Site Architecture is very important for success of the Web today. For optimum success, you need to match your techincal & information architecture to the search services in which you wish to feature – this should be worked it out in advance.

Architecture can be broken down into two segments: Informational Architecture and Technical Architecture

Informational Architecture basically covers:
the mission of your site
the content therein
organization of the site
navigation on the site
labelling of keywords
search in the site

Technical Architecture basically covers:
Hardware and software
Web search services

You site’s mission should be to match the searhers’ mission. To do so, you need to do keyword research to find out what their mission is and work out what you need to include on your site and how to present it. You need to do things that make the most sense for humans – search engines will pick up on that and follow.

The best skeleton for any site is a hierarchical structure. It allows you to:

• have easy site management
• easily predict pages that spiders will or will not crawl
• easily cutt of spiders using robots.txt
• easily create links from other sites and directorires
• easily build a site with breadcrumb trails

Some tips:

• Keywords in anchor text help search engines work out the most relevant pages on the site for those keywords
• The words you use in your labels makes a huge difference.
• Try to always link to items from the homepage that you think will be of interest to most visitors
• Bury items more deeply if they are more niche interest

Look for these issues:

• Canonicalisation – this occurs when several different URL’s are treated as the same URL
• Same content at different URL’s – often on dynamic websites and also, if same site available at different domains

Avoid these issues by:

• making good use of robots.txt and robots meta tag
• only build links to primary domains
• make good use of redirects (301 (permanent), 302 (temp), 307′s)

Normalization is another issue worth mentioning. This determines the standard form of a set of identical URL’s. Search Engines’s normalise all URL’s before queing to index.

Also, look at location. Use a local domain name .co.za for South Africa and .co.uk for UK or similar and host your site in the country you’re in or the country that your content is relvant to. Make sure primary Name Servers are in the host country and get links from other sites and directories that are in that country.

Another important thing is that all visitors should see the same URL. When using dynamice URL’s, make sure that it makes sense because people will link to it. If you don’t, inbound links will be harder to obtain and rankings wil lbe harder to achieve

Problem summary
Same content at different url’s
different content at same URL

Best advice is that “one piece of unique content should be indexed once, at the best URL”

Final ideas:

• make correct use of domains, subdomain, subdirectories
• keyword rich domains don’t really help
• SE’s don’t crawl links in JS or are JS
• SE’s don’t index content written in JS or pulled in by Ajax
• SE don’t support cookies – don’t rely on cookies
• 404′s – install 404 handlers to catch errors
• offer sitemaps for faster indexing
• link naturally out from your site – don’t use rel=nofollow