Google introduced a new search algorithm Hummingbird a few months ago that moved beyond keyword search into the world of entity search. The session covered differences between entity and keyword search on Google & Bing. A discussion on how SEOs can and should take advantage of Hummingbird was also held.
Moderator:
Danny Sullivan (@dannysullivan), Founding Editor, Search Engine Land
Q&A Coordinator:
Mark Munroe (@markemunroe), SEO, SEORadar
Speakers:
Warren Lee (@Warrenleemedia), VP marketing, Lee
Bill Slawski (@bill_slawski), Director of Search Marketing, Go Fish Digital
Marcus Tober (@marcustober), CTO, Searchmetrics Inc.
The Evolution of Google and Entity Search
Warren started the session and discussed the immediate effects of the Hummingbird algorithm. He said that the algorithm is impacting 90% of the search queries on Google and that basically the update is released to cater to the mobile users in a better way. It is also an attempt to learn the contextual meaning behind the searches.
Warren said that links are still important and so is the site architecture and quality of content.
Knowledge Graph was the first baby step of Google and Warren said that Hummingbird used the Knowledge Graph to understand both – search entities and entity search. It doesn't change the original ranking factors, it adds additional relevancy signals that can also impact rankings.
Explaining the search entity, Warren said it includes:
- Keyword
- Organic search results
- Query session information
- Time frame
- Paid search results
- Anchor text in a link
- Domain name
Warren added "Google is trying to use search entities for learning about relationships between keywords".
Coming to Entity Search, Warren said that it is contextual targeting based on search history. He shared several methods for optimizing for these changes. The key takeaways of the session were:
- Long tail is as important as ever for providing search entity relational information
- Query expansion is the part of Hummingbird and it’s huge
Warren shared resources to get an idea of keywords related to your primary phase:
- Type query and see what Google recommends
- LSI Keywords.com
- Ubersuggest.com
Warren provided an example for search of a word "headache" and explained that earlier Google would show results on what a headache is but after Hummingbird, it showed treatment results such as Tylenol's page.
Hummingbird & The Entity Revolution – Knowledge and Search, Two Swords Sharpening One Another
Bill Slawski is brilliant at researching Google's patents to provide insights on what the search engine giant is doing. "Patterns" was the one word he repeatedly used for describing what Google is doing with the Hummingbird.
Bill said that Google is monitoring search patterns to learn more about the contextual relationships between words. He added that patents show that Google is using individual user behavioural data and aggregate to impact search results (Entity Search). This is helping Google in understanding the meaning behind more ambiguous searches and to provide more relevant results to users.
Summing up the content, he said that Google is:
- Viewing entities as search entities
- Possibly using RDF (Resource Description Framework) schema for tracking information to calculate the probabilities of what classification is meant by a query
- Searching for patterns in queries and on pages to answer questions
- Looking for Schema markup and schema-related facts & attributes information to create and understand the context
Like Warren, Bill explained the Hummingbird impact with "cats" word. Normally Google used to show pages about the pet, but after Bill searched for Broadway plays in New York, the results showed to him were for the musical cats. So, Google is learning about individual search patterns and preferences, is what Bill said.
He also discussed a new Google Knowledge Vault, which is extracting charts across the web to try and fill up any gaps in the Knowledge Graph's Knowledge base. However, Bill said it is yet to find out how largely it will affect the search. Concluding the session, Bill said that we all now Google is collecting a tremendous amount of data.
What is Hummingbird & The Entity Search (R)Evolution?
Last was Marcus who impressively covered the current effects of Hummingbird and its long-term implications.
He said that Google Hummingbird covers:
- Contextual search
- Conversation search
- Semantic search
The SEO paradigm has shifted from Keyword/variations & landing page to strategically finding the topic and creating a landing page.
The most important aspect we believe that was mentioned by Marcus was that how Google is looking for more than just the primary keywords. It is also looking for related keyword phrases in the content. Google looks at the page to have a holistic view of the topic that will provide complete answer to the searchers. He specifically stated that SEO has grown from go-down-the-list tactic to strategy, which includes understanding a topic and building a complete view in order to answer all the queries of the searchers.
Because of these changes, an interesting effect has taken place. Marcus revealed that since the launch of Hummingbird, link diversity has decreased by 7%. So, less number of websites that Google trusts are ranking now, where lesser known websites used to rank earlier. He said that it will be interesting to see if the trend continues and the effect it will have on the smaller websites. But, it implicates that SEO has to adapt to survive.
Summarizing the session, Marcus said that "Hummingbird is not just a standard newly released algorithm, it is a learning algorithm that will get better and better over time".
