How Google Hummingbird Affects Daily Life

What do we really know about Google Hummingbird? – Not a lot…exactly.

Google doesn’t tell us a lot about the algorithm, how it works, or even when it was launched (most agree August 2013…roughly…probably rolled out in stages to different geo locations). To understand the changes Google has made to its algorithm it’s good to start by looking at Google’s search history.


Google’s History – This graphic is a good visual of where Google has come over the last 15 years, offering insight into where they are heading. Google Search Graphic. The Google Search Timeline includes the following milestones:

  • Did You Mean? – 2001 – Google gives search results that include guesses as what you might have wanted to search.
  • Synonyms – 2002 – Google includes similar search terms.
  • Google Local – 2004 – Later it becomes Google Places.
  • Auto-complete – 2005 – Google automatically fixes spelling errors.
  • Universal Search – 2007 – Search Results include “Vertical Searches” in other search databases like web, images, news, local, sports, stocks etc.
  • Google Instant – 2010 – Google’s search bar automatically tries to guess what you are typing while you type it, giving you suggestions before you are finished typing.
  • Knowledge Graph – 2012 – Knowledge Graphic offers suggestions of content that may not be links to websites or other vertical searches but Google content and even answers to questions.

When Google started in the late 1990s, their goal was to revolutionize web search by offing the most relevant search results base a technique called page rank. While other search engines showed results based on which sites included the greatest number of a given search terms (keywords) on a specific page (later known as keyword stuffing), Google looked at which sites were being linking to (back-links, inbound links or incoming links) and then gave results accordingly. Essentially Google determined which sites were most relevant based on a new set of criteria. At this stage Google was only interested in finding websites, they were not interested in other types of searches, content or devices… not yet.

However, as the years have progressed Google has included more and more into its search results and into the algorithm that generates these results. Synonyms, Autocomplete, Universal Search, Google Instant and Knowledge Graph are all examples of Google attempting to generate more accurate and more specific results. These results are then tied into Google’s growing database of information. Information that now extends well beyond the original scope of just websites. Vertical search can include places, images, video, news or even answers to simple questions.

Now the emphasis is on generating “Semantic Search” results that get to the intent of search. Intent can be based on the context of the search—where the search was located, the user’s network and/or the user’s device. Search is now not just about finding websites but determining intent and finding the relevant information and answers. A search result can be a person, data, directions, news or sports results. What the user really wants is determined by the semantics for the search and the search entity.

What does this have to do with Hummingbird?

From what we can tell, Google has been heading in the direction of semantic search,  mapped to a variety of vertical search date sources for at least a few years. However, many of the previous updates to the algorithm have been patches to the original search engine. According to Google, Hummingbird is a brand new engine. It is safe to assume that this new engine takes full advantage of Google’s existing content, properties and strategies and that it does it in a much more efficient and scalable fashion. It’s also safe to assume that Google intends to transition to newer semantic-based search results slowly, over time, and not abruptly (hence, not announcing the Hummingbird until a month after it had been implemented). It’s also safe to assume that the techniques Google employs to cut back on spam search and “Black Hat SEO” are still in place and that there is a dial that Google can turn up if search results become too spammy.

 “Keyword not provided” – ouch!

If you are like me, and often spend hours deep in Google Analytics, you might have noticed a greater number of searches listed as “Keyword not provided”. This can be a problem for a search engine marketing (SEM) professionals because it eliminates a key strategy when trying to determine how people are finding your site and what search terms they are using. Google would say that they are moving away from specific keyword searches and would prefer that SEMs focus more on quality content and effective landing pages. This is always a good idea regardless, but it looks like with the roll-out of Hummingbird, the keyword data that we came to rely on is going bye-bye.

What’s better with Hummingbird?

Again, the search engine itself is a black box, only the engineers at Google really know what’s going on. However, Google will tell you that their goals include better conversational search, faster search, better voice recognition and a better understanding of context; place, contacts, and interests etc.

Why does this matter for my business?

Because marketing strategies, particularly bad ones, can be hurt by the new algorithm, dramatically effecting ROI, it’s important to understand the changes at Google. However, if you have been employing best practices with your online marketing, there isn’t much to worry about. The take-away should be that Google is going to continue to increase the accuracy and breadth of its search results, focusing more on semantic search.

What the experts say:

  • “We don’t anticipate making any dramatic changes in what we are doing. What we are talking about is how we create quality, engaging, shareable, linkable content. It has become a core piece of our SEO strategy.” – Christy Belden of LEAP
  • Jenny Halasz of Archology believes SEOs have become so keyword focused that they’re putting emphasis on the wrong things, explaining that many are, “Trying to reverse engineer data that really isn’t actionable.” She thinks SEO should be less about keyword data and more about customer engagement.
  • “Trust is now king. The primary goals of semantic search is weeding out irrelevant resources from SERPs.”  – Trond Lyngbø of Metronet in Norway (SERPs = search engine results pages)
  • “When Google switched to Hummingbird, it’s as if it dropped the old engine out of a car and put in a new one. It also did this so quickly that no one really noticed the switch.” – Danny Sullivan, author Search Engine Land
  • “Better content over more content.” –  Rand Fishkin of Moz

Key Concepts



Key Players