Google's Hummingbird Update

September 2013

POV

Google’s Hummingbird Update

The Hummingbird Rollout

On September 26, Google revealed the existence and prior rollout of a new algorithm approach, which it code-named “Hummingbird.” The Hummingbird update was specifically targeted at better understanding and breaking down very long search phrases and questions in order to arrive at results that better match the intent of questions.

In the same vein of Google’s Knowledge Graph updates and move toward semantic signals, the key concept of the Hummingbird update is user intent. While the Penguin and Panda updates adjusted a pre-existing formula to knock out low-quality content and links, Hummingbird appears to be an update to the underlying engine—along the lines of Google's increased ability to map synonyms over time. In this way, Hummingbird is similar to Google’s Caffeine update, which made indexing faster and put a higher premium on pages that are more recent. The Hummingbird algorithm will still include many old factors (such as PageRank) but will be more effective at including newer items like schema.org semantic markup.

Though Hummingbird applies to a much wider body of keywords than Penguin and Panda, this shift that affected 90% of search queries (by Google’s estimation) happened in a relatively unnoticed fashion versus the tumult surrounding Penguin and Panda. This is likely because the traffic affected by Hummingbird’s improvements, while making up a huge percentage of the unique search phrases used by Google search visitors on a daily basis, constitutes very low volume compared to the shorter phrases targeted, monitored, and seen by most brands and site owners. Hummingbird affects the long-tail of the long-tail.

Where Google is Heading

Not only do these very long search phrases collectively (if not individually) drive a significant portion of searches, but given Android's continued expansion in mobile generally and voice commands specifically, these kinds of phrases are a strong match to those you’d expect to arrive at when asking your phone to help find something—potentially enhancing the value and quality of its mobile search

results.

By increasing drive toward semantic search and actively encouraging sites to include semantic markup such as schema.org to their content, Hummingbird represents another step by Google toward attempting to correctly arrive at user intent even more frequently. This makes the richness and depth of the content on your site all the more important, as does filling in as many contextual clues via keyword inclusion and semantic markup as possible.

The ability to better parse long queries also aligns with improving the experience for mobile users, who are increasingly turning to voice search and voice commands on Google’s growing Android platform and utilizing voice searches on Google across all devices. Smaller screens in the mobile world also make delivering information faster and more direct—such as those in Google’s Knowledge Graph (also fueled in part by semantic markup). By improving the ability to arrive at a quick answer to a long question via Hummingbird, Google is equipping itself to better serve both mobile users in general and voice search users in particular.

This also aligns well with some of the innovations and approaches taken within Google Glass’s platform, by which a user may start with a single small phrase (a person’s name, for example) and carry the context of that search with them into their next question (‘when was he born?’) without having to explicitly say the person’s name again. Search as a conversation, being a natural extension of a more semantic approach, would be well served by improved abilities to make sense of longer phrases.

Impact & Considerations

Based on the areas and phrases affected by Hummingbird, iProspect recommends reviewing the following areas of your SEO and content strategies:

Increase the detail of your article and product-level content. As Google works harder to determine intent and meaning behind search phrases, a high level of detail in your content—especially for pages targeted toward longer-tail terms—opens increased opportunities to rank for a wider range of terms

and to provide Google with content that can be matched to a wider range of search phrases.

Expand your content. Developing deeper catalogs of meaningful content can also position your brand to be the answer to a wider range of queries. Integrating user-generated content (UGC) modules (whether adding reviews to existing pages or allowing wider posting by your visitors on forums and image/video collections) can not only help enhance the reach of your current pages but more quickly expand your content catalog—while keeping your content fresh.

Include schema.org markup on your pages and templates. The drive toward semantic search is definitely not going away. Hummingbird comes on the heels of Google’s efforts to encourage sites to include semantic markup—specifically Schema.org markup—in their content to more thoroughly identify data points and create connections to wider concepts. If you aren't adding semantic markup to our sites, or making use of Google Webmaster Tools' data highlighter as a temporary alternative, you're a step behind those who are, and the gap is growing fast.

Pose content as answers to questions. Long-tail queries and question-based informational searches make up a tremendous amount of search volume collectively. While many brands focus their efforts on searches that are more closely tied to direct responses, answering questions that shoppers and your audience are asking—even those that may be only tangentially related to what you ultimately want them to consider buying—can act as a huge draw for visitors and brand exposure. This in turn builds a larger audience for your brand and offerings over time, creating a long-term business case for investing in developing content related to query-based searches, long-tail concepts, and a wider content base that looks beyond immediate direct response.

Optimize for the mobile user. In addition to supporting semantic search, the ability to better parse voice commands and questions should add further attention to optimizing your site for mobile search (via Google's recommended base platform of a responsive design website) and to developing content that can attract and enable mobile-specific searchers (who tend to use phrases that are more localized in nature).