Most SEO companies are aware that semantic code has an effect on how Google indexes and ranks a website. Many web master’s, however, are unaware of the extent it’s influence has, so let’s cover the basics.
Coding semantically simply means following development standards and using best practices which have evolved over several years to build the best website or application possible. It is often referred to as structural code by some developers.
It’s a simple idea, but the benefits this can have on SEO could be the difference between Google loving your site and hating your website!
Common sense tells me that Google would need metrics with which to rank their location and social based results. I believe, as there is only so much a robot can guess, that these new metrics will be taken from data such as post date on forums and blogs or location information from on-page or in profile information.
Other changes, such as Google’s golden ratio page layout algorithm update, are targeted more at user experience while on a website. These kinds of changes could be very difficult to change on an active site developed a few years ago.
Most of the changes sound fairly small on their own, but the large number of small changes quickly adds up. It may amount to little more than making sure all post have dates and locations set, but the boost to well coded sites may mean old sites and less proficient developers may be indirectly penalised simply by not being standards compliant. The common theme I see in all of Google’s recent updates implies closer adherence to development guidelines. Google seems to be aiming for better coded websites all round, which is a great idea… If you can code perfectly.
Several new features in HTML5 allow designers and developers to define content on a far more granular level than HTML4. The <article> and <nav> tags, for example, in theory make it easier for Google to tell where the site’s main content blocks and navigations are located with the hope of reducing the time spent crawling data it is aware of and dedicating more time to other aspects.
Google is, however, sitting on the fence towards HTML5 by saying things such as “it won’t help, but it can’t hurt” but I believe they will start to place more weight on such factors moving forward.
Technically, HTML5 is not semantic code as it’s in a specification which is still being worked on and could change before public release. This is unlikely, but a possibility, nonetheless.
Server-side code should always be written to perform well which may sound painfully obvious but is often overlooked when time lines are involved. Google allocates a set amount of time and resources to each site it crawls so if it has to wait for poorly coded pages to load or process data it will detract from the time it spends performing other tasks. This is a small oversight which can lead to massive penalties if a site generates a lot of content or traffic.
Each language has different ways to improve performance, but most are simple to implement retroactively, so there is no excuse. A few examples in PHP would be to use comma concatenation when printing many text variables or definitions (instead of full stops), using a ++$i instead of $i++ incrementer inside really big loops, using echo instead of print, using full file locations on include and require statements, avoid using @ error suppression.
Faster code means more time for Google to do what you want… Index your content!
Time to update?
It may appear to be an altruistic act on Google’s part as it would provide results more based on site quality, but it could seriously damage a companies income to be overtaken by competitors with less resource intense pages or simply newer sites.
I think it’s great that Google is moving away from the links + content = rank approach, but a little more transparency would be nice.
If you have a website designed using tables, with many images, or simply an using an ineffective class or method it may be time to consider updating your code. Google seems to be saying: Keep up-to-date or be left behind!