Evidence of Page Level Google Penalties?

Jun 18, 2009 • 7:41 am | comments (8) by twitter Google+ | Filed Under Google Search Engine Optimization
 

Richard at SEO Gadget showed how Google seemed to have penalized specific pages of his site from ranking in the Google index. The penalty seemed to be fair, in that there were nasty comments that slipped through his comment spam filter.

The drop in traffic can be seen by the keyword phrases that page ranked well for. He noticed a ~70% drop in traffic for that phrase, which in his case resulted in a 15% drop in his Google traffic and a 5% drop in overall traffic.

What I find extra fun is that a Google Search Quality Analyst, @filiber, tweeted:

Google Page level penalty for comment spam – rankings and traffic drop http://bit.ly/JNAly (via @AndyBeard) <- interesting read!

Of course that is not admission to this as a fact, but it wouldn't be too hard to believe that bad comments caused such a decline.

Now, I don't think this would be considered a keyword-specific penalty, which most SEOs believe in, but rather a specific page being penalized.

Forum discussion at Sphinn.

Previous story: Google AdSense Gives US Font Size Control
 

Comments:

C

06/18/2009 01:24 pm

We have seen a site keep its primary keyword rankings, whilst every single subpage that ranked has been demoted to between 150 - 700 positions. This happened for the first time last month, and has now happened again.

Michael Martinez

06/18/2009 05:17 pm

It's not necessarily keyword-related. It's page-related. I've only seen this happen on pages in competitive queries, where Google knows a lot of eyeballs are looking at the search results. I'm surprised that it kicked in the SafeSearch filter but in retrospect that does make some sort of sense.

Chris

06/18/2009 09:42 pm

Hey all. I agree in part with some points here but would like to throw my white hat into the ring. I have an SEO client who was penalized for 4 specific terms because of spammy blog linking. They hired us to unravel the problem. The evidence I've seen supports it being keyword specific and not page specific. To test, try to optimize another page for your penalized keyword and see if you get your ranking back. In my (unfortunate) experience it didn't work which makes me believe it is keyword and not just the page. (In this case the penalized keywords also pointed to the home page, tsk tsk, but the home page still ranks well for other terms...) I'm all ears if you have evidence to the contrary and would LOVE to hear that I'm wrong but in this test case, removing the keyword from titles and metas and moving to another page that has good rankings didn't work and their site still doesn't rank at all for the "penalized" terms... Anyone know a good blog-eraser? :) -CWalk

Chris

06/18/2009 09:45 pm

Forgot to mention - the "move" I made was closely considered and a relevant page was chosen after a site search in the Googz. Content was there and the page ranked well for closely associated keywords... It wasn't just an arbitrary switch. Also, I didn't expect rankings to just snap to and be great all of a sudden. Just seeing them in the top 100 again would have been enough of a result. Still nada.

Michael Martinez

06/18/2009 10:45 pm

Keyword-specific penalties have been reported and discussed through the years. What is being suggested above is that a page-level penalty has been incurred. Some people have noticed this effect before when discussing the so-called "-950 penalty" (Googlers has confirmed such a penalty exists but it's really a group of penalties, the severity of which depends on undisclosed factors). Some people call this the "Overoptimization penalty" because, they claim, they have seen their search results improve after unoptimizing the pages (which, in effect, means the pages weren't optimized at all). The connection to keywords, in my opinion, seems to be that the more active a query, the more likely these penalties will kick in. That way Google would not have to maintain a table of protected queries -- they could instead merely set up some criteria for triggering a new set of filters and then like magic "overoptimized" pages would start to drop in the search results.

Chris

06/19/2009 12:24 am

I appreciate the feedback Michael! I don't want to disagree but the evidence I've seen doesn't seem to be page-specific. (Maybe I'm wrong - I have been more than a few times :) Their home page ranked very well and got a ton of traffic from a specific keyword, then after the penalty was applied the traffic (and top 5 position) completely vanished. Not just a drop, it jumped off a cliff. BUT, the home page still ranks well for other terms. So we created an interior page named www.xyz.com/penalized-keyword.html, optimized the heck out of it knowing that they used to rank so well for the term, created great internal linking and content (top of left-nav), submitted a reconsideration request and have yet to see even a blip in top 100 positions for the penalized keyword... If you search for modifier+penalized+keyword guess who is #2 in Google. Penalized home page! Now I'm depressed.

Andy Beard

06/19/2009 10:08 am

I have mentioned it in a few tweets over the last 6-8 months that I have seen quite suggestive data that Google can apply penalties to individual pages or sections of a site due to paid links, preventing them passing PageRank or possibly other ranking factors. It affected TBPR sitewide It did not affect overall traffic Rankings shifted from one type of page which matched the search query to another on the same domain. The pages that lost traffic primarily gained juice from offending pages. When paid links were removed, in under a week, ranking for the original pages reverted to normal levels. Google is doing all kinds of fun things, interesting on the retweet, I saw it but the correlation didn't trigger in my head.

Kevin Lam

06/20/2009 03:54 pm

So it's a penalty for excessive display of specific keywords? Why don't Google just filter it like they do with duplicate content? In this case, it's duplicate keywords once it reaches a certain number of repeats or a percentage of a particular page.

blog comments powered by Disqus