homepage Welcome to WebmasterWorld Guest from 50.19.206.49
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / Google / Google SEO News and Discussion
Forum Library, Charter, Moderators: Robert Charlton & aakk9999 & brotherhood of lan & goodroi

Google SEO News and Discussion Forum

This 36 message thread spans 2 pages: 36 ( [1] 2 > >     
Over Optimization Penalty? Wrong Page Ranking For Keyword
Planet13




msg:4291497
 4:04 pm on Apr 3, 2011 (gmt 0)

Hi there, Everyone:

I wonder if one of my pages might be have an over optimization penalty, and hope you might have ideas on how to determine if I do, and what the best method of correcting it might be.

The Situation:

An Example: I am trying to rank for the keywords "widget photos"

my domain name is: thewidgetpanorama . com/ and it is about 16 months old.

I want my page /widget-photos.html to rank for the "widget photos" keywords. However, it does NOT appear in the first 300 search results in google

my page /widget-images.html currently ranks around 59th in the serps

My widget-photos.html page is linked to directly from the home page. The widget-images.html page is NOT. In fact, it is only linked to from the widget-photos.html page.

The title text of my widget-photos.html page is: Widget Photos and Images: Reproductions of Widget Pictures

The title text of my widget-images.html page is: Widget Images

my widget-photos.html contains not only about a dozen products, but a pretty extensive unique history of widget photography

The Plot Thickens...

While this site is 16 months old, it was "in competition" with another site of mine, which had similar material (but different enough to hopefully not get a duplicate content penalty).

On February 18th, I took all the widget images material off the old site and 301 redirected it to the most relevant pages on the new site.

The home page of the OLD site (which DID have the words widget photos in the title) ranks 48th for the keyword widget photos (it was ranking around 16th with an indented listing to its widget-photos.html page back in mid February when I moved the content to the new site, and back in the day, it ranked in the top 5 many years ago).

The old site has a link on the home page in the first paragraph of text to the widget-photos.html page of the new site. The anchor text is widget photos

the widget-photos.html page of the old site has been 301 redirected to the widget-photos.html page of the new site.

And Google Says...

Google analytics tells me that the widget-photos.html page of the new site is the most popular page on my site, but the keywords people are using to find it are things like "widget photos history" or "widget photos techniques" " or "meaning of widget photos" - basically, NOT money making keywords.

Immediately after 301 redirecting the old widget-photos.html page to the new site's widget-photos.html page, the new site's widget-photos.html page's traffic quadrupled.

Within three days though, it was down to only being double the traffic it had previously got. (I guess that it took google about three days to "index" the 301 redirect, which is when the traffic dropped from about quadruple to just double.)

Link Profile:

I don't have many links. Like maybe 20 (mostly reciprocals) to the home page - none to the widget-photos.html page, and none to the widget-images.html page. I do have about 25 to a different section of my site that is unrelated to widget photography (so different, in fact, that I will probably be breaking it off to another site in the next 6 months or so).

Any suggestions on how to analyze this or what direction to move forward would be greatly appreciated.

Thanks for reading and thanks in advance for any suggestions.

 

Planet13




msg:4291536
 6:21 pm on Apr 3, 2011 (gmt 0)

Ok, this was a long post already, but just to add to it...

the page widget-photos.html seems to be dropping in the SERPs for a few other phrases over the last two weeks as well. Although one drop was only down 5 spaces, so it doesn't look like a penalty.

aristotle




msg:4291668
 1:16 am on Apr 4, 2011 (gmt 0)

Immediately after 301 redirecting the old widget-photos.html page to the new site's widget-photos.html page, the new site's widget-photos.html page's traffic quadrupled.

Within three days though, it was down to only being double the traffic it had previously got. (I guess that it took google about three days to "index" the 301 redirect, which is when the traffic dropped from about quadruple to just double.)



I'm having trouble interpreting this part. Do you have any data about the traffic from different sources during this period?

Planet13




msg:4291686
 2:55 am on Apr 4, 2011 (gmt 0)

Hi there, Aristotle:

First, let me give you some general traffic data

Immediately after the redirect, From Feb 19, 2011 - Feb 20, 2011, traffic was up 291.43% to the widget-photos.html page (I guess that is only triple, not quadruple), and the percentage of site traffic was 7.87% (before that, the percentage of traffic for that page was only 3.52%)

When i look at that page for entrance sources for the first two days after the redirect, compared to the two days prior to the redirect, it looks really weird. It says that google traffic was up 3,883.33% - but for the period prior, it shows several sources (yahoo, (direct), ask, and google.com (which I guess is different from regular google) as having 0 traffic before the 301 redirect (and then a handful of entrances after the 301 redirect).

But when i look at Feb 21 to 22nd (compared to Feb 19 to 20th, immediately after the redirct), it is kind of the opposite:

google -79.92%

yahoo +100%

my-original-site.com/ +100%

ask +25%

bing +80%

search -33%

google.com -83%

Oddly enough, it says that the number of entrances to widget-photos.html using the keyword widget photos doubled during that time.

Anyway, it is weird because I seem to get pretty good traffic for three-word keywords that include widget photos in the keyword phrase, but I don't seem to rank well for just widget photos (I haven't checked my rankings for those three-word keywords, but I do get traffic from them).

I'm kind of at a loss...

tedster




msg:4291689
 3:13 am on Apr 4, 2011 (gmt 0)

I have fought similar battles over past years where Google just didn't "want to" rank what I considered the best page for the query. In almost every case, things only stabilized when other sites began to reinforce the "best page" with their natural backlinks.

For some 2-word queries, this was an area where I never found a way to direct Google to the best page using on-site methods. And having the phrase in the file path never seemed to do the trick either.

Planet13




msg:4291717
 5:51 am on Apr 4, 2011 (gmt 0)

@tedster:

Thanks for the input. I looked at some of your previous advice to others about some ranking problems and just decided to pull some of the 301 redirects from the older site to the new widget-photos.html page.

I realized that I had several products that were discontinued and, because they basically were in the widget photos CATEGORY, I had decided to 301 those individual products to the widget-photos.html page.

I figured I had already given enough signals to google what the widget-photos.html page is about by using keywords in the title, h1 tags, the one-to-one redirect from the old site, and internal anchor text to the page, that I DON'T need to have a bunch of 301 redirects from "similar but not exactly the same" pages. That might just confuse google.

Also tedster, if memory serves, I remember you had mentioned that you had one instance where you had changed the anchor text on a link that was on the page that you wanted to rank that was pointing to another change on your site, and that had helped to push it over the hill. Does that sound familiar? And am I thinking of the right context?

tedster




msg:4291722
 6:18 am on Apr 4, 2011 (gmt 0)

Good memory - I wasn't thinking of that case. Yes, I actually created new anchor text as an ON PAGE element, but I did not use the target query terms. Instead I used a "co-occurring" phrase. Think about a page that I want to rank for "cider press" and I added the anchor text "apple orchard" in a new link on that very page.

I remember now what a last ditch effort that was - and how astounded I felt when the page jumped from #28 to #4 within days of that change. I still can't say for certain that the new link made all the difference, because we were also reaching out very intesively for backlinks at the time. But the timing sure seemed to say "Aha, now we know what you're page is about."

Planet13




msg:4291925
 6:21 pm on Apr 4, 2011 (gmt 0)

@tedster

Instead I used a "co-occurring" phrase. Think about a page that I want to rank for "cider press" and I added the anchor text "apple orchard" in a new link on that very page.


Do you remember how you decided on the particular phrase? Was it something you spend some time analyzing? Or was it more of a gut feeling that led you to "apple orchard"?

And did the page you lined to already exist? Or was it created especially for this?

...because we were also reaching out very intesively for backlinks at the time.


Do you remember if they were targeted to that page, or the home page of the site? And were you targeting specific keywords in the anchor text? Or not so much...

thanks in advance.

aristotle




msg:4291945
 7:13 pm on Apr 4, 2011 (gmt 0)

I wonder if one of my pages might be have an over optimization penalty


From what you've described, it very well could be an over-optization penalty, although there's no way to know for sure. I suggest that you GRADUALLY try to "de-optimize" it -- One step would be to Replace keywords with synonyms in both content and anchor text. Removing some of the redirects was probably also a good step. Also, in addition to de-optimizing, you could also try to get some backlinks directly to the page.

I think Tedster was talking about de-optimizing an outlink. The same principal would apply - use synonyms and substitutes instead of keywords in the anchor.

In SUM -- Add variety but do it gradually.

tedster




msg:4291954
 7:36 pm on Apr 4, 2011 (gmt 0)

I think Tedster was talking about de-optimizing an outlink.

Not exactly - this was a newly added outlink but it pointed to an already existing page. I was intentionally working to enhance the semantic content of the page by using co-occurring phrases. The original content had been written with too much of a single keyword focus - very old school SEO.

First I did a text analysis of the top 20 ranking pages to find common co-occurring phrases. I chose just a few that were a good fit to the page whose content I was working with. I added 2 phrases as pure content and two others as anchor text. The pages they both pointed to already existed on the site but they had not been linked to from this particular page.

This content change also enhanced the landing page for the visitor (by a lot). Onc I did the analysis, I was almost embarrassed by the "puerly SEO" nature of the previous content. But the desired ranking result also happened within 48 hours of the change - this landing page jumped from page 3 to #4 for its intended query phrase.

The landing page itself was getting new backlinks at the time, so it's not a very controlled test. But the jump in ranking came when I added the new phrases, after months of struggle with other attempts. The page had already been collecting new backlinks and the needle was still stuck - then it jumped within 48 hours of adding those phrases.

This is a very touchy technique. I know of people who have completely tanked their rankings by adding a lot of co-occurring phrases at once. The phrase-based indexing patents [webmasterworld.com] include some heavy duty spam prevention components.

tedster




msg:4291968
 7:50 pm on Apr 4, 2011 (gmt 0)

I just want to add that these new words were NOT synonyms, they were co-occurring phrases [webmasterworld.com]. That is a significant difference.

Going over the threshold for co-occurring phrases can get you smacked even worse than garden variety keyword stuffing.

Planet13




msg:4291973
 7:59 pm on Apr 4, 2011 (gmt 0)

Thanks for the input, guys.

First I did a text analysis of the top 20 ranking pages to find common co-occurring phrases.


Is there a tool available that might help one accomplish this?

The page I am trying to rank ( widget-photos.html ) might have TOO much content in it, since not only to I link to the main products on this page, but I have links to the subcategories (like Black and White Widget Photos) and I have a few paragraphs of text about the history, a few paragraphs about the materials used, and a few paragraphs about the techniques used in making widget photos.

Maybe it would be good to move some of that content to new pages and link to it ONLY from the widget-photos.html landing page? They might end up being a bit thin (although i have certainly seen thinner pages out there).

Otherwise, I don't know how to make the widget-photos.html page seem less keyword stuffed. I have tried to use synonyms, but there is about 800 words of text on the page, so the phrase "widget photos" shows up a lot - especially since it appears in the links to ALL of the 11 products on that page and ALL of the five subcategories on that page.

One other strange thing: The home page of the site doesn't rank for the widget photos keyword, despite having that keyword as the first phrase in the title of the home page. I wonder if that has something to do with it? (On my old site, the home page DID rank for that phrase with an indented result).

aristotle




msg:4291994
 8:38 pm on Apr 4, 2011 (gmt 0)

If you write a thorough article, most of the normal co-occurring phrases for the subject will appear in it naturally anyway. I think the Google algorithm mainly uses them to evaluate credibility. But to really stand out from the crowd, an article needs to have some additional unique content in addition to the normal co-occurring phrases.

aristotle




msg:4292013
 9:09 pm on Apr 4, 2011 (gmt 0)

Planet13 - Google Search has a tool called "Wonder Wheel" that you can use to find co-occurring phrases.

Planet13




msg:4292026
 9:18 pm on Apr 4, 2011 (gmt 0)

Thank You, Aristotle!

I remember hearing about that in a video. I will look into it.

aristotle




msg:4292035
 9:25 pm on Apr 4, 2011 (gmt 0)

Well, I just took a look at the Wonder Wheel myself, and it doesn't seem to work as well as I remembered. It's been at least a year since I looked at it. So it's probably not as useful as I thought.

indyank




msg:4292168
 3:02 am on Apr 5, 2011 (gmt 0)

Yes the patent that Tedster had linked to in another thread seem to have been designed to counter this and one should be very careful with this.I guess it is best to have them included naturally by writing a thought article as Aristotle suggests rather than using tools to discover them and include them.

Tedster, does the page that you linked to had that keyword in title or content body or anywhere else("apple orchard" in your example)?

I guess that it is also important that this page on "apple orchard" is an existing one and google knows about its existence in its index.Is that true?

tedster




msg:4292170
 3:15 am on Apr 5, 2011 (gmt 0)

The page I linked to used the co-occurring phrase prominently in the content area, but not in the title element.

Back in 2004-2007 I did a lot of testing for how to use co-occurring phrases. One time, we tried adding every Related Search phrase that Google returned for our target query phrase. The result was a six month minus 950 penalty. Several other dramatic experiences taught me that probing in the area of phrase-based indexing is like working with unstable nitroglycerin.

From that point on I've added co-occurring terms to indexed content very gradually when I do it at all. Today I mostly add co-occurring vocabulary before publication, usually when proposed content reads too much like it was written by an SEO. It's very rare that I play around with already indexed content.

indyank




msg:4292177
 3:53 am on Apr 5, 2011 (gmt 0)

I guess that it is also important that this page on "apple orchard" is an existing one and google knows about its existence in its index.Is that true?


Tedster, I meant the page that is "linked to" with the co-occurring phrase when I asked the above question and not the page that we work on. sorry if that wasn't clear.

tedster




msg:4292179
 3:57 am on Apr 5, 2011 (gmt 0)

It was clear to me - and yes, the page I linked to already existed. I'm not sure if that was a critical factor or not. My guess would be that it isn't important, but it helps.

suzukik




msg:4292615
 6:57 pm on Apr 5, 2011 (gmt 0)

First I did a text analysis of the top 20 ranking pages to find common co-occurring phrases.


@tedster
Did you do the analysis manually or use some tools?

I have another question:
did you find this method working at other pages/sites than you showed us as the example?

tedster




msg:4292647
 7:35 pm on Apr 5, 2011 (gmt 0)

I used to use a tool - one that is no longer available because too many websites were not understanding it and they were killing their rankings. Now I just collect the text and then eyeball it. The tool used to spit out some unusable ideas anyway, and human judgment is essential.

aristotle




msg:4292657
 7:56 pm on Apr 5, 2011 (gmt 0)

To find co-occurring terms, read some good articles about the subject. The terms will naturally appear in the content of the articles.

Planet13




msg:4292756
 10:33 pm on Apr 5, 2011 (gmt 0)

I am tempted to look for a tool, too. But after reading wheel's most recent thread, maybe I will just link to pages that would be of most value to the reader.

aristotle




msg:4292809
 1:31 am on Apr 6, 2011 (gmt 0)

I'm puzzled about why anyone is looking for co-occurring phrases. If you're writing an article, you should already know what they are. If you're not writing an article, then what do you intend to do with them?

tedster




msg:4292825
 2:21 am on Apr 6, 2011 (gmt 0)

I used to make half my income writing. Sometimes you write and write - and have blinders on. This is especially true for web writing from the old days, when you needed to shout at search engines to get your relevance across to their first generation, nearly brain dead text-match algorithms.

I've found that a co-occurring terms check can often breathe fresh life into dead copy, and also awaken the writer to aspects of the topic that the audience is interested in but that somehow slipped past. There are also times where it helps your copy to take what is implicitly understood and state it explicitly.

Using naturally extended vocabulary is another way to avoid the so-called "Over Optimization Penalty" while it also wakes up the page.

Planet13




msg:4292862
 4:46 am on Apr 6, 2011 (gmt 0)

Using naturally extended vocabulary is another way to avoid the so-called "Over Optimization Penalty" while it also wakes up the page.


I hate to admit that I don't know what "naturally extended vocabulary" means exactly. My first thought is that it means synonyms. Am I headed in the right direction?

Planet13




msg:4292864
 4:54 am on Apr 6, 2011 (gmt 0)

@aristotle


I'm puzzled about why anyone is looking for co-occurring phrases. If you're writing an article, you should already know what they are. If you're not writing an article, then what do you intend to do with them?


Well, that is a good question...

My writing instinct is to not look at them as co-occurring phrases, but to think of them as just related topics.

However, I have been writing for the search engines more and more, and this one page that is not ranking at all for the keywords is really bugging me.

So I am just looking for the one phrase or two that will bring things up - much the way that tedster had mentioned he had done that before.

tedster




msg:4292875
 5:52 am on Apr 6, 2011 (gmt 0)

I don't exactly mean synonyms. I mean instead of writing content with nearly a lockstep on the target keywords and their stemmed variations, use a wider vocabulary that fit naturally with the topic. You want the page to rank for "doctor's office visit" and you write about all kinds of things related to that event - "prescription pads", "blood pressure tests", "pharmaceutical samples", nurse assistant" and so on.

Take the tight reins off the copy and use more words that naturally co-occur with the main topic. Maybe your content already does this - but clearly there is a lot of web content today that does not.

Planet13




msg:4292879
 5:59 am on Apr 6, 2011 (gmt 0)

Getting back to the original subject of Over Optimization penalty for a moment:

While the page is not ranking for the two-word keyword phrase that I am after (namely, widget photos), it does rank number one for unique phrases that are four or more words in length.

for example: while it doesn't rank AT ALL for "widget photos" it does rank #1 for "Each widget photos does" . It also ranks number three or four for other three-word keywords that contain widget photos in it (I am out ranked by ehow at #1 and two listings by amazon; a product listing and an askville page from amazon).

So if there is a penalty, it has to be specifically for the two word phrase "widget photos" only, but not for any phrase with other words prepended or appended.

Hmmm... it seems a few sites are using a couple of paragraphs of text from this page - one of them being an amazon listmania site. How much text is fair use?

This 36 message thread spans 2 pages: 36 ( [1] 2 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google SEO News and Discussion
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved