views:

106

answers:

3

Standard SEO caveat: It's a black box, and the algorithms are proprietary, and trying to predict and game the search engines is a crappy way to make a living.

That said, what are the baseline steps you want to take to make sure your content is visible to the major search engines (Google, Bing, etc.)

I'm specifically curious as to what role your URI Information Architecture plays. It's common wisdom that you want keywords in your URI, and you want to avoid the query-string laden approach, but what else beyond that?

A quick example of what I'm talking about. Based on something I read on a forum, I recently exposed a /category/* hierarchy on my site. In the following weeks I noticed a sharp uptick in my page views.

I'm curious what other basic steps a site/weblog should take with its URIs to ensure a baseline visibility.

+1  A: 

some simple things ...

  1. meaningful and accurate meta fields (especially description, keywords)

  2. a valid hn hierarchy on every page (e.g. h1 h2 h3 h2 h2 h3 h3 h4 h3 h2)

  3. all (text) content accessible to a text browser

  4. check spellings

  5. keep content and display functionality separated (e.g. use HTML and CSS fully)

  6. validate CSS and (X)HTML and use standard DOCTYPES

  7. relevant <title> for each page

  8. sensible site hierarchy and no orphan pages

mas
+1  A: 

A few URI tips that have kept me ranking:

  • Write URIs in English but include a unique ID. SO does this well: http://stackoverflow.com/questions/1278157/seo-and-uri-structure
  • Stay consistent when linking to a page: domain.com/, domain.com/index and domain.com/index.php are different URIs
  • Use .html extensions, or purely /one/two/ directories for pages

There's probably hundreds of other tips! The structure of linking plays a very important role too...

  • Logically break your site down into many categories/subcategories
  • Link all pages back to your homepage
  • Don't link to hundreds of pages from your homepage

EDIT: Oh I forgot a very important one - a proper 404 response!

Hopefully that helps a bit

Al
Could you expand on the logic (if you know it) of the unique ID AND English URLs. I've always been under the impression getting the ID *out* of the URIs was a better practice.
Alan Storm
In this answer http://stackoverflow.com/questions/1142480/adding-redundant-information-to-a-mvc-route/1142508#1142508 I discussed the practicalities of having the id and English in the URLs. Also, the ID of a resource won't change where as the English part might. If you include the English but work from the ID you're in a much safer place.
Garry Shutler
AFAIK it adds an element of trust from the search engine's point of view. In my limited experiments GoogleBot was happy to munch through more pages if they have UIDs. It was also important to be listed in Google News but I'm not sure that's relevant any more. From a programmer's point of view it's great :)
Al
Right, from the programmers point of view it makes sense to keep the ID in there, but going from a pure, gut level, ivory tower late 90s IA point of view I always assumed having an ID stick around is bad. Assuming it IS good, it does answer a lingering question I've had about dates in the URL of weblog posts. Given that a date provides some level of uniqueness, it's probably a good idea to use them.
Alan Storm
A: 

1) Don't use www subdomain if you do not have to. If you or your company has made the mistake of using subdomains for asset management then you likely forced into using www just to be safe.

2) The biggest problem faced by search engines is redundant URIs for the same page. This problem is solved by using a canonical link tag in your HTML. This will perhaps help you more than any other single SEO factor.

3) Make your URIs meaningful. If people can remember URIs well enough to type them out your SEO will be significantly improved.

The most important factors with URIs is easy to remember and the ability to specify uniqueness to the search engine. Nothing else matters with regard to URIs and SEO.