29 May 2013

SEO Automation: Page Title Wins and Canonical Conundrums

Over the years SEO’s and developers have worked together to design, test and implement automation methods which support technical and content search engine optimization goals. From the first use of dynamic insertion of database elements into page titles and product headlines, to the advanced eCommerce “tricks” we see in play today, automation is the key to SEO efficiency. This is especially true for large publishing or eCommerce sites which generate new and updated content on a daily basis.

CogSearch engines themselves have empowered Webmasters to try new ways to streamline content and deliver the right pages and digital assets to appropriate searchers. Beyond the continually evolving authenticated Webmaster Tools environments, protocols supported across Google, Bing, and Yahoo have been regularly launched. Examples include dinosaurs such as Meta coding, robots.txt and XML Sitemap files to the middle age <nofollow> attribute, canonical tagging, and microformatting, and on to current techniques including updated Schema.org and rel= coding which can define content authorship and brand ownership.

The goal to allow page-level code to specifically speak to search crawlers and provide boundaries and paths for ideal indexation is a noble one, but I feel that overambitious SEO automation has led to many problems when it comes to resolving conflicts during requirement-definition and implementation. The following are a few examples of wins and losses that I have recently come across when doing industry research. (Please note that this is not meant to act as an endorsement or critique of those performing SEO for any of the sites mentioned. I don’t know who did it and understand that many SEO environments have included “lots of chefs” of both the SEO and I.T. sides over the years)


A Couple Wins:

Page Title Dynamic Optimization – this is one area I a rarely ever see butchered by SEO. Most big eCommerce brands that leverage product level have mastered the concatenation requirements to make decent looking titles.Google Site Colon Dicks Sproting Goods

Note that the above example also shows evidence of sound Meta description automation, with clear product details and a call to action included

Publishers have long also benefited from the “just use the article title for the page title” trick, as made most famous by WordPress blogs:Google Site Colon Huffington Post

The above example is one reason many SMBs are considering WordPress as not just a blog platform but also a CMS to host all of their site content.


A Couple Losses:

The biggest problem I have come across recently is the use of Canonical tagging as part of an automated content delivery technology. In some more catastrophic implementations (I will not name names here to protect the possibly innocent), multiple pages end up being tagged as the “chosen” one. This must be controlled.

More of a conundrum can be found in the following search, with a Footlocker result at #10:Google Search

 The page above has a canonical tag pointed to the one in the results:

****link rel=”canonical” href=http://www.footlocker.com/Womens/Running/Shoes/_-_/N-25Zf9Zrj****

The good news is that this page is the true canonical for the “Women’s Running Shoes” category. Some would argue that a little more time spent on optimizing the URL would be worthwhile, but the canonical seems to have worked. The “loss” from an SEO perspective is that they just stopped with that canonical for all versions of the pages that can be generated by the system. As you can see from the below example, even though many pages have the canonical applied correctly pointing to the main version, the others are still getting indexed and likely weighing the core page down due to perceived duplicate content:

Google Footlocker Text Snippet Search

In the above example, some people would say it’s Google’s fault for indexing that many versions, but don’t hate the player…

One more loss I see commonly is when Titles are automated and Meta Description tags are not, or not properly implemented:

Google Site Colon Brooks Brothers

In this example, you can see where some of the top level pages which appear at the bottom of the results shown do have “clean” Meta descriptions. These may or may not have been automated. The majority of the results however show very weak description snippets, which could be much cleaner through SEO automation.

In conclusion, SEO automation can be a very powerful ally in the drive towards better organic search performance. We didn’t go into details about microformatting which is a huge miss these days for many big and small brands, as well as the proper implementation of the Open Graph protocol. Using the tools we have available to us is the only way we can continue as marketers to be thematically relevant to the search engines.