Dan Hanley

Handy hints: optimising for Google

I think Google’s blog posts are always a good starting point for site optimisation ideas, and last Thursday’s instalment is no exception.

As usual, the post focuses on ongoing changes to search, and there is enough detail to provide a glimpse of what matters to the search engines but not enough to make it easy to game them. This is perfectly understandable – if Google gives too much detail then things can become skewed very quickly, as we’ve seen in the past.

However, of the thirty or so updates this month, three are particularly worthy of comment, and provide good scope for making improvements to your site.

Soft 404 detection

A “soft 404″ is what happens when a user performs a search, and the web server returns a good page, but not the one you were looking for.

What should happen in cases like these is that the web server sends a page telling you it can’t find what you want, with a status code of 404 (“not found error”). With a “soft 404”, what actually happens is that the user still sees the page which tells them it can’t find what they want, but with a status code of 200 (which indicates that the page is ok).

A soft 404 makes no difference to the user, because they still see the page telling them the page they want hasn’t been found. However, it does make a difference for automated tools like the Google crawler. Because the page has a status code of 200, there is no obvious way to distinguish it from the page the user asked for. This error happens surprisingly often, and regularly from sites whose mangers should know better.

Over the years I’ve heard many sorry excuses for soft 404s, but it usually begins with a simple small accident: someone configuring the web server uses an absolute file path instead of a relative path to point to the 404 page.

This eventually snowballs and sometimes results in towering edifices being built to support an infrastructure for managing soft 404, because they are believed to be “the way the system works”.

The best way to deal with this error is to avoid having broken links in the first place – but if you do have them, make sure you configure your servers to correspond correctly.

More Rich Snippets

Rich snippets are the pieces of information Google takes from your web page and pulls into its search results – and they’re getting more sophisticated. Google is concentrating less on keywords and using semantic technology to get to the heart of what a page is about – throwing up product names, prices and telephone numbers, to name a few examples.

There are special tags you can add to your content to make sure Google identifies the “right”  information for your page. The standard for these tags is RDFa. To make sure you’re heading in the right direction, clients of ActiveStandards can use the “Aboutness” reports to identify areas where they may wish to use RDFa tags to ensure their content is delivering the right results.

At Magus, we’re big fans of semantic markup, and the use of rich snippets is just the tip of the iceberg for these technologies. We’ll be adding improvements to the Activestandards Aboutness feature in the coming year which should make handling rich snippets even easier.

Faster Mobile Browsing

Mobile internet use is expanding at an exponential rate. And what do users want? Speed. You can speed things up for them by linking directly to your mobile site’s URL – this helps avoid all the browser sniffing and page redirects users generally have to contend with. It’s a simple, safe and sane optimisation to make.

So, there we have it – three Google improvements that you can use as a basis for ongoing site optimisation.

Of course, I’ve managed to get to the end of this post without mentioning the most controversial Google update of all: integrating Google+ into search results. I’ll deal with that next time.