Search engine optimization is a complex and ever-changing game, with a variety of dimensions to consider, but good SEO will always require an intimate understanding of what is actually "optimal" in the eyes of a search engine like Google, Yahoo, or Bing. What will be surprising to most web startups is how little exact keyword matches are beginning to matter in improving your PageRank.
As Google's algorithms are adjusted and refined over time to perfect search relevance, internet marketers and SEO specialists must stay ahead of the curve by understanding how "relevance" is actually being defined by the search engines they are trying to appeal to. Search engine optimization is a complex and ever-changing game, with a variety of dimensions to consider, but good SEO will always require an intimate understanding of what is actually "optimal" in the eyes of a search engine like Google, Yahoo, or Bing. What will be surprising to most web startups is how little exact keyword matches are beginning to matter in improving your PageRank.
For years, SEO in natural language searches was defined by the ability to predict and exploit relevant keyword phrases, bolstered by factors such as content, traffic, and social media linking. Semantic search has been used for several years as an indicator of relevance, but until recently it was limited to keyword synonyms in close proximity, with keyword matches augmented by good content being the gold standard for true relevance. This has begun to change rapidly. The increased preference for semantic keyword search is partly in response to the proliferation of black-hat SEO tactics like keyword (and keyword synonym) stuffing, but mostly it comes from an understanding that the Internet has simply gotten too big for exact keyword matches to play any significant role. Search engine users are rarely willing or even particularly capable of doing a targeted keyword search, and indexing every website that responds to "mexican food los angeles" isn't likely to serve as an efficient filter or narrow down the results in a meaningful way. In their never-ending quest to break down these natural search terms and discover true relevance for their users, Google has responded to such changes by, among other things, expanding and improving the semantic search capabilities of their Hummingbird algorithm.
Traditionally, a search engine with semantic search function will start with exact keyword matches, and then index those results based upon their close proximity to similar keywords, synonyms, or related searches by users who began with the same keyword phrase. The number and placement of these search synonyms provide a grounding for what one might call "likely relevance", and results are then further refined by content and reliability factors like external links or bounce rate. The problem with this strategy, at least at a simplistic level, is that it is just as easy to stack your pages with keyword synonyms as it is keyword phrases. By generating three or four likely alternatives to possible search terms, web designers can keep the number of keyword matches on a single page to a safe level, while stacking the rest of an otherwise irrelevant page with semantic keywords that will light up Google's algorithm. This is textbook black-hat SEO, and not the basis for the solid web presence you want to build upon. So how exactly is Google responding to this abuse of its semantic search term function? By simply broadening the field.
Last year, a Google search for something like "movie tickets online" would trigger semantic search terms like "movie tickets", "buy movie tickets online" or "buy tickets online". Now, Google's semantic search has expanded to account for longer phrases and more broadly related text, text which may not even contain a single word from the initial keyword phrase. Continuing with the hypothetical search for "movie tickets online", semantic search may now account for phrases like "search showtimes", "sell tickets online", "find local theaters", or "upcoming releases". Hummingbird is attempting to anticipate what its users want, to determine not just what they are looking for, but why they are looking for it. And for those attempting to establish themselves as a local business, or an up-and-coming online service, this can actually be a good thing. Predicting semantic search terms may become more difficult, but providing good content remains key to a good PageRank, and may prove to be a more powerful predictor of success as Google gets better and better at what they do. Generating relevant keywords and likely search synonyms remains important, but for the sake of SEO, designers may be better off spreading these trigger words out over a longer page, with text that explains how the keywords relate to what you do as a company. The most successful companies online are those that provide a service that users really need. With the content perfected and the demand established, Google Search will eventually find its way to you.
Learn more about best SEO practices and how we can help your business.