Show you blog some love

Though "normal" blogging sites such as mine are probably below the radar of governments agencies, European laws mandate for sites to ask about users consent before storing cookies - even though no modern site cannot do without. As I use Google Analytics to keep track of visits, I prefer to try to be compliant…​ just in case. Google provides an out-of-the-box script for web masters with the relevant documentation. Two flavors are available, one pop-up and one bar. You can choose it by calling the relevant method.

I thought I achieved some degree of SEO via the SEO Smart Links Wordpress plugin ages ago. In essence, I just installed the plugin, the configuration itself being very crude: you just had to edit the different files.

I recently stumbled upon another plugin that is much easier to configure, the WordPress SEO by Yoast. It provides a friendly user interface with which you can design the pattern of the SEO metadata e.g. blog name plus post title, the type of metadata - Facebook OpenGraph, Twitter Card and Google+. This can of course be overridden on a case-by-case basis. Moreover, it offers a preview of the snippet as it would be displayed on Google search page. There are also plenty of configuration settings I’ve yet to use (and understand).

The preview provided by Yoast’s SEO is nice but doesn’t replace the real thing. Fortunately, Google offers a neat near-real preview page to test in real-time the final result of your changes (no caching involved). The greatest advantage of this preview feature is that you may add set an URL but also raw HTML code instead. This lets you check how Google parses the metadata of your unpublished pages.

For sake of completeness, there’s a similar tool provided by Yandex (the Google from Russia) that gives slightly different results. From my microscopic experience, Yandex is slower, but seems to less lenient and more importantly, provides solutions. For example, it tells me that the date value is not in the correct format (ISO 8601) and that I forgot the itemscope attribute at some point, which corrected the error I had on both Google and Yandex.

Both were very effective in pointing out one problem: it seems the post’s author linked to a URL that displayed the home page. Since it didn’t give out a 404, I had no hint about that. (I fixed that by a simple 301 Redirect)

All this automated stuff is nice but somethings cannot just be automated. For example, I’d like my book Integration Testing from the Trenches to be referenced on the page with the right kind of information and this, I have to do by hand. I’ve searched for metadata frameworks and it seems there are two W3C standards, RDFa and Microdata and OpenGraph by Facebook. The latter seems to be quite crude, so I chose to use RDFa and Microdata. For example, for the page about my book, I use the Book schema and for the page introducing me, I use the Person schema.

Interestingly enough, the book is quite straightforward: either add the relevant itemscope, itemtype and typeof or itemprop and property attributes  to an existing tag or to a new span and you’re done. However, this is getting harder on the Me page; challenges include:

In the end, I noticed the whole SEO thing is still very in its infancy. There are a lot of contradictory norms, standards and proprietary stuff around. The good thing, is that I think I achieved some things:

On the other hand, I’ve checked other sites, and it seems they do only the automated kind of SEO, and not manual. I can only infer why: it’s going to cost you a lot of time. All what I did is probably not necessary, I could have used the Yoast plugin to only override the title, the description and the image. But as always, it made me learn stuff.