The real benefit of valid markup
I happened across this article a couple days ago, and immediately it had my attention. SEO Testing and the effects of having valid semantic HTML are of great interest to me. The findings were not completely surprising, but at the same time shocking. Is Google’s algorithms really that leanient on bad markup, and can they really be that easily manipulated by basically anyone who can type out the ugliest of markup? Yes.
The problem is simple, but too tricky even for them to figure out. Most websites are not coded properly and would all have to be lowered, making any site that is coded properly stand out, regardless of the content quality.
My proposed resolution: a search engine which only indexes valid content. While being parsed, it is checked to be valid and that the Doctype is followed. If an error is found then disregard that page but still collect and follow the links if possible. That search engine would be a dream.
So what do you really get from writing good markup then? Satisfaction that your doing things properly, that your code will be readable in the future, that your site is better than 90% of other sites, and most of all community recognition. Suddenly your a fellow saint in the web community, well… maybe its not quite that glamorous. A valid site is so often praised that, even though Google may not look at it any differently, the new found links to your site will certainly help that PageRank out. But who really cares about PageRank anymore anyway.