I’ve been updating themes recently. One of the many great things about WordPress is that it generates markup with microformat classes, thereby giving your posts a ‘semantic boost’. This means search engines will be able to extract information such as dates, authors and titles easily from your posts, and — once micordata has been properly adopted — more sophisticated information, such as entry summaries (using the entry-summary class).

I enhance WordPress’s work by adding more classes from the hatom microformat (including said entry-summary class). While I was updating my Scherzo theme a thought struck me. If Google is able to extract semantic data from a page it must look at the text enclosed by tags with a class attribute. What if it used this functionality beyond the scope of microformats? For example, what if it extricated some about information from this HTML?:

<small>© 2010 Leon Paternoster ∼ <span class="about">A blog about web design and writing</span>. Built on WordPress & the Scherzo theme by Leon Paternoster.</small>

I decided to use this HTML (about isn’t, as far as I’m aware, part of any microformat) on my blog in a place where you wouldn’t expect to find it (namely, the footer) so that I could be reasonably sure Google wasn’t looking at any other structural factors when extracting data. Note that the code appeared on every page of my website.

I also added a meta tag with some different content because Google normally uses this text when it displays search engine result text:

<meta name="description" content="A blog about web design, writing and the internet.">

I waited a few days and Googled myself. Here’s what Google displayed in its search results:

So you’ll see that Google did in fact extract the text wrapped in my about span. Interestingly, Google appears to give this markup more weight than the meta tag.

You’ll also note that, for some reason, Google displayed the text around the extracted meta data, resulting in a rather messy entry. But that’s not the only problem. The search for ‘Leon Paternoster’ returns several of my blog entries (Improving WordPress search in the above image). Instead of displaying the text wrapped in the post’s entry-summary, or some text from the post content, it adheres to the text it finds in the about span in the footer, which is pretty useless as a description of the post’s content (and shows that Google interprets about as information about the document rather than the author).

It’s good that Google attempts to extract meaning from a page’s class and ID structure, but I think that straying from a set of pre–defined rules (such as a microformat) makes it difficult to predict how your markup will affect search engine listings (or understand exactly what you’re marking up) —you should certainly use the about class with care.