SEO Correlation Between Code and Search Engine Rankings

[PaulDylan]

I posted this on my blog and wanted to get everyones opinion on this (http://palatnikfactor.com/2011/06/07/seo-correlation-between-code-and-search-engine-rankings/)

I’m always looking to see what top ranking websites may be doing to get the rankings they do. One of the tasks of any SEO I guess is to really analyze competitors, right?

I want to really stress that what I am writing here is completely opinion based and have not (due to time) validated this correlation enough but would like to get the discussion started. Nevertheless, I did enough research to see that there may be a correlation between code validation and top ranking websites, at least for certain queries where the number of real big players/brands is limited or non-existent.

So, what do I mean? http://validator.w3.org/ validates code on websites. This tool shows you errors and warnings that may be making it harder for search engines to crawl your website. Looking at top competitors for certain niches, I was surprised to find that top sites had very few errors compared to 2+ page rankings. That’s not to say that all the sites on the first page had fewer errors (cleaner code) than websites in the 2^nd page plus. However, again, top ranking websites for keywords that I was looking at had cleaner code which may have a correlation in regards to organic rankings.

What’s your take? Does this have any effect in regards to SEO?

[RyanKent]

I have a lot of thoughts on this subject. If I was to make a blog entry on this topic, it would span multiple pages or have to be broken down into sub-topics.

I do think there is a correlation between good code and search engine rankings. I do not think there is a correlation between a w3c validated page and search engine rankings. The validator is not current enough, nor flexible enough, to accommodate the real world situations which websites encounter.

Example A: HTML 5 is recognized by all major browsers. W3C validation of HTML 5 is still experimental. A specific example that applies to SEO is the canonical tag. According the the W3C validation site, the canonical tag is not currently valid. Take a snippet of HTML5 code which passes validation, add a canonical meta tag to the header, and the code will no longer pass validation. This is a direct conflict between best practices and validation.

Example B:  The world's most popular web page, google.com, does not pass validation. Matt Cutts discussed the topic. In short, they had a choice between providing code which validated, or code which worked. They chose to go with the working code.

Example The standard facebook widget code, youtube video code, and other popular code does not pass validation.

Whenever I design a website, I check the code in the validator to look for errors. Initially, I will find numerous errors related to code outside of my control such as social sharing widgets or youtube videos. Once I remove that code, the page often validates. I have researched the issue and it is possible to modify the facebook code or youtube code so that it still functions and passes validation. Doing so requires extra effort, it provides absolutely zero benefit other then saying "hey, I pass validation", and there are often drawbacks such as having to add extra javascript to your site which can otherwise be viewed as unnecessary code.

[MarieHaynes]

It's totally conjecture, but I would say that code validation doesn't have any bearing on ranking.  Some of my early websites that I created that are not validated are ranking pretty well.  (But then...I haven't checked whether my competitors are validated at all.)

Here's my theory...if a site is validated it means that some quality work has been put into it.  So, if someone put that kind of quality work into the coding, they likely have put good quality work into the SEO and every other aspect of the site.