The majestic hummingbird – a strange, nectareating bird captured all the SEO buzz last week.

That noise you hear in the distance is the sound of SEOs polishing the keys on their keyboards, trying to answer a simple question.

“What is Google Hummingbird?”

And they’re making big claims about knowing the answer. If you don’t believe me, click on one of the links at the top of this paragraph. Even though there’s enough Hummingbird content to keep you reading well into the night, you shouldn’t bother.

Why not?

Because it’s noise

Rand Fishkin says hummingbird is about longtail.

Danny Sullivan says hummingbird is about conversational search.

Ammon Johns and Bill Slawski say long-tail optimization is gone, and big brands may have been gifted more leverage than ever. 

3 experts. 3 Answers. Let’s cut the BS. What is Google’s Hummingbird?


What do we actually know about Hummingbird?

Does the SEO industry have any idea what’s going on?

I don’t think so.

What I’ve seen out of the SEO industry since Hummingbird was announced has let me down. I’m sure I’m not the only one.  There’s no lack of content on the topic. But most of it is conjecture and outright guessing without anything even beginning to resemble a quantifiable hypothesis.

Here’s a revelation: No one has any clue as to what Hummingbird actually is!

Now that we have that out of the way, let’s walk through what we DO know about hummingbird. After we do that, I’ll treat you to my hypothesis and what I think it means for SEO agencies and businesses trying to stay ahead of the curve.

Quotes from Google (and others) on Hummingbird:

– A word cloud of all of the google hummingbird coverage this week.

“Shortly, there will be new ways to get answers from the Knowledge Graph.”

“Second, conversations with Google will get even more natural.”

“Google is working on a new design language to help with this.”

Danny Sullivan relates a story from a Google event he attended:

“[A searcher] asks Google to show pictures of the Eiffel Tower. Now she asks “How tall is it” then after that “when was it built” and gets answers based on Google knowing what “it” is.

She goes on,”Show me pictures of the construction.”

Hummingbird helps with complex queries but also impacts over 90% of searches worldwide now.

“It’s not easy to build a new algorithm that’s so good”

Q: How’s [the algorithm] different? A: People are asking more complicated questions.

So how [does Google] keep results relevant in light of this. Hummingbird impacts all types of queries, but it is far more effective on these long queries we get now.

Gave us an opportunity, Hummingbird did, to take synonyms and knowledge graph and other things Google has been doing to understand meaning to rethink how we can use the power of all these things to combined meaning and predict how to match your query to the document i terms of what the query is really wanting and are the connections available in the documents. and not just random coincidence that could be the case in early search engines.”

Richard Taylor has more: 

“At a presentation on Thursday, the search giant was short on specifics but said Hummingbird is especially useful for longer and more complex queries. Google stressed that a new algorithm is important as users expect more natural and conversational interactions with a search engine – for example, using their voice to speak requests into mobile phones, smart watches and other wearable technology.

Hummingbird is focused more on ranking information based on a more intelligent understanding of search requests, unlike its predecessor, Caffeine, which was targeted at better indexing of websites.   It is more capable of understanding concepts and the relationships between them rather than simply words, which leads to more fluid interactions. In that sense, it is an extension of Google’s “Knowledge Graph” concept introduced last year aimed at making interactions more human.

In one example, shown at the presentation, a Google executive showed off a voice search through her mobile phone, asking for pictures of the Eiffel Tower. After the pictures appeared, she then asked how tall it was. After Google correctly spoke back the correct answer, she then asked “show me pictures of the construction” – at which point a list of images appeared.”

Last Hummingbird quote, from Gerry Brown:

“Although Google has not given away many details, it said that Hummingbird is focused on ranking information based on a more intelligent understanding of search requests. As Internet data volumes explode we increasingly have to type more and more words into Google Search to gain greater accuracy of results. Often we need to conduct multiple searches to find the information we are looking for, which is frustrating and time consuming. This is because the Search results we currently receive reflect the matching combination of key words that a search phrase contains, rather than the true meaning of the sentence itself.

Search results produced by Hummingbird will reflect the full semantic meaning of longer search phrases, and should in theory produce more accurate results.

For example Hummingbird will more greatly consider question words like “how” “why”, “where” and “when” in search phrases, in addition to content keywords. Hence Hummingbird moves the emphasis of search from “results” to “answers”. Google also has acknowledged that the number of mobile and voice-based searches is increasing. Such voice searches are in natural language, and may not therefore contain the keywords we might finesse on a computer keyboard. These ‘on the fly’ searches are likely to return poor results using a keyword search system.

The semantic search capabilities of Hummingbird aim to address this need. It should be noted however that the most-used medium for mobile voice-based search is Apple iPhone’s Siri, which uses Yelp and WolframAlpha rather than Google for semantic search. WolframAlpha has had a semantic search capability since 2012, so there is undoubtedly a competitive response angle to the Hummingbird move.

The future is therefore “conversational search” or “hot wording” as Google refers to it. By this Google means that a user can simply voice prompt the Google search engine by saying “OK, Google”. The latter is also the voice catch-phrase used to operate the wearable Google Glass spectacles.”

My hypothesis after gathering the facts is…

I bolded a few keywords in the quotes because they’re the meat of the technical explanation we have for Hummingbird so far.

Google (and others) are giving us a definition of Hummingbird:

Hummingbird is “semantic understanding of queries by connecting the algorithm to Knowledge Graph.”

Wait, what?! What’s a knowledge graph?

Don’t read SEO blogs for the answer. Everything you need to know is right here, provided by Google: knowledge_graph2

Google is building a database of relationships. They want to know about people, places and things. Google wants to know how everyone and everywhere relates to everything else. That’s knowledge graph.

Knowledge graph is what Google sees when it shakes the magic 8 ball after asking “What’s the future of search?

What does it all mean for webspam?

It’s a lot easier to detect a blackhat lead gen site for free credit reports when the website isn’t connected to a company that has natural connections to the credit reporting industry. Knowledge graph makes it easier for Google to see who is connected to the credit industry’s most important companies, i.e. Equifax, TransUnion, and Experian.

If Google knows your company isn’t connected to the credit reporting industry’s most important companies, they won’t return your website for the phrase “Free Credit Report” no matter how many other signals they have that say you should be there (i.e. link spam and page relevancy). It’s the start of a brave new world, folks.

It makes a lot of sense to incorporate knowledge graph into the core algorithm to reduce spam. But it’s great for other things, too. Things like, you guessed it, “conversational search” i.e. “the semantic understanding of queries.”


Because Google can use knowledge graph data to give you more topically relevant results, based on the relationships between the people, places, and things you’re searching for. Language is a web of relationships and interactions that connects stored bits of understanding. If I say “Google” you think “search engine”. Then you think “Internet,” followed by “computer” or “microchip” or “circuit board” until you run out of words you associate with “Google.”

Everyone uses their core understanding of a concept to determine what words and ideas are associated with it. Once you realize that all of those connections drive how you use and understand words, knowledge graph starts to look a lot like the same process you use to understand language.

Pretty cool, isn’t it? When’s this thing going to start thinking for itself? Anyone seeing Skynet (aka Google) becoming self aware yet?

What am I doing from here?

Nothing yet.

Smart businesses will position themselves for the future, using what they know about Google’s trajectory. This is one of those times when you need to step back and evaluate what is about to happen and make appropriate changes to safeguard your success.

The “takeaway”

Hummingbird’s infrastructure involves knowledge graph architecture.

Due to the complexity of calculations related to knowledge graph, effects will be intermittent or Panda and Penguin-like in terms of their update freshness. In other words, updates will be very complex and look at the web in ways unfamiliar to you if you are say, still operating like it’s 2010. Believe it or not, there are a lot of companies and SEOs thinking with an outdated and dangerous mindset. It’s time for them to step their game up, or they may not have much Google traffic in 2014.

  • Google wants to make authorship more important as they work to expand what knowledge graph knows about people. If you’re not already using Google authorship, now’s the time to start learning more about it.
  • Google will always want to eliminate spam, but they’ll take it to the next level by detecting who and what companies are really ‘in’ specific industries and who is faking it. In other words, Google will soon be able to tell if you’re lying about being one of the “cool kids.”
  • Your business’s NAP (name, address, and phone number) will become even more important. Make sure you’re consistent across your social and business profiles.
  • Start interacting in ways real businesses do by participating in RCS (Real Company Shit) instead of playing on the Internet.
  • As a side effect, they’ll be able to associate the people, places, and things in your searches to their relationships through knowledge graph.

Google is one step closer to understanding, in context, what it is you’re really asking about when you perform a search.

Let us know what you think below. Do you buy into Hummingbird being all about Knowledge Graph?