Any announcement of changes by Google makes SEO professionals very nervous.

Google was secretive before Hummingbird was implemented. It’s never mentioned in official documents and nothing was added in an announcement made by Amit Singhal, leaving people to wonder what the change would mean.

In the past, Google updates have meant waiting for results rather than act on intuition. These updates usually have to be analyzed to find the right answers. There is often not one single correct answer to meeting the requirements.

The press has continuously analyzed documents including the original announcement; an FAQ by Danny Sullivan, which only references AMIT Singhal’s statement; and documentation attributed to related patents. Since Hummingbird represents such a deep change in Google’s fundamental algorithm, it is compared to Caffeine in 2010. The update changed how Google indexed billions of documents.

Hummingbird, instead of changing how the search engine indexes things, analyzes what the user is doing to offer the best results. The infrastructure change is both dependent on the Caffeine update and solves the issues seen in Google’s algorithm after the 2010 alteration. These included poor-quality results on SERPs. Patches such as Panda, Penguin, and others were implemented to mitigate such problems.

The series of updates since were fine for short queries, but long tail queries still caused issues. Programmers have sought to enable Google with natural language recognition. Technologies from Metaweb and Knowledge Graph, for example, were implemented to customize searches and handle long-tail queries so more complex concepts could be interpreted.

The Inner Workings of Hummingbird

Synonyms have been important to Google for a while; at least since 2002. Even in 2001, disambiguation was used analyze queries. A post on Search Engine Land last year, by Vanessa Fox, indicated synonyms were being used to reflect the user’s intent and optimize the results of their queries. The post also discusses how using a thesaurus alone does not ensure the most relevant results.

Even Amit Singhal has admitted this before. Synonyms may have similar meaning, but are not always used in the same context. Hummingbird provides Google with the tools to understand the context of the user’s language. While synonyms are still important, Google now focuses more on search entities, reflecting an objectification of the concepts defined by words. The context changes the relationship between words and this is what the search engine looks for.

These search entities are identified by determining the possible meanings of a word. A probability score is then used to determine what information is retrieved from the web. Knowledge Graph identifies the user’s intent in a similar way, while the Hummingbird patent also factors in co-occurrences of terms.

What the Resulting Algorithm Enables Google to Do

Google can now understand the intent of user queries and find a larger number of documents which may serve to satisfy the user’s search. Also, information delivery is simplified, as the search engine does not have to suggest alternate pages. It can also increase the probability the search results will meet the user’s needs and provide more effective ads.

Google announced Hummingbird would affect 90% of queries. Since there were no changes in traffic and rankings reported, users questioned this fact. The mix-up in SERPs during late summer was coincidental to Hummingbird’s release, as conversational searches are not often tracked by SEO experts. No one could track long-tail keyword rankings because Hummingbird does not concentrate on keywords.

Traditional ranking factors still count, but the update reflects a new framework. Relevancy and authoritativeness are just as important for ranking. Most likely, sites impacted were those concentrating on very long tail keywords, without much authority. In theory, one can use linkable and shareable content semantically related to long-tail keywords. Poor quality pages simply will not do the trick any more.

On Using Schema.org for Site Rankings

While rumors suggested Hummingbird would rely on structured data, concentrating on Schema.org’s model will not address the concerns of SEOs. The goal is to create understandable context on a page. Structured data helps. Pages must be authoritative to rank well under the Hummingbird model.

When the update was announced, so were new features to Knowledge Graph. People speculated Hummingbird was just a translation of this, but no direct relationship is likely and both probably share a similar philosophy. Knowledge Graph and Answer Card elements are presented on SERPs, but that’s only based on simple correlations and the ability of Google’s updated algorithm to simplify wordy searches.

In terms of the Link Graph, Hummingbird still applies PageRank and link-related components to Google’s algorithm. Good links still contribute to a page’s authority. The context of the link is more important than ever, an aspect introduced with Penguin. Experts still believe the way in which Google handles links needs to be improved. Brand co-occurrences and co-citations have more weight these days, however.

Relationship to the 100% Update

Hummingbird and 100% were introduced nearly simultaneously. Keyword data is only a part of the equation, along with search entities, improved information retrieval, and a more thorough analysis of queries. Users now have to concentrate on optimizing the topical content of their sites. It won’t help to think in terms of synonyms alone. Semantic connections are significant factors.

When optimizing a site, one must now:

• Think about the audience viewing the content.
• Consider SEO best practices such as a clear presentation of information, original information on each page, and providing enough content to satisfy the reader.
• Create a semantic content model to address the connections between ideas.
• Add natural language to the content on each page of a website. Thinking in terms of a landing page will help to tailor content accordingly.
• Diversify the format of content on the same site.
• Make use of tools such as Schema.org, Open Graph, and other semantic markup tools.

It’s also important to see if link building can increase brand visibility, increase referral traffic, create a sense of leadership, and make a site more topically relevant. To benefit from Hummingbird, one should also consider the brand visibility, discussion power, mentions and link content, and traffic from social media.

Considering these helps to refine content and be more creative and engaging to increase site rankings and traffic. These concepts, in contrast, can help identify weaknesses and get on track with the best practices in SEO. Hummingbird helps to steer away from ineffective SEO and take advantage of being authoritative.