Penguin Experiment - Changing Anchors on blog network

Finally I'm making some traction figuring out this Penguin shit.

My clients took such a hard hit, I've been unable to do hardly ANY of my own affiliate marketing, and it's pissing me off...

So I decided to take matters into my own hands and used Keyword Canine and dissect each of my clients backlink profiles, looking for patterns.

Pattern #1:
Just as the MicroSite Masters blog post mentioned, clients with 2 full URL variations in their top 5 anchors were not affected as badly.

Pattern #2:
Clients with a close ratio of D0FOLL0W to NOFOLLOW were not affected as bad as clients with majority D0FOLL0W links.

Pattern #3:
Clients with indexed pages on Angie's List and BBB appear to be near invincible. I have 1 client with only ~20% Full URL anchors, but his Angie and BBB page have the Full URL, PR, and are indexed. Oddly enough, he's the ONLY one that still ranks #1, and all the competitors were destroyed by Penguin.

Experiment:
My first test was to change Anchor links on my blog network to Full URL instead of keywords. I started with PR2-3 blogs, being very careful with the anchors. I used Keyword Canine to give me an idea of which "style" of Full URL variation I should be using.

Results:
In 24 hours, with NO ADDITIONAL links built, here is the graph from one client:

panel.micrositemasters.com/reports/8ead4ee28b085aa8b6612503ano3s9s.html

O_O....? Duh...? Penguin is running a background filter just like Panda, if that wasn't obvious already.

I have now changed anchors for several clients, and in addition launched a 30 day UD Wiki blast to account for the lack of NOFOLLOW links and anchor diversity.

I will post updated results here over the next few weeks.

Great info. Raw anchors is indeed the silver bullet that is working.

By changing existing anchor text to raw urls, on your blog network, it's a double whammy to Google. Your backlink profile falls into line with their new "standard" much faster than those that are building new links to overcome those that can't be modified.

Keep up the great work and sharing your updates with us!

I think I found another pattern...

This makes total sense if you think of how Google wants to kill High PR Homepage networks.

When Keyword Canine indexes a link from a HPRHP site, it pulls the data from the Homepage because it has more authority.

Problem: The Homepage is NOT relevant at all, as it has many posts and a random Title tag etc.
(Unless you use HPR domains with relevant content only, time consuming and expensive, but does work.)

I think this is why Wikis are working so well. On EVERY Wiki link that KW Canine pulls, the specific page has a relevant title.

I ran a test on the same keywords... 1 BH site (iWriter) vs 1 WH (hand written) site. The WH site randomly received more blog network links, and is hardly ranking anywhere on the SERPs. Majority of links are Full URL, but it has too many Blog network posts on the homepage.

The BH site has majority Wiki links, only a few Generic HPRHP links, and both sites have a few *Relevant HPRHP links (aged domains I used specifically for this niche).

BH site is kicking the WH site's ass...

If the algorithm is counting the link from the "authority" page, in this case the High PR homepage, it would make sense it sees them as "irrelevant" links.

So in that case... I wonder if it's MORE effective to let .. ermm... some* posts roll off the homepage, as Google will then place the authority on their dedicated page.
(Of course leave some on the homepage if for nothing more than increasing PR.)

Any opinions on this?

Thx

Someone said in a thread once that URL anchors still pass some form of link juice. I think the quote was "Google knows what your site is about, all links pass juice"
Any truth to this?

I think you may be confusing what exactly "Link Juice" is.

Any link from a High PR page is going to pass juice to your site.

Yes I do believe the algorithm is becoming smart enough to rank pages without anchor text backlinks.

The BH site I mentioned in my above example has NO Exact keyword anchors at all (but does have a few LSI anchors), and it destroying the WH site that has exact anchors.

Proof is in the pudding on that experiment... But I still need to run more tests, this was only in one market.

So in that case... I wonder if it's MORE effective to let .. ermm... some* posts roll off the homepage, as Google will then place the authority on their dedicated page.
(Of course leave some on the homepage if for nothing more than increasing PR.)

Any opinions on this?

Click to expand...

The way many blog networks were/are constructed is that they display the entire post on the homepage. Some blogs set to 10 posts, 20 posts, etc. The problem is that these are all duplicated on category and individual pages. While the homepage may have the most authority, the text of a paid blog post is diluted with other irrelevant text (including the blogs own meta data). It's not healthy for the blog and the posts that appear on the homepage. Using the "more" option for blog posts is essential. Looking at a couple nuked blog networks, BMR had the "more" option and HPRS did not. Although HPRS had higher pagerank blogs, the posts were far less effective than BMR when constructed with the "more" option.

Although blog posts do roll off of the homepage, many blog networks categorize posts. Once again, if the "more" option is not used then duplicated content is present at the category level. Generally speaking the categories are linked to from the homepage, so that would carry far more weight then individual pages. If the individual blog post page is duplicated text from the category page, the category may win out in the battle of relevance leaving ones blog post smack dab in the middle of other unrelated posts.

Depending on how your blog network is constructed, it will make a big difference in the relevance of the posts/links that appear within it.

My thoughts exactly! I had actually been using the More tag already, in an attempt to pass more relevance to the dedicated page.

Very interesting, I'm glad I noticed that on KW Canine. Gives me a whole new way to think about how G treats blog network backlinks.

Thanks for the reply