SEO Principles to Bear in Mind when Setting up a Blog

It is of primary importance that Google (or any search engine for that matter) find each article in your blog on one url only.

I would also recommend that it see this article belonging to as few categories as possible - one if possible.

Snippets

Snippets are an introduction to each article. These (instead of the article) are what should be displayed on the blog's home page and any category pages.

You can use the first paragraph or two of the main article as the snippet, but I prefer to have a unique snippet that is a quick summary of (or better still, a teaser for) the article.

By using snippets we guarantee that the main article text appears on one page only. This will go a long way to solving duplicate content issues, and should also ensure that anyone minded to link to the article is clear which page they need to link to.

Categories vs Tags

Basically the distinction I use is this: categories do not share content, tags do. In other words, an article could have multiple tags, but will only belong to one category. Therefore, even if you use snippets, the more tags that are used per article, the more likely it is that there will be duplicate or near-duplicate pages in the site.

The point is to plan your content and how it will be organised. Because you can tag an article after you've written it, many people get lazy with their classification systems. Instead, spend some time thinking about what you will be writing about and draw up a system to organise it in one consistent manner.

Using Categories

Each article belongs to one category only, and has only one possible url path. This means that you can use url-rewriting to create virtual folders and a site hierarchy, so instead of urls like this:

www.somesite.com/blog/article.php?category_id=5&item_id=23

we could first swap the ID numbers for text identifiers:

www.somesite.com/blog/article.php?category=prime-ministers&item=tony-blair

which can then be rewritten to: www.somesite.com/blog/prime-ministers/tony-blair

The second is more user-friendly (people don't trust cryptic urls) and better for SEO. Furthermore, because there is one path to each article we can use 'breadcrumb navigation' to help people see where they are in the site i.e.

You are here: Home > Blog > Prime Ministers > Tony Blair

Using Tags

An article belongs to every tag it has. If we use the tag in the url path we will end up with duplicate pages. Using our 'Tony Blair' example, and supposing the article was tagged 'labour politicians', 'election candidates' and 'prime ministers', we'd have these pages:

www.somesite.com/blog/labour-politicians/tony-blair
www.somesite.com/blog/election-candidates/tony-blair
www.somesite.com/blog/prime-ministers/tony-blair

This is obviously no good as we've created three versions of the same article. We can fix this in two ways:

1) Abandon the Folder Hierarchy

Each article sits only in the blog folder i.e. www.somesite.com/blog/tony-blair. We lose our folder hierarchy, but Google will still associate the item with the paths it takes to reach it (i.e. via the tag pages).

2) Use the Google Canonical Tag

The rel=canonical tag was developed by Google as a fix for problems like this.

Where you have pages that are very similar or practically identical, you decide which is your 'main' page and which are the 'dupe' pages. Then you place a tag in each dupe page. So if we wanted Google to see

www.somesite.com/blog/labour-politicians/tony-blair

as the main page we’d place this tag:

<link rel="canonical" href="http://www.somesite.com/blog/labour-politicians/tony-blair">

in the other two pages. Google says that it will then effectively amalgamate the three pages; it will only index the main page, and no PageRank will be lost. We can also keep our breadcrumb navigation.

Bear in mind however that this will require programming a system whereby the blog automatically

• inserts the correct canonical tag into any dupe pages as and when they are created (by further tags being attached to the article), and
• either doesn't allow removal of the main tag from an article or, if it is removed, then assigns one of the dupe pages to be the new main page and alters the tags on the remaining pages accordingly.

What about the breadcrumb navigation?

I've seen cookies suggested as a means of controlling the breadcrumb navigation - in other words, show the hierarchy that was used to get to the page. I don't like this because it doesn't help people who end up on the page without hitting a tag page first. I would prefer to show either one path or all paths to the item in the breadcrumb navigation i.e.

You are here: Home > Blog > Prime Ministers > Tony Blair (see also: 'Labour Politicians', 'Election Candidates')

Different snippets depending on the tag?

If you're only using a few tags per article, then you could write more than one snippet per article. This will

• eliminate duplicate content (in fact you create extra content because you have more category pages)
• enable you to angle each snippet to the subject of the tag page
• send more PageRank to each article

The Date Archive

This is a feature on most blogs, but I don't think it is always required. The important thing is to think about the purpose of the blog.

If you intend the blog only to be used as a simple 'news' section and don't plan to update often, then it could be that a date archive would be your only system of navigation.

If you plan to use it more often and for different purposes, then categories or tags would be recommended. But do you need a date archive too? Ask yourself this: is it important for people to be able to see what you wrote on your blog in December 2009 - or will they only care what you have written about a specific subject?

If there is an element of 'current events' to the blog then a calendar would be useful. If however the content in the blog is 'evergreen', then I personally would deem a calendar less useful or even unnecessary.

If a date archive is deemed to be important enough to keep in the site, then the simplest way to deal with it is to insert

<meta name="robots" content="noindex,follow">

on all of the pages in the date archive. This will stop Google from indexing them but allow PageRank to flow through them to the actual articles and the rest of the site.

Using Ajax 

If you are going to have more than one system of navigation in your blog then I would use Ajax to stop Google crawling any but the main one.

In other words, work out if you want to use categories, tags or the date archive as your primary means of navigation. Then import the secondary (or tertiary) navigation menus via Ajax, where the Javascript function is in a directory that robots.txt forbids Google to crawl.

This means that you offer humans all the functionality that they expect on a blog, without compromising your blog's search engine friendliness in any way.  

Is that all?

Yes - and no. So far we've covered your content and how to present it to Google in the best way. In the next and last section of this article I'm going to talk about other people's content on your site - comments on your blog. If it's on your site, it is effectively your content because you are responsible to Google for it. 

previous: Blogs vs SEO: Problems with Wordpress and other Blog Software
next: Stopping Blog Comment Abuse and Spam