| 9:20 pm on Jul 10, 2007 (gmt 0)|
I had a similar situation and I did remove the category descriptions from page 1, 2, 3 and etc.. because I wanted the copy to be unique only to the main category page not the numbered pages.
I still made sure the user knew that they were still part of that category but I did remove the description. The next pages will always have unique content so why add the description from the main page there.
| 10:34 pm on Jul 10, 2007 (gmt 0)|
I find it appalling that we are forced to sacrifice user experience in order to satisfy search engines' ambiguous (and, to the best of my knowledge, unpublished) definitions of "duplicate content".
How about petitioning for them to either give us precise definitions or provide "how-to-do-it-without-incurring-the-wrath" examples for common situations like photo galleries where one photo can end up in several of them, multiple post categories in blogs, etc?
| 11:47 pm on Jul 10, 2007 (gmt 0)|
I agree. This isn't the first time I've had to weight my users' satisfaction against Google's satisfaction.
| 12:34 am on Jul 11, 2007 (gmt 0)|
Ask yourself how a SERP with 50 identical titles in it guides the user as to which one to click on... and you'll know what to do.
| 1:03 am on Jul 11, 2007 (gmt 0)|
If I were Google I'd pick the one with the most internal links pointing to it. Are you saying that the duplicate descriptions are OK because of that, or are you saying I should remove the duplicate descriptions?
| 1:14 am on Jul 11, 2007 (gmt 0)|
How are they supposed to attach semantic meaning to the following:
<p>This is a description of my category</p>
<h2 class="category">My Category</h2>
<p class="description">This is a description of my category</p>
<h2 class="cat">My Category</h2>
<p class="description">This is a description of my category</p>
The first one is a h2/p tag combination. Sure you know its a category and description, but how are they supposed to know this or any h2/p combination for that matter?
In the second case, do you really believe that Google can perfectly derive semantic meaning based on class/id alone?
In the third case, even Google did make some decision about id/class meaning, how do you know they won't analyze it to be a description of a cat?
So can you blame them? Its not their fault. Its not yours. Its the markup language and agreed upon understanding of what you are trying to express is what is wrong. HTML isn't it.
| 2:26 am on Jul 11, 2007 (gmt 0)|
I don't blame them, it's just frustrating from a whole-business perspective.
Do you recommend removing the descriptions from the pages beyond #1 then?
| 4:56 am on Jul 11, 2007 (gmt 0)|
To give G due they have updated their Webmaster Guidelines recently and the content has hugely increased regarding duplicate and boiler plate.
Check out the new guidelines, they might help you in your approach to content.
| 12:12 pm on Jul 11, 2007 (gmt 0)|
Pagination is OK if every page is unique. Check for: meta titles , descriptions , content . If these are OK you should be sailing, provided you have no other issues.
If you cannot alter this [ which would defeat the purpose of pagination ] , then G will likely pick the strongest pages aka PR , IBL links .
[edited by: Whitey at 12:13 pm (utc) on July 11, 2007]
| 1:06 pm on Jul 11, 2007 (gmt 0)|
|I find it appalling that we are forced to sacrifice user experience in order to satisfy search engines' ambiguous (and, to the best of my knowledge, unpublished) definitions of "duplicate content". |
You're not, so don't get riled!
The more content a page has on it that is reproduced on other pages in that site or others, and the less original content it has on it, the less important it will seem to Google to feature that page or feature it prominently. With limited space, they have to try to determine 'quality' algorithmically. As g1msd said, think about it from their perspective.
If it's one paragraph and you have tons of unique content underneath then why worry?
And who said you can't have it 'on the page' as far as humans go?
| 2:14 pm on Jul 11, 2007 (gmt 0)|
Ok, it sounds like a unique title, meta description, h1, and main <p> description are what I should do to be safe. Thanks everyone.
| 2:18 pm on Jul 11, 2007 (gmt 0)|
..... and content
| 2:07 am on Jul 13, 2007 (gmt 0)|
I've removed the duplicated <p> descriptions and meta descriptions for all pages greater than #1.
The more I think about duplicate content though, the more I'm not sure where to draw the line. My logo, navigation bar, and menu structure all appear on every page. Is that hurting me? The same type of question can be asked of many of the elements on all category pages, all product pages, etc.
| 2:27 am on Jul 13, 2007 (gmt 0)|
No, those universally repeated elements are not hurting you. All the search engines, including Google, recognize the commonly recurring sections of a template and do not consider them as content.
Here's where good semantic mark-up can be an excellent help to a site's rankings, IMO. If each "section" of a page is marked up within a container element of some kind, that makes this algorithmic job of finding the content section much easier -- and consequently the site is less prone to taking on collateral damage.
It's duplicated content within three major spots -- the content section, title element, and meta description -- that can cause trouble. By definition, these elements are "supposed to be" both unique and specific to the url where they appear.
| 4:53 pm on Jul 13, 2007 (gmt 0)|
What you're saying makes sense, but have you seen fairly concrete evidence of it?
One of the main goals for my HTML and CSS has been to have as little of it on the page as possible. Creating extra container divs would be going in the other direction, but maybe it would be helpful in this case.
| 5:32 pm on Jul 13, 2007 (gmt 0)|
Yes, I have seen very suggestive data about search engines (not just Google) working to isolate the various blocks of the template and score them differently.
You may have read the often repeated opinion that a backlink from within the content section is weighted more heavily than other links. This is one widely discussed bit of evidence, that somehow the algorithm is also looking for WHERE the link appears on the page, and not just THAT it appears. Also the recently discussed issues with "footer links" fall into this same territory.
| 7:45 pm on Jul 13, 2007 (gmt 0)|
About the only Google patent that I ever tried to wade through (some years ago) was one that talked about "method to identify a list, or paragraph, of internal links as being in-site navigation" and talked about "method to identify a common paragraph of text, one appearing on most pages of the site, as being the footer text" and so on. So, this technology is available, and assumed to be already in use for quite some time.
| 11:14 pm on Jul 13, 2007 (gmt 0)|
I believe it's in use, I just wanted to make sure it's based on container divs before I start nesting everything.
| 11:20 pm on Jul 13, 2007 (gmt 0)|
As I see it, container divs just make page sections easier to detect - but the actual algo seems to be more robust than that.