Optimising your blogs, forums, images and search results for search engines

One of  Google’s early patents sheds some light on how a blog post will fare in a search, particularly a search on Google’s Blog Search. I credit Bill Slawski of SEO By The Sea for highlighting the patent’s relevance to blog ranking. That post is over a year old but his interpretations of a couple of patents will get your mind thinking about algorithms. You can read his post but, in essence, there are two scores involved in a Google blog search result:

1. Relevance Score, which could (note “could”) be based on:

• How often does the searched keyphrase occurs in a post.
• Placement of the searched keyphrase, e.g. in title tag, H1, etc.
• Enhancement of the searched keyphrase, e.g. bold, italics, font size, etc.
  

2. Quality Score, which could be (implied in this patent) based on:

 Positive Signals, such as:

• Subscriptions from news aggregators such as FeedBurner.
• Number of searches for a keyphrase and the number of times a particular blog is selected.
• Listing of the blog in blogrolls, particularly in reputable blogs.
• User-generated tags such as the ones enabled with WP-Cumulus for WordPress
• Forum posts, references in emails, chat sessions that contain a link to the blog.
• PageRank of the blog’s home page and, eventually, the PR of the post.
  

Negative Signals, such as:

• Infrequent posts.
• Mismatch between the post content and what was in the RSS feed.
• Presence of spammy words in the post.
• Identical length of posts.
• Link concentration to one main target site.
• Ads in the body of posts (ads in the sidebar are OK).
  

SEO for WordPress

WordPress is a free blogging platform that you can install on your site very easily if your host has given you a control panel with such an option (if you have cPanel as the control panel, then the Fantastico option leads to many packages you can install, including WordPress, Joomla and many others).



You upload WordPress plugins to the /wp-content/plugins directory.

WordPress PlugIns

WordPress can be enhanced with the help of free plugins, which you can download from www.wordpress.org or from other sites. Here are the ones I use when I have SEO in mind:

• All in One SEO Pack: It enables you to generate meta tags, optimised page titles and avoids indexing of duplicate content.
• Headspace2: It is a new competitor to the above plugin.
• Akismet: This spam filter comes with WordPress but you have to get a special key to enable it.
• Google Sitemap Generator: You can generate the sitemap.xml file for all search engines, not just Google.
• Math Comment Spam Protection: Commenters have to solve a simple arithmetic test, e.g. 2+7 to prove that they are human and not a spam robot.
• FeedBurner FeedSmith: Helps to publicise your blog to subscribers.
• Link to Me Textbox: Makes it easy for someone to cut and paste code to link back to your post
• Robots Meta: Adds numerous robot-related controls to prevent some pages from being indexed, such as author and date-based archives (duplicate content concern), opting out of Yahoo and DMOZ directory descriptions being used in SERPs, and so on.

SEO for Content Management Systems

There are thousands of content management systems (CMS). Joomla is my open-source CMS of choice and it is free. You can install it on your site very easily if your host has given you a control panel with such an option. About half my sites run the old version 1.0.x while the rest run 1.5.x. I am using 1.0.x examples here but most of the tips will work for 1.5.x.

Basic SEO for Joomla

Out of the box, Joomla is not as search-engine-friendly as we would prefer. You can take certain precautions to improve this deficiency:

• Joomla comes with sample data and datafeeds. Remove this data, or better still, opt out of installing it if you are using a control panel installer such as Fantastico. Otherwise your site will contain information about Joomla and Linux, which might not match the theme of your site. I have one such site where I am still discovering sample data in hidden places.
• The Site Name field in the configuration area should be short, e.g. APC Magazine. Don’t stuff it with keywords, else you may find that Joomla uses it for unexpected purposes, such as in the Subject field of emails sent from your site.
• Joomla can generate PDF files from your content, accessed through a PDF icon on the page. While this may look impressive, it is duplicate content and the PDF file has no links that a spider can follow, so turn off the PDF feature. This also applies to the Printer-friendly icon.
• Don’t add buttons for social bookmarking sites unless you can use rel=“nofollow” for those links. If your content is unlikely to be featured on the home page of those sites, don’t use such buttons. You don’t want to annoy such sites.
• Redirect the non-www URL of your site to www.example.com using an entry in .htaccess (covered in Part 7, March 2009).
• Enable the cache, as it will speed up the page serving to the spiders and humans alike.
• Use search-engine-friendly titles for Sections and Categories, for they appear in URLs.
• Check the Google Sitemap file before submitting it, as it might contain pages that are not useful to a visitor. (Joomla can generate duplicate URLs that end up in the sitemap.)
• Check the default Joomla meta data and empty it or replace it with something relevant to your site. Note that updating Joomla replaces your customised meta data, so keep a copy and re-enter it after the update.
  

Which SEF Component?

You should create search-engine friendly URLs to replace the long ones generated by Joomla. I use a discontinued version called OpenSEF on my Joomla 1.0.x sites and SmartSEF for version 1.5.x. Others prefer one called SH404SEF – the choice is vast.

Example
Internal URL: index.php?option=com_content&task=view&id=13&Itemid=30
Friendly URL: /courses.html

OpenSEF can mislead you if you follow its default settings, as it will create the Google sitemap in its own component directory. Google will not accept this, as the sitemap needs to be at the root level.

So, create an empty text file (0 bytes) called sitemap.xml and upload it to the root directory.  Make it writeable. Then, in OpenSEF, supply the sitemap XML file name as “sitemap” with no extension. The location field should be left blank.

Then go to the Google Sitemap option and choose Generate XML.

Next, save the XML file.

Next, submit it to Google Sitemaps.

Login to Google Webmaster Tools and explicitly add this sitemap. You are now done.

SEO for Forums

Discussion forums pose a problem to spiders because their architecture is based on user convenience and not search engine ranking. For example, a discussion thread (the lowest level in the file hierarchy) can be reached via multiple URLs, such as (partial list):

• The direct URL.
• A list of new threads on the home page.
• Printer-friendly version.
• Member list and their recent posts.

This can create a duplicate situation as well as sending the spiders in a loop that they cannot escape (other than their own time limit). Since you want the spiders to crawl important content just once, you can adjust the robots.txt file accordingly. The following is from a forum that uses vBulletin, considered one of the best scripts for this purpose.

User-agent: *
Disallow: /test.html
Disallow: /forums/login.php
Disallow: /forums/printthread.php
Disallow: /forums/member.php
Disallow: /forums/memberlist.php
Disallow: /forums/calendar.php
Disallow: /forums/sendmessage.php
Disallow: /forums/subscription.php
Disallow: /forums/showpost.php
Disallow: /forums/showgroups.php
Disallow: /forums/search.php
Disallow: /forums/report.php
Disallow: /forums/image.php
Disallow: /forums/misc.php
Disallow: /forums/register.php
Disallow: /forum/attachment.php
Disallow: /forum/avatar.php
Disallow: /forums/newthread.php
Disallow: /forums/usercp.php
Disallow: /forums/newreply.php
Disallow: /forums/private.php
Disallow: /forums/index.php?referrerid
Disallow: /forums/sendtofriend.php
Disallow: /forums/forumdisplay.php?page
Disallow: /forums/forumdisplay.php?sort
Disallow: /forums/forumdisplay.php?order
Disallow: /forums/forumdisplay.php?pp
Disallow: /forums/forumdisplay.php?daysprune
Disallow: /forums/forumdisplay.php?do
Disallow: /forums/showthread.php?page
Disallow: /forums/showthread.php?pp
Disallow: /forums/showthread.php?p
Disallow: /forums/showthread.php?mode
Disallow: /forums/showthread.php?goto
Disallow: /forums/showthread.php?post
Disallow: /forums/showprofile.php

The following is a robots.txt file from a forum site that runs phpBB software:

# Applies to all robots
User-agent: *
Disallow: /forums/posting.php
Disallow: /forums/search.php
Disallow: /forums/post-
Disallow: /forums/updates-topic.php
Disallow: /forums/stop-updates-topic.php
Disallow: /forums/ptopic
Disallow: /forums/ntopic
Disallow: /sys/
Disallow: /calendar/
Disallow: /include/
Disallow: /resources/

SEO for Image Search

Image Search in Google can also be exploited with SEO techniques. Sometimes an image result appears above the first organic result, so it is worthwhile to optimise images for this type of search. Depending on the reason for someone searching for an image, there can be an advantage in certain industries. For example, an image search for a type of jewellery could be someone researching to buy, whereas a search for a picture of an African elephant could be a student researching a school project.
Currently, Google looks at the text around an image to get a context for it. It also uses other markers such as Alt tags, anchor text, and so on. A recent paper by Google researchers proposes a better way to analyse the images themselves and refers to “VisualRank”, which is similar in concept to PageRank for pages.
The following techniques are known to help rank high in an image search:

• Larger images (300 pixels by 300 pixels or more) tend to be favoured for appearing in search results.
• Use keywords for file names, anchor text, Alt tags, and in the page containing the image, just as you would otherwise optimise the page.
• Include images in a site map, or build a site map purely for images.
  

SEO for Local Search

Local search is very important for small businesses. There are many services whose customer base is within a few kilometres from their outlets.  The broad guideline is to ensure that the website contains geo-location or at least postal and postcode reference that can be reached by the spiders. Additionally, there are location-specific websites that are particularly beneficial for SEO value.

GeoTags: This site helps you to create geo meta tags such as:

• <meta name="geo.placename" content="Milpitas, California" />
• <meta name="geo.region" content="US-CA" />
  

There is no current SEO benefit in using this method but some SEOs are using such tags.

Microformats: The hCard generator enables you to code your name and contact information in a geo-location-friendly format.

GeoURL: Submit your URL after you have placed an ICBM meta tag such as this:

• <meta name="ICBM" content="-37.53,145.05" />

Craigslist: Craigslist is a very popular free classifieds site that is available in many countries. A listing in a remote town lasts for 45 days. Be very careful not to spam it.

Google Local: You can register your own site and have it show up in a local search.

Address information: Sites that display properly formatted addresses and telephone numbers can be picked up and displayed in a local search result.


 
APC Magazine in Google Local Search

SEO for Mobile Search

To optimise pages for mobile search, consider the following tips:

• Ensure strict compliance with W3C XHTML standards.
• Follow accessibility (508) guidelines.
• Build lean pages with the use of linked CSS files to keep formatting off the page. Provide a separate style sheet for mobile users.
• Some people recommend using “display:none” in CSS as a way to show some content to desktop users and not to mobile users. 

Next: Diagnosing a Sick Ranking
Previous: Getting SEO benefits from social media
All: The full SEO & Web Marketing series so far