Search Engine Optimization and Semantics

Frederick Townes 

Despite the awesome drawing power of the www and its ability to sell products and broadcast messages, as site designers and owners, we've only begun to harness the true power of a fully-compliant semantic web.

Semantics aren't new. They've been around since scribes wrote on mud tablets. We use semantics everyday in the way we write and speak. Semantics are nothing more than a set of communications conventions to which we all agree.

We rely on punctuation, grammar, spelling and sentence structure to effectively communicate with others. However, throughout the world, a different set of semantics has been developed for each language. In English, the umlaut isn't used except by Motley Crüe. And you won't see Õ, ĝ or ¿ used in English writing because they aren't English semantics. The differences between distinct linguistic conventions correlate closely to the problems of accurate and effective communication via an asemantic web.

HTML: The Universal WWW Language

It stands for HyperText Markup Language and it's the accepted programming code used for the design and development of web sites. HTML is easy-to-learn, very flexible, it handles text, graphics and other multimedia efficiently and it's totally scalable, providing the same results for mega- and mom-and-pop sites alike.

However, HTML has one striking drawback. HTML defines where to place what content on a site page, but it can't discern relationships between the content, or how it could be used to provide more valuable results to user queries. That's why SEs exist in the first place – to try to develop100% relevancy in a world with “too much” information.

A page of product listings. HTML defines that this is <catalog page 06>. However, it doesn't identify item #4432 as a hair brush costing $4.95. HTML is limited to identifying a block of text describing the hair brush that should appear next to the product image. However, it isn't able to determine whether the content is about a hair brush or a manure spreader. HTML places content, makes it interactive with web users and provides site navigation via coded links. What it can't do is read and index content to make that information more accessible (read useful) to visitors.

Using proper semantics, SEs are able to identify relationships between various chunks of site text, and from a site skin POV, semantics better equip visitors to more easily identify these relationships, whether via browser or as human eyeballs reading the screen.

The use of compliant semantic standards accomplishes two more important objectives for site owners. One, standards enable site designers to more accurately describe site content to SEs and two, semantics facilitate clearer communication between site and SE spiders. These two key benefits deliver more targeted, motivated site traffic.

The bottom line is this. With today's HTML technology, search engines are able to spider and index web sites. However, they can not index, assess (weigh), correlate and deliver

specific content based on the context of a user's keyword search.

RDF, OWL, XML, et al

To address the limitations of HTML, compatible programming languages and machine-readability apps are available in the site designer's resource mix. RDF, Resource Description Framework, is one such tool. So is Ontology Web Language (OWL). And the data-driven XML, eXtensible Markup Language, has vastly expanded the ability of web designers to catch the attention of search engines through the development of indexable content.

The implications are enormous, increasing the usefulness of the web exponentially. Here's an example. Using HTML alone, a search engine won't be able to produce useful links to the keywords ‘website designers in Boston'. Instead, the SERPs show lots of designer sites all over the world, some of which might be located in the Boston area. Though the sought-after information is available, it's contained within the pages of many sites – all indecipherable to spiders.

Further, site content is useful within a variety of contexts. What if we substituted the keywords ‘website designers in Massachusetts' referencing the example above? This SE user has changed the context of the search (and therefore relevant content) from links to Boston-based design firms to links for the same across the entire state. A semantic web would not only deliver specific data, it would do so within a wide variety of contexts.

Instead of simply indexing an entire site, semantically-enabled SEs index content for use in a wide variety of contexts, compile this information based on a user's query and deliver SERPs with links that take users to the specific information for which they're looking.

W3C and Semantic Code

The World Wide Web Consortium, or W3C, has embarked on an important mission – the standardization, improvement and extension of the www. Not a simple task. It entails the standardization of programming languages, development tools, browser specs and now a semantic web – a more powerful, useful and dynamic interface.

However, W3C has many challenges to overcome. First, semantic web technology is still under construction.

Second, there's rarely agreement within the web community on which apps are best, which languages best serve global needs and so on.

Third, universal semantics are impossible. Semantics for one language won't sync up with another. Even writing styles – academic with lots of footnotes vs. a blog entry – employ different, accepted semantics. Thus, a universal set of semantics would involve cataloging types of content – a virtually impossible task.

Instead, semantics should be viewed as a set of guidelines to help programmers, designers, software developers and others to use the W3C-compliant tools and protocols properly. Indeed, because the task of connecting web users to specific information is shifting from search engines to site developers, all that remains is for semantics tools to be used properly.

SEO and Semantics

The whole point in developing semantics standards is to make websites machine readable. This will allow search engines to spider and index content with vastly improved specificity. Good SEOs do this while optimizing sites, usually without realizing just what they're doing. Imagine the results of semantics optimization when developer and SE marketer collaborate to make best use of this emerging technology.

Search engine users will be able to access local news and sports, local sales, locate and compile similar content from widely-disparate sources and in general, access more data within any user-defined context.

With a semantics-compliant web, SEs will be able to better identify relevancy of and relationships within the copy itself and thus save search time and deliver more traffic to that website. The more semantic-compliant a site, the more relevant traffic it will generate.

Site Semantics and Design

If your site designer isn't up to speed on w3 semantics, find another designer. Yes, new and improved semantics tools are in development but the big SEs are constantly improving their ability to match context and relationships instead of simply identifying literal matches of character strings. And, if your site isn't semantically optimized, it's also not optimized for conversions, so you're missing the best SEOpportunity to come along since the advent of search engines.

Can you afford to ignore site semantics? Not if you intend to be around tomorrow.