5 Tips on Optimizing for Semantic Search

The Hummingbird update to the Google algorithm in the fall of last year was, and continues to be, a game changer for SEO. Unlike the Penguin and Panda updates to come before—both partial updates to the existing algorithm—Hummingbird was a completely new algorithm that changed much of the way that search functions. Known for its focus on providing intuitive search results based primarily on user intent, Hummingbird understands the relationship of keywords or phrases to one another and uses this to rank websites relevant to what the user has searched.

Semantic Search

Semantic, or conversational, search existed before Hummingbird, although the new algorithm is the first to integrate it so completely into its ranking system. Semantic search allows users to search entire sentences or phrases and receive results based on the collective meaning of the keywords in those phrases. While there are many other features of Hummingbird that SEOs should keep in mind, semantic search is important as it changes the way we think about keywords.

Almost one year after the unveiling of Hummingbird, SEOs are still discussing how to optimize for semantic search. Here are five tips to keep in mind:

1.)  Revise Your Keyword Research Strategies

“’Semantics’ refers to the meaning or interpretation of a word or phrase”, according to Search Engine Journal. With that in mind, keyword research is the best place to begin. How will a page on your site’s keywords be interpreted when Google crawls it? By taking a holistic approach and breaking your keywords into three tiers, a well-rounded list that accounts for variations in user intent can be presented to the search engines.

  • Level 1 – Core Keywords: This list is comprised of keywords closely related to your initial target keywords. They should be variations of your targets close enough in meaning so Google can consider your site if any one of the core keywords are searched.
  • Level 2 – Thematic Keywords: Whereas the keywords in List 1 are somewhat synonymous with one another, thematic keywords are further removed from your initial targets, yet are conceptually related. If your target keyword for a page is “manhattan realtors”, a list of thematic terms like “new york city apartments” can help you potentially rank for the query, “low rent new york city apartments”.
  • Level 3 – Stem Keywords: Your third level should include keywords that answer users’ questions. These keywords anticipate the information users need after they have found your page, and should be integrated into the content to naturally provide answers. Once a user has found your page by searching “low rent new york city apartments”, it is likely they are seeking information on “finding low rent new york city apartments” or “renting affordable new york city apartments. Your Level 3 Keywords may be some variation of “rent new york city apartments” or “new york city apartment listings”. 

Ideally, keywords from all three levels should be incorporated into your content.

2.)  Create a Robust Content Outline

This isn’t new to Hummingbird—its just good SEO. But with Hummingbird focusing on user intent, pages that rank are pages that, in addition to being optimized, provide relevant content.

Keyword stuffing has never been best practice. When optimizing for semantic search, it is important that your keywords are well placed, organically integrated into your content, and not overused. Create content around the questions and searches your keywords seek to answer. Outline which keywords are targeted for which pages, and how the content on each page will meet the user’s needs. From there, build out your content to be robust and to directly relate to the keywords you’ve delegated to that page.

3.)  Integrate Social Media

Social media is a significant component of Hummingbird’s personalization of search results. Google can now provide more refined, intuitive results pulling from personal data in users’ social media profiles. Search Engine Land had a great article on the topic that stated:

This suggests that paying attention to social search is becoming more and more critical, and that social media is playing a larger role in search results, sending strong signals to the search engines. Leveraging (and discovering) your target audience’s interest graph is key to producing content that will bring them to your website.

Essentially, fully integrating social media into your marketing strategy—creating content that reflects your audience’s social interests and sharing that content on your networks—is a large part of optimizing for semantic search.

4.)  Structure Your Data

Semantic search depends on structured data. Information on your website needs to be properly tagged, marked up, and organized in order for search engines to crawl your page. More than ever before, SEOs must be well versed with the back end technical details of a website that search engines recognize as indicators of a page’s relevancy. Working closely with webmasters, or familiarizing yourself with HTML markups will ensure that your site is easily crawlable. Schema.org provides a collection of templates and markups that Google, Bing, Yahoo!, and Yandex rely on.

5.)  Continue What You’re Doing

Since the launch of Hummingbird, Google has maintained that not much should change day-to-day for SEOs. While there are some specific things SEOs should keep in mind when optimizing for semantic search, that statement is, for the most part, true. Hummingbird rewards SEO best practices, and actually penalizes sites that use black hat tactics. A poorly optimized page will simply not rank. But pages that are properly marked up and focus on user intent will benefit.

Lastly, cross-pollination is important: collaborate with social media and other disciplines in search marketing, like SEM. Search engines are becoming more intelligent and responsive to users, and SEOs also need to think about how to anticipate what users ultimately need out of their search.

Link to this post:

<a href="http://fuzeseo.co/5-tips-optimizing-semantic-search/">5 Tips on Optimizing for Semantic Search</a>