Five Years of SEO: Hummingbird and Knowlegde Graph
Where changes like Panda, Penguin and “not provided” are changes that have the most impact on the ‘back end’ of things, aka the work of an SEO, changes like Hummingbird and Knowledge Graph have much more impact on what we actually see in the front end. And it gives the SEO’s headaches as well.
What are they?
The reason I’m putting the both in the same article is because the two changes are very much connected. The way we’ll approach it in this article is therefore very semantic. Because that’s what it is :-).
When Google rolled out the Hummingbird Update in 2013 Danny Sullivan described it as ”Google changing their entire engine, but keeping the car intact.” This to emphasize the size of the change: it was huge.
Anders Hjorth also feels it is a huge change:
“This is the promise of moving away from keywords and into real language and deeper understanding of intent.”
But what had exactly changed? Everything! Well, sort of. Google seemed to have changed the way it looked at search so it seems. Where before Google looked at search as a keyword-based action (you are looking for something so you type that into the search bar), they now looked at search as ‘communication’. Upon the actual launch Google hinted at the changing way of looking at search very clearly:
“Users expect more natural and conversational interactions with a search engine”.
Google is expecting people to have much more of a communication with the search engine, they are not typing in keywords anymore, they are asking questions. So we can talk to the engine and get answers back now.
Take a look at this ‘conversation’ I had with Google about the Eiffel Tower:
You may notice a few things:
First of all, it’s a conversation; they are not search queries on its own. When I search for Eiffel Tower related elements it gives me that information as expected. As soon as I start leaving out the actual name ‘Eiffel Tower’ Google however still understands that my questions are related. This is Hummingbird at it’s best: it has a better understanding of whom Google is ‘talking to’.
Semantics are key here. It is based on the search behavior of course, but Google is also taking a lot of other things into account, like your social profile and data from the people around you. See the third video below for an example on how ‘personalized’ this can get.
The answers come from the Knowledge Graph. Knowledge Graph is something Google has been doing for a few years now. It is trying to give people answers to their questions, as quickly as possible. Which basically means you don’t have to click through to a website anymore, Google will just give you the answer. Want to know how tall Justin Bieber is? Google to the rescue!
Knowledge Graph also gives other information, which means that in many cases we don’t need to click through to websites anymore. Google is working with partners on this so it means they aren’t ‘stealing’ the information from websites. But, as Barry Adams points out, they are taking away traffic:
“Google has broken the unwritten rule of web search:
Websites provide content to search engines, and in return search engines provide traffic to websites.” Increasingly, Google is taking websites’ content and not giving traffic back. Knowledge box SERPs, ever more common, are stealing traffic from websites that publish original content, and the new site search box is another example of Google’s nefarious tactics to steal traffic from websites so it can show more ads.”
This taking away traffic is based on the information, which is shown in the SERPS on some results:
No need to click through anymore, because Google is giving us the answers directly in the SERPS!
The Knowledge Graph clearly shows the intent of Google. At SES San Francisco in August, Matt Cutts said:
“…one of the key focuses for Google is to move away from being a search engine and focus on becoming a knowledge engine. Google is so committed to this that Google’s Search Quality team has been renamed to Google’s Knowledge Team.”
Google combines the Hummingbird Update and the Knowledge Graph into a way of working as they said they would: a conversation with an answering machine.
Take a look at this video for example which shows how Hummingbird and Knowledge Graph work together very nicely:
There is something to be said for the opinion that Knowledge Graph and Hummingbird have had a much larger impact on the work of an SEO than Panda or Penguin. This because it means a completely different approach is asked for which is not ‘technical’ or even taking or not taking certain steps. It is much more about understanding the audience.
This means a lot more effort for the SEO’s and also a different approach, much less analytical, much more human.
The impact of that it is more difficult to be seen, as Andre Alpar points out:
“The Knowledge Graph keeps drastically lowering publishers’ opportunities whose business model is to monetize users (or their eyeballs) with advertising.”
This means the combination of Hummingbird and Knowledge Graph has ‘hurt’ SEO’s and certain website owners. It also means a new way of thinking has to be implemented: not trying to give the answers, but being visible in the conversation. Which is a lot harder.