The name of Google’s new search algorithm is called “Hummingbird”. Derived from being “precise and fast”, it’s been the biggest change in Google’s search algorithm since the  Caffeine updat e in 2010.

There were also some major changes before like Penguin and Panda, but what sets Hummingbird apart from them is that two previous ones were just changes done to some parts of the old algorithm, while Hummingbird is an entire replacement of the old one.

One of the new search activities Google features is “Conversational Search”, which is meant for smart phone users who may find having conversations more convenient when doing research.

For example, you could ask, “What’s the nearest taco store from my home?” The traditional search engine will most probably focus on finding matches for your keywords like “taco” or “store”, and then lead you to a website owned by a restaurant that sells tacos but may not be necessarily close to your home.

Hummingbird makes Google almost human with the way it responds to queries. Instead of looking at keywords, Hummingbird makes Google look deeper and focus on the whole statement or question.

In relation to the example above, the new update helps the search engine understand that you’re looking for a physical store near your area of residence that sells tacos, provided that you use the Google domain based in your country.

In a nutshell, Google responds to whole statements and questions in a way similar to how a human being would by looking at the query as a whole, and not segmenting its keywords. This then makes the results to match the query better since it looks for concepts, not words.

The Hummingbird algorithm responds to end users by means of processing the whole statements or questions, and then looks for websites or webpages that has the most relevant concept to the query.
But how does Hummingbird do this exactly?

Hummingbird applies a statistical language approach in a way that it rewrites search queries to make it simpler. This can be done by browsing the large database of synonyms or other related terms that might be substituted in place of one or more terms in the query.

If a query indicates, “Where is the best place to buy burgers?”, Google will most likely try to find a substitute for the terms “best”, “place”, “buy”, and “burgers”. The words “where”, “is”, “the”, and “to” would require no substitutes since they hold no great significance – these words are called “skip words”.

The synonyms that are stored in the database were created through observation of what end users search for in their queries. If an end user, for example, submitted two consecutive queries, the words he sent will give the hint that the words are related to the ones sent in the previous one.

To illustrate, when an end user searches for “Korean idol” and then “Korean star”, Google will understand that the word “idol” and “star” are of the same context.

It was mentioned above that there are some words that Google doesn’t bother finding synonyms of. That’s because keeping synonyms of words takes up a lot of space in their database. Moreover, relevant search results can still be produced without the use of some words. Queries like “The mp3 player” and “mp3 player” would most likely have the same results. Processing the latter would be less time consuming because Google only has to find alternatives for the terms “mp3” and “player”.

Google’s “Conversational Search” is not really new since it was released almost two years ago. Once you speak through the microphone of your smartphone, you’ll see your words appear on the screen, but the difference now is that the “search by voice” function speaks back to you.

Google will answer you back in full sentences with a computerized voice while an  in formation car d  with the answer is placed on your screen. Sometimes, Knowledge Graphs will be employed if applicable.

But what’s magical about this application is that it makes Google almost human in a way that you can have conversations with it. Google now has the ability to understand a query’s context.

For example, I first asked the question “How old is Zooey Deschanel?” Google answered: “Zooey Deschanel is 33 years old”. Then I asked, “Who is her boyfriend?”. Google answered: “Her boyfriend is Jamie Linden.” The second answer was surprising because I did not directly refer whose boyfriend I was asking about by replacing “Zooey Deschanel” with the pronoun “her” and it still got the right answer.

Any other regular search engine cannot make a relevant response if the main subject of a query is replaced with a pronoun or other shortcuts that refer back to the previous statement. And that’s what separates Google from the rest.

The Knowledge Graph helps solve queries by conveniently showing images of what the end user is searching for, along with an information card containing important details. This lets the end user acquire a lot of information in very little time and with very little effort.

For example, you type in “Charles Darwin”. Knowledge Graph shows you his profile – his picture, his date of birth, his date of death, his famous works, and other important details of his life and career.

Knowledge Graph also understands that many terms have different meanings with reference to the several contexts these terms may be used. Hence, when typing a query, Google automatically suggests whole statements that were patterned to popular questions and statements that other end users type on the search bar.

Say, if you type in “Rhino”, you might be referring to the endangered animal in Africa, the bulky long-haired professional wrestler, or to an American entertainment company. Google Knowledge Graph suggests queries by posting some  images and some details  in the right side of the screen.

Also, Google’s so smart now that it can make comparisons between two elements. If you type in “Coke vs Pepsi”, a 2-column table will appear with a picture of both beverage, and below them will be details about them like its calorie count, total carbohydrate, and so on.

SEOs have long dreaded Google’s updates since they believe that these would mess with the ranks they’ve worked so hard for.

But Google said that as long as you have been following their age-old rule to make original and high quality content, then there’s nothing really to worry about since the Hummingbird was just meant to process information in a different way.        

In a recent blog post, Krystian Włodarczyk suggests that Google is intent on reducing the importance of keywords. What will always matter is the quality of the content that an author produces. Hence, the new algorithm only has massive effect on those sites which use dirty tactics in order to achieve their ends.

Although SEO’s purpose is to get traffic, it’s not about manipulating search engines in devious ways, sending out spams, or killing web design. It’s all about making quality content that’s friendly with search engines and people alike, in a way that both would have legitimate reasons to rank it high and share it.