I solved (at least partly) my problem maintaining rankings over time. Old pages that used to get traffic are once again seeing inbound search visits. I did it by tweaking my blog’s category pages, but I think the lessons are also helpful for ecommerce sites.
By default, Wordpress category pages show 10 posts. Similarly, most ecommerce category pages show 9 products per page (in a 3×3 grid).
This sucks for SEO, because your individual blog posts or category products get less and less internal linklove over time as they get pushed to the secondary category pages.
There are two types of secondary category pages – the extremely deep ones that no one will ever get to because it takes 10 “next page” clicks to get there. For this problem, Egghead created Wordpress pagination. Adding “1,” “2,” “3,” … links is just a partial solution, though. And as Rand highlights in this fantastic Whiteboard Friday (prompted by a question of mine, I believe 
), pagination is still not great.
Search engines don’t have a particular reason to crawl deeper than the main category page, as a general rule.
This is more true for ecommerce sites where the products tend to resemble each other (e.g. diamond rings) and a deeper crawl is unlikely to make the results returned to searchers significantly better. Search engines also have that incentive in a Wordpress blog context since a percentage of the newer posts likely rehash older ones, though the rehash is less pronounced relative to ecommerce sites.
Further compounding the problem is the fact, as I just recently found out, that All In One SEO – probably the single most popular Wordpress SEO plugin around – adds a meta robots noindex tag to your category pages! (Or perhaps it writes to robots.txt; I didn’t pay exact attention but that’s a technical detail that is beside the point.) If I were a search engineer, I would treat links from noindexed pages as less valuable than those from pages meant to be in my index. No, I haven’t run this test scientifically – this part is just a hunch.
The Takeaway Tips on Optimizing Archival Material:
1) Go to Settings (aka Options in older versions of Wordpress) – All In One SEO – and remove the checkmark next to ‘noindex category pages’.
2) Go to Settings – Reading. Set “Blog pages show at most: ” to 100. Rand suggests (in that Whiteboard Friday) that this won’t solve all your problems, because it doesn’t scale. So here’s how to scale it…
3) Ironically, the answer is in another SEOmoz post. Consolidate the repetitive long tail post selection by moving all your content on a narrow subject to one particular post. This will reduce the number of posts (and hence the depth to which spiders need to crawl) as well as make for a single, more comprehensive post.
That makes it more linkable, incidentally. And it should help bring in more longtail traffic too, due to the mix of previously absent words with those already present! While we’re getting all Wikipedia style, why not throw in a table of contents for good measure, since your post is going to be getting longer?
4) Take Michael Gray’s tip on putting posts in just one category (see the vid below). This will enable you to maximize the value of each category page in getting posts indexed (or each subcategory’s power, as Rand cleverly advises in the Whiteboard Friday video).
For example, I recently wrote that my idea for measuring social media results could be coded using the Friendfeed API. Well, the Friendfeed API blurb is going to be integrated into the main post as soon as I find time. Similarly, a number of my link buying posts can be simplified and consolidated.
The Ideal Category Archive SEO Solution:
- Ajax – 100 posts can be intimidating, especially when that scroll button gets all tiny. Instead, why not have 100 posts loaded, but only 10 or 20 visible? To get to the others you click an AJAX link. Only when you get to say, the 100th or 200th post does a new page need to be loaded. This makes for decent usability while keeping things search engine friendly.
- Value-add tagging and sorting system – Essentially, you’d be able to tag something on the backend on a scale of 1-5 on how useful a post is. Alternatively, you could base this on user ratings of how good a post was, but this depends on having a large enough user base to sustain this.
Then, users viewing your category archives could sort them by how valuable or popular an individual post is. This ’sort-by-importance’ view would be the default for search engines, ensuring your best content gets the most crawler attention.
What do you make of these ideas? Would you do any of this differently? Have additional tips on this? Why does All In One SEO noindex category pages by default, and likewise Joost’s robots plugin?
Gab Goldenberg writes on SEO at his SEO blog. Check out his services if you’re in the market for something professional.
Related posts:
- Wordpress SEO: How to Choose a Permalink Structure For this post we’re going to be taking look at...
- Wordpress SEO: Wordpress Security Why it Matters to SEO In recent weeks wordpress security, or more correctly the lack...
- Wordpress SEO: How to Use RSS and Scrapers to Build Links When you run wordpress or any other blog for that...
{ 25 comments }
I remember a fantastic Plugin, Custom Query String.
With CQS was possible to set the number of posts for every single page (author, archive, category,…).
Unfortunately this plugin is orphan and doesn’t work with WP 2.7:
http://codex.wordpress.org/User:MichaelH/Orphaned_Plugins_needing_Adoption/Custom_Query_String
Once you’ve un-noindexed the category pages, it’s a useful thing to add ‘page X’ to the titles and meta descriptions of pages 2 and higher. This makes them unique and helps users understand what the results are in SERPS. You can do it with a small bit of PHP:
<?php if ( $pagedMore here: http://www.malcolmcoles.co.uk/blog/avoid-duplicate-meta-descriptions-in-pages-2-and-higher-of-the-wordpress-loop/
They are noindex by default because I think it is better to show post detail page to searcher instead of category page.
Great post.
Another way to increase archival indexing is to rotate which products/posts appear. One week show 12 posts (or products), the next week show a different set. This is particularly good for new products which typically get dumped far back in the pagination.
You could also reverse index, meaning the oldest post show up first on a category page while the newer posts are in the back. Theory here is that since new posts are likely indexed already (because they are new and on the home page), you could afford to show the older posts in the category first.
Not sure how that’d be for the user experience, but that’s another story
In terms of why plug-ins no-index categories by default there was a lot of chatter about how the indexing of Wordpress categories lead to duplicate content issues and page rank “spilling” a year or so ago. This was especially relevant for blogs with 100 categories and 100 friends in their blog roll.
So many of the top “guru’s” suggested that no-indexing your categories, etc was a smart thing to do. For some it worked well. For others no effect. For a few it had a negative effect.
But that’s why you see many of these plugin’s no-indexing categories.
Gab
Great post .. Im going to test the 100 posts setting, but already the load time of 100 blog post is horrible. I guess we will see.
Right Chung, why not nofollow the links to category pages and block them in the robots.txt so they dont get indexed in the first place.
I adopted CQS and then someone adopted mine when tags switched over to WP core
http://moshublog.com/2007/10/30/custom-query-string-reloaded-for-wordpress-23-with-tag-support/
I believe that version is WP2.7 compatible
No additions other than “Thank you! I needed this!” Now to review the video and resources referenced!
Nice job, Gab!
@Chung – That’s where you and I disagree; I think that offering choice amongst hyper-relevant offerings is better. E.g. affiliate marketing posts on niche selection, affiliate manager relationships, etc.
Even if we assume that post detail pages are superior, you mostly should not be targeting the same keywords with posts and archive pages. The former should go for longtail (since anyways their topic is narrower), the latter (category pages) should hit the short tail, since the greater variety of topics increases the likelihood of satisfying the user query.
@Alberto, Malcolm, ANdy – thanks guys, that’s very helpful.
@ Garrett – That does make sense as far as cannibalization. G has some issues with a particular post of mine, and keeps returning the post’s category page when I search for the post itself on G, despite using unique longtail keyword combos. But then, that may be because my internal link building plugin directs a lot of juice to my category page, which targets a very close keyword.
All of you, thanks for adding to the post’s quality with your insights and valuable knowledge!
Malcolm, I’ve bookmarked that for personal use and added your feed to my reader.
Garrett, if I’m in the voip market, I’ll know where to go. As to your comment, great insights, though I disagree with the first two. User experience isn’t great (I swear the product was here when I bookmarked it!) and methinks there’s better workarounds for search.
But as to the category noindex bit, that may explain issues I’m having with a particular post. Or it may just be that I’m directing too much competing internal anchor text to a category, relative to that post.
Dana, thanks for the kind words and Andy, way to go on the plugin!
If you disable the follow,noindex tag in all-in-one-SEO for categories then how do you prevent content duplication caused by category pages? I really think an authority post that recommends turning off a widely accepted and beneficial feature should address this point.
We remove category pages for Google for a Good reason. We don’t like seeing a single post in Google’s index more than once. Sure Google will let us slide, most the time. But, what if Google decides not to let it slide? The impact on the risk is too high to ignore.
Please respond to the duplicate content problem created by removing noindex on your category pages so that I can make an informed decision on your article.
Thank you for this one. I own a site that offers well designed shopping cart templates and this post will really help me in gaining traffics to other pages of my site that are not often visited.
Thanks. very nice…
Thanks Gab…You gave a solution to the problem i was suffering from long.The information you provided here will help alot.
To add to Garrett’s comment. Here is a plugin that seems to work really well for rotating old posts back to the front page, with the option of making it visible in the feed or not: http://www.dagondesign.com/articles/scheduled-post-shift-plugin-for-wordpress/
Why not add a sticky to your category pages with a couple of paragraphs of unique content, then limit your category post list to title links – so none of the content is duplicated? Then allow the cats to be indexed…
You could also silo the category pages so that no cross linking contamination occurs…
If you disable the follow,noindex tag in all-in-one-SEO for categories then how do you prevent content duplication caused by category pages?
Google doesn’t care about this, surely. Content duplication is a filter, not a penalty. Google is not going to think your category page is the correct URL for an individual post. And no one is going to link to a category page mistaking it for your post, so there won’t be any link dilution. So this whole business with category pages and duplicate content seems like a red herring to me.
The one issue I do see is that often pages 2 and higher of a category loop tend to have the same title and meta description. So why not add page 2, page 3 etc to these automatically?
Here’s some PHP for wordpress to do this: http://www.malcolmcoles.co.uk/blog/avoid-duplicate-meta-descriptions-in-pages-2-and-higher-of-the-wordpress-loop/
Garrett, I’m working with some guys to create just such a plugin for cat pages!
@Greg – Just don’t make full posts available on category pages. use the “more” quicktag to cut off what can be seen. See also Malcolm’s second comment.
This is great, thanks for the tips!
I got rid of all of my categories a month or two ago.
I just didn’t display them initially, then I realized that I don’t even need them, because they seemed to affect results from the “Related Posts’ plugin. That also clears up a bunch of space on the sidebar.
This is exactly what tags are invented for. They allow you fine tuning of the content so you do not have a problem where you have to put 100 posts up. A typical (optimized) tag page will contain only up to 5 posts.
I made an experiment using phrase ‘fresh yoghurt drinks’ for which my tag page still ranks no1 in google without any other relevant content other then title.
http://www.prelovac.com/vladimir/improve-search-engine-placement-with-tags
What are you doing with the space then, Billy?
Some spot on advice here, in particular for e-commerce sites.
Ghaz
Wonderful tips… Thanx…
Comments on this entry are closed.