Hummingbird Algorithm Update Guide
Reading Time: 4 minutes
What is the Google Hummingbird update 2018 algorithm?
Occasionally Google updates its algorithm in an attempt to refine the search results it delivers. Striking fear into the hearts of website owners and SEO professionals alike, an algorithmic update can alter, and hopefully improve, the results a searcher receives. This can have a dramatic impact on what is displayed and to whom. In this regard, Hummingbird was a little different to previous updates such as Panda and Penguin; these two were more focused on the quality of an inbound link profile and of the actual content itself, whereas Hummingbird seems more focused on how search is delivered rather than the state of the content being delivered.
Released sometime towards the end of Summer in 2013, Hummingbird looks more closely at searcher intent when it comes to delivering results; focusing on matching search queries based on an increased understanding of semantics and context.
SEO experts tend to agree that the upshot from a digital marketing perspective is that whilst Panda and Penguin contributed to significant ranking turmoil and reports of lost traffic and previous indexing, Hummingbird’s impact was more widely felt in local search engine results and didn’t correlate so much with reports of dropped rankings and fewer resulting website visits.
At the same time, it seemed to positively improve the relevancy and accuracy of Google’s knowledge graph; Google’s repository of data used to populate info boxes that surround regular SERPs (search engine results pages) from a variety of different sources that aim to better deliver a contextualized and semantically relevant set of answers and solutions to search queries.
Semantic search and the knowledge graph
Hummingbird’s most significant impact was felt in Google’s semantic search abilities; an advanced search feature that is learning to infer search meaning by parsing intent and context, and in it’s knowledge base, referred to as the knowledge graph, used to deliver more contextually relevant information.
What is semantic search?
The objective of semantic search is to generate a more accurate set of SERPs against search queries based on a far broader understanding of the meaning of the words beyond their individual definitions; context changes everything when words are aligned with each other and the subtle changes in meaning and intent are not always clear to humans, let alone learning machines. Google’s Director of Engineering and futurologist, Ray Kurzweil said:
“Natural language understanding has evolved substantially in the past few years, in part due to the development of word vectors that enable algorithms to learn about the relationships between words, based on examples of actual language usage. These vector models map semantically similar phrases to nearby points based on equivalence, similarity or relatedness of ideas and language.”
We can see this research becoming apparent not just in Hummingbird’s SERP delivery but also in the way we are offered suggested responses in Gmail or advice in how to finish a sentence we have only just started writing.
An example of this interpreted searcher intent might be in a simple search query such as “best place to get Thai”. The words individually are ambiguous at best unless we understand the searcher’s intent and the context in which the question is being asked. Semantic search might enable Google to infer ‘best place’ to mean ‘restaurant or takeaway’ and ‘Thai’ to mean a type of cuisine rather than any of the many other meanings that the word alone might otherwise be taken for. How does Google then use the knowledge graph to deliver results based on the query intent and context?
What is the knowledge graph?
Google’s knowledge graph attempts to gather data from various sources and deliver them in response to a semantically enhanced search. Alongside the regular, organic search results on a search results page, you will quite often see additional information displayed in a box to the right of the listings. For example, if you search for a celebrity, TV show or country (or pretty much anything as the knowledge base grows), as well as returning search listings from the web, you’ll almost always see a knowledge panel containing an image and various bits of key information about the subject. If you search for food, you may also see a list of potential recipes, or if you search for a place you may see a map and so on.
If we refer back to our previous example, ‘best place to get Thai’; we may expect to see a list of local restaurants in order of distance from our current location, along with reviews, quick links and even telephone numbers. The objective is to return contextually relevant information based on the subject and the perceived intent regarding the subject.
The future of Hummingbird for search and SEO
There is no doubt that Google will be continuously working to improve and develop Hummingbird as part of their continuing and clearly signaled commitment to delivering increasingly more relevant search results. One area we can expect to see more of this is in voice search. As our lives become more integrated with the Internet of Things and we are increasingly surrounded by devices with which we can vocally interact (Google Assistant, Amazon Alexa, Apple’s Siri et al) so we can expect to see more focus on how these devices return answers and search results to us.
Clearly, trying to infer intent from individual, keyword by keyword, phrases will not work to deliver very accurate results but by having a clearer understanding of how the words combine contextually and gain a better understanding of exactly what a searcher is looking for, so will machine learning algorithms like Hummingbird become more advanced by necessity.
For the SEO community, this means more opportunities to gain contextually relevant visibility and better target their content at their audiences. More accurate and less spam-filled, local search results will additionally work well for businesses with specific geographic relevance and help to deliver increasingly more accurate, geo-targeted results to potential customers.
For the content creators, it’s more about creating high quality content that speaks to the intended audience more than it does to the search engines as we are entering an era in which search engine results are going to be tailored far more accurately to the needs of the searcher; and if website owners, SEOs and copywriters generate work with Hummingbird in mind, everybody wins.