What is Google Hummingbird?

Hummingbird was different from the Penguin and Panda algorithm updates. Panda and Penguin were changes to parts of Google's existing algorithm, targeting low-quality content and link spam respectively, while Hummingbird replaced the core algorithm itself, focusing on the meaning of the whole search query rather than the individual words earlier versions matched. This meant Hummingbird could better understand search intent, even when a query used multiple keywords or entirely new terminology.

The practical effect: Hummingbird helped Google return more relevant results on its search engine results pages (SERPs), and it made matching search intent more important to SEO than repeating an exact keyword.

Hummingbird also helped Google return more accurate results for voice searches. Voice queries tend to be longer and more natural-sounding than typed ones, which made them hard for the old keyword-matching engine to handle.

How Hummingbird interpreted queries

Instead of scoring each keyword separately, Hummingbird considered the context of the words together, so pages matching the meaning of a query could do better than pages matching just a few of its words. It weighed synonyms and theme-related results, and it aimed to send you to the most relevant page on a site rather than just the homepage.

Google's launch examples showed the shift. A search for "acid reflux prescription" used to return lists of drugs; under Hummingbird, it returned information about treating acid reflux, which is what the question was really asking. The Guardian's September 2013 coverage noted the same shift for "pizza hut calories per slice," which began pulling the answer directly from Pizza Hut's site.

One thing this page leaves to its sibling: the launch story. Our Google's Hummingbird Update entry covers Hummingbird as an event, including when it rolled out, how Google announced it, and how much of search it touched. This page covers Hummingbird as a system: how it read queries, and what became of it.

Hummingbird vs. Panda and Penguin

Danny Sullivan of Search Engine Land, who documented the launch, compared the algorithm to an engine: Panda and Penguin were changes to parts of the old one, like a new oil filter or an improved pump, while Hummingbird was a brand-new engine that continued to use some of those older parts. The difference for site owners was what happened to rankings:

• Panda (2011): a content-quality filter. Sites with thin or copied content could be demoted by it, then recover after cleaning up.
• Penguin: a link-spam filter with the same dynamic. A site hit by it could clean up its links and recover.
• Hummingbird (2013): a rewrite of how the core engine interpreted queries. It was not a penalty, so there was nothing to be hit by and nothing to recover from.

Hummingbird today: a retired system

Hummingbird no longer runs as a separately named system. Google's guide to Search ranking systems, published in November 2022, lists Hummingbird among systems that are no longer in use or have been incorporated into other systems, alongside the original Panda and Penguin. Its whole-query approach carried forward into Google's later language systems: RankBrain, which relates words to concepts, and BERT, which reads how combinations of words express different meanings.

What Hummingbird means for your SEO

Hummingbird never demoted anyone, so there's no recovery checklist to run. The lesson that has lasted is the one that came with the launch: write in natural language for the intent behind a query instead of forcing an exact keyword into every sentence, cover the topic fully enough to answer the question the searcher actually asked, and let long-tail phrasing occur on its own. Before you publish a page, ask what question it answers, then check that the answer appears in the first paragraph.