 30 pages (preferably under 20). Then I analyze those sites further. There are likely many cases where there were no “15+ page niche sites” in the SERPs before the update, leaving all the domains in the top 10 with lots of pages. I have a feeling that a 15 page niche site would likely work in any SERP with the generic articles/answers sites ranking (answers.com, Y! Answers, ezinearticles, ehow, low rank squidoo lenses and hubpages etc.). Don’t hold me to this though! I will be working on making my keyword tool also check and record pages indexed (though I’ll probably scrape bing or use seomoz instead because of annoying proxy issues with Google) and also check for occurrences of any of the giveaway generic sites. Once flagged I will set up a batch and report on how they do. One of the other many changes Google made to their algo involves understanding search intent and promoting certain types of results for certain types of queries. They have been promoting e-commerce sites for shopping queries, and travel review sites for hotels and destinations for ages now, but since the “site-clustering” update, I am seeing a lot more health sites and educational sites being returned now (which had traditionally been a MNS gold mine).
Update October 17: I have created 22 15 page sites. They are all in varying stages of development, but about half of them have all 15 articles published and link building began yesterday and today. For the other half I plan on adding an additional 15 pages of content (making them 30 page sites). I also plan on adding an additional 10 15 page sites using a slightly different content method, and another 10 30 page sites using this same method (more on that later). This last EMD update seemed like a lowering of the threshold of the classic OOP (Over Optimization Penalty) more than anything else. With the last EMD update I have changed several things:
Site Level: On a site-wide level, I am only creating 4-5 articles targeting a phrase match term that is uncompetitive yet profitable.. The other 10 pages of content are all based on LSI terms that are NOT phrase match and have high phrase match traffic in GATK along with a high Suggested Bid on the contextual tool. I don’t care about ranking for the exact match term, I hope for some long tail alone, but mainly they are here to support the 4-5 main terms. I am also interlinking my articles a bit within the site.
Example: If my main term is “blue widget” then I would have 4-5 secondary phrase match terms:
- Buy Blue Widgets
- Wholesale Blue Widgets
- Cheap Blue Widgets
- Blue Widget Accessories
Then I’d have 10 terms that are not phrase match, like “What are the best widgets for men”? Or “The history of Acme Inc” (I’m assuming Acme makes widgets??), “WebmasterWorld Widgets”, etc.
Page-Level On a page level, I’ve drastically reduced the number of mentions or my target term. I try to have it just twice. I’ve also made sure the title tag had more than just the keyword in it, and the keyword is not mentioned in the meta description.
Link Building I have a suspicion that phrase match anchors got hit (finally!!). That is to say anchor like “Blue Widgets from Blue Widgets.com”, “Cheap Blue Widgets Discount”, etc. are having a Penguin-like effect. I’m keeping my link building strictly to URL for now, and will add a few highly relevant brand anchors if that’s not enough to rank. As usual if you have a question please leave a comment! Now that the internships are over I should have some more time for the blog. Cheers, Hayden