Analysis: SEO Results of Article Submissions

I said in another thread that I would eat my hat if I didn't get a lot of links from article submissions. I think I may need some salt, but I let the community decide. It should be noted that these results are not indicative of the traffic that articles can generate. This is only an analysis of the SEO effects of article submissions.

The Setup
My Site (After 5 Web Design) is pretty new. It was completed about the last week of October. Google's December 19th Backlink export showed only 47 links for my site. To gain some popularity and traffic, I wrote a freeware program called Article Distributor that helps authors submit articles to a few hundred article directories. It was released on Dec 3rd. Because I offer it free, I include a short phrase at the bottom of the author's resource box. The phrase is "Submitted with Article Distributor", and "Article Distributor" is linked to the software's download page (http://www.after5webdesign.com/software.html). This accurately simulates the act of submitting hundreds of articles.

The one monkey wrench in the equation is that I also submitted a pad file for the software to a lot of shareware and freeware software directories. This has resulted in a few backlinks not from article submissions.

The Results
Using a multi-data center search tool like the one at SEO Logs (http://seologs.com/data-center-check.html), I have kept track of my unique phrase "Submitted with Article Distributor". The old data centers report 15,500 occurances at the time of this writing. The BigDaddy centers report 35,200 occurances. As previously stated, my index page only had 47 backlinks by Dec 19th, and the software page showed 0. With the export currently underway, some data centers are showing only 60 links to my index page (http://64.233.167.104/ | http://64.233.167.147/ | http://64.233.167.99/). These same data centers report 273 backlinks to the software.html page. A few more report just one backlink, and a majority are still showing 0. Since the export is still underway, this doesn't really surprise me. The toolbar is showing a PR of 3 for both pages. Both pages previously had a PR of 0 according to the toolbar.

Conclusions
Looking at the reported backlinks, I see several instances where the same article on two different sites shows up as a backlink. I also went from 0 PR to 3 PR. This pretty well proves that submitting the same article to lots of sites does not incur a duplicate content penalty for your site.

However, I am perplexed by the huge discrepancy between the number of occurances of my phrase and the number of back links reported. I know Google only reports a fraction of the number of backlinks they see, but 273 backlinks from 35,200 occurances is well below 1%. Even using the 15000 reported occurances from the non-bigdady centers is only 1.7%. Could this be due to link aging? Will I show an even larger number of backlinks after the next export?

So, submitting articles will help your site, but they don't exactly appear to be the most amazing way ever to generate backlinks as some claim. So, what's the verdict? Do I need a recipe book for how to cook hats?

 

I think it's still way to early to tell anything about PR.

 

I should also note that I'm a bit surprised by the fact that 273 reported backlinks only resulted in a PR of 3. The only explaination I can think of is that the actual PR passed by the article pages is tiny or these pages themselves have little to no PR assigned yet.

 

That's a good point. I may have to wait a week or two before I can get good, solid results.

 

Google does not report backlinks correctly anyhow so how are you chercking?

 

link:www.after5webdesign.com, and I know G doesn't report backlinks correctly, but even their flawed results should be indicative of the results one could expect to acheive by submitting articles for SEO purposes.

By comparison, MSN shows 783 backlinks for software.html and 984 for the site (link:www.after5webdesign.com). Yahoo reports 375 for the site (link:http://www.after5webdesign.com) and 2510 for link:http://www.after5webdesign.com/software.html

allinurl:http://www.after5webdesign.com/ reports:
G: 33 (bigdaddy centers don't appear to obey this command)
Y: 0
M: 0

 

And what sort of serp increases have you had for the particular keywords in the backlinks?

That would be the interesting part

Brad

 

The first article I ever wrote was picked up by some major sites and I got a ton of backlinks from it. Then I got lazy and started submitting 300 word generic articles that were crap because I thought it was easy. I didn’t get the same amount of exposure that I had previously. I guess my point is that just "writing articles" doesn’t mean you get traffic and backlinks. The articles have to be something that people actually want to read and will link to. If you browse through the majority of the stuff on GoArticles, Emazines, etc. you might find that most of the articles are useless crap. But thats just IMHO.

 

I'm starting to think google is getting better at assigning value to links that extends beyond the simple pagerank algorithm. For example, I'm finding large amount of links that occur only at the bottom of pages is not as valuable as when they appear at the top. Definitely still early though.

 

I'm ranked number one for Article Distributor. The other 9 results are all download sites where you can get Article Distributor. Article Distributor is a pretty narrow keyword though. Had I been smarter (and I think less ethical) I would have used the phrase "Submit Articles" or some such thing. I do not rank in the top 393 results Google displays when you just search for "Article".

Very true. Crap articles won't get you a link from anywhere but the article directories that accept it. A good article for your industry can get you hundreds of links from related sites though.

I haven't seen all of the articles that have been submitted with the software, but I'm confident it is well over 1500. Many are probably junk, but I have to believe that at least a couple are good, quality articles. Regardless, this should provide a good baseline of results. I wouldn't expect to see a huge number of reprint links showing up in the first few months anyhow.

I think you're very right David. My guess is that there are several things SE's can do to value links better. For example capping the count of links with anchor text "xyz" would discount the practice of buying sitewide links. Or they could devalue any links found with 5 to 10 lines of the closing body tag.

I guess my bottom line is this: If my results are positive for the distribution of articles to just article directories, a user can expect even better results when they submit a good quality article. If my results show that there is no SEO benefit to submitting articles, then we have to examine the traffic benefits more closely. If they don't generate traffic, then a whole segment of the internet is debunked. I don't want anecdotal evidence that article submissions work, I want good solid numbers.

 

Whoa! I just checked the rankings in MSN and Yahoo for article distributor, and was surprised to see that I do not rank 1st on MSN. Number 1 spot on MSN is the number 12 spot from Google. It is some web site with the URL ArticleDistributor.com. I rank second. (For the record, I was not aware of their site before I started this project. I chose Article Distributor because it was focused on sites using the Article Dashboard script. Get it? Dashboards and distributors are a part of a car. Plus, distributor seemed appropriate.)

It's worse with Yahoo. I rank 71st. Two download sites that are carrying Article Distributor rank in the top ten, but that's as close as I come. Yahoo certainly does like that Title tag though, huh?

 

Wow great report Tim.

I have been attempting this same type of experiment now for a client and I agree with a lot of what you have written as it shows in my results as well.

YES: Lots of allinanchor backlinks pointing to your domain.
YES: Dupe content does not seem to be triggered.
YES: MSN/Y seem to love it.
NO: Immediate benefits from this in G.

I think it really goes back to link age as the link age factor is what G used to combat the mass article submission and similiar tactics. G is not nearly as forgiving as G.

Having said that I still think article submission is a necessary strategy for any content developers - but it has to be a long term chore...And even with an automated submission tool it is a Chore with a capital C.

I will post back here when I have more data.

 

I obviously wrote this analysis too soon. My PR has been jumping around like crazy. PR0, PR3, PR5, etc. The "Current PR" tool at http://www.seologs.com/pagerank/current-pagerank/currentpr.html says my PR is 5, but I have yet to see that on the toolbar with any consistency. Most of the Google datacenters are showing 273 links for my software page now. The PR check from each data center (found at http://www.seologs.com/pr-check/pagerank-dc.html) shows 0's and 3's but no 5's.

One of the other things I've seen:
My links in MSN have completely stagnated around the 930-960 mark. This seems to imply that they cap the number of links counted from one domain to another. In otherwords, sitewides don't help much when it comes to links.

I'd love to see your data Miles. Maybe we can co-author an article and submit it!

 

I completely agree with your views Tim.
Google has really changed their algo a lot this time around. In February a website of my client (I just took over his project) was #33 for a keyword ( a lot of competition for it ), I did some on-page optimization and got a #25 on G. The worst part was that the webpage I was trying to optimize did not have any backlinks. I was about to start the working on the links part this week and all of a sudden the SERP's threw the website to a #51-52. (This was after the update). To my surprise the content optimization led to a rise in the PR of the two pages from 0 --> 7
I think this speaks a lot in itself. Now I am getting some one-way links and seo articles for the clients website. I am keeping my fingers crossed (getting links-links-links) and hoping for the SERP's to change soon

 

How long does it take for the backlinks generated by article distribution to show up on Google, MSN and Yahoo? Are they instant i.e. when the page is crawled by bots or can they take months.

 

That's good news wisdom. Congrats!

Qwest,
Define "Show up". If you mean "when will they appear using a link: command", possibly never with G and M because G only reports a sample of links they know about, and M appears to either cap the number of links that are counted from a particular domain or only report a sample of links like G. If you mean "when will they be counted", then the answer is "as soon as the page is spidered".

Keep in mind that the SEO benefits of submitting articles are not immediate. Articles are a great long term SEO strategy. You will see a boost in backlinks from the article directories the articles are published in, but it will take time for other websites to start picking these articles up and reprinting them on their site. This second phase of links can occur years after you originally published the article. You may have even forgotten about the article, but if a site finds it and reprints it, Ta Da! you got a new backlink.