Google’s Hummingbird Update: What It Is, What It Means, And How It Affects Your SEO

The web has been all aflutter over the past week with the news of Google’s most recent algorithm update. Known as Hummingbird, it is so named for the lightning-fast, agile little creature that the search giant believes embodies the efficiency and precision Google will achieve in the coming years.

It’s no surprise, either, that this update has generated so much, err, buzz. Google declared the change, which went into effect about a month before the official announcement, to be the most substantial makeover to its search algorithm since 2001. Whereas the recent Panda and Penguin updates added new rules to the existing model, Google is describing Hummingbird as a fundamental reworking of the way it approaches search queries: indeed, it’s projected to affect some 90% of all searches.

We know that those two updates created a whole lot of chatter over what SEO practices are now relics of the past, and which ones are here to stay. So it’s natural to ask yourself: if minor changes to the algorithm made entire batches of SEO tactics obsolete, how in the world are SEOs supposed to adapt to such a supposedly revolutionary change?

The short answer is fairly simple: if you’ve been paying attention to the best practices in SEO, you’re already in the clear. The long answer has a few more pieces to it, so to delve into that, we first need to look at what the update actually does – which, fortunately for us, we’ve been able to learn more or less straight from the horse’s mouth.

What Has Hummingbird Actually Changed?

In short, not all that much. Or at least, much less than one would expect in light of such a dramatic announcement from Google. In actuality, Hummingbird is less about what types of practices help sites rank, and more about the types of queries Google is looking to answer.

The Rise Of Semantics-Driven Search

Google’s earliest incarnations focused on the keyword as the primary indicator of what a given search user was looking for. This method, while far from foolproof, was effective enough at matching the keywords in a query to keywords on a page, and combined with Google’s other ranking factors, it provided a constantly-improving way to link users to the content they wanted to see.

As one would expect, Google has gotten a bit smarter about user intent since its beginnings in the late 90s. One of the defining features of Hummingbird-era search is Google’s focus on context over keyword. Rather than trying to play an advanced game of word association, the search engine evaluates context clues in your entire query – coupled with available information about location, demographic, and the device you’re using – to provide the result that it believes best matches your intent.

Those context clues can be simple or complex, depending on the nature of your query. One of Google’s main objectives in this new approach to search is answering what it calls “conversational” queries – drawing a distinction between the way we’d ask a person for information, and the way we have traditionally asked search engines. Google wants to eliminate this distinction.

Why Is User Intent So Important To Google?

To put it simply, it’s easier to provide accurate results for a given search if you can paint a broader picture of what the user is trying to find – rather than simply associating words from their search with other words on the web. By figuring out a user’s intent, Google can easily eliminate results that may match a query based on keyword density, but which are irrelevant in terms of actual context.

The other part of it is that, as it always has, the web is continuing to evolve at an incredibly rapid pace. Two trends, in particular, are likely behind this transition for Google:
Web Users in 2013 Use Search Differently. As search engines have gotten better and better at returning relevant results, web behavior has changed. It used to be that throwing together disjointed keywords in the search bar would get you about as far as you could go in your quest for relevant content, especially if those keywords were head terms.

These days, people search with much greater confidence that their specific queries will yield results at the top of the first page, and they’re usually right – for which we can all be thankful. (Remember digging through the fourth and fifth pages to find a site you knew was popular? I would hazard a guess that, if you do, you don’t miss it.)

In order to keep up with users’ rapidly changing expectations, and their increasingly specific searches, Google is adapting its process to find the meaning in complex queries so it can continue to show the most relevant results.

Mobile Search Matters More Than Ever. You don’t have to look too far to find crazy statistics on the rapid rise of mobile technology. Bloomberg reports the world will have around 2 billion smartphone users by 2015. Mobile traffic in every industry increases 3.5% each month. And, for the small businesses out there, here’s the big one: mobile local search is projected to exceed desktop local search for each of the next three years.

That has a number of implications for the future of SEO, of course, but for our purposes here, one of the biggest is that complex, long-tail queries, including conversational types, are becoming more and more the norm in search. This is especially true when considering the ascendancy of voice search, an innovation that could only catch on in a world of widespread mobile technology.

The bottom line is simple: more and more people are finding information on the go by asking their mobile devices, the same way they might ask a help desk or a friendly local. That means Google needs to be able to interpret those queries even in the absence of numerous explicit keywords: in other words, by determining user intent through context.

So we have some idea what Hummingbird is trying to accomplish. Now let’s take a look at how that plays out in practice:

What Makes Hummingbird Different?

Hummingbird brings at least two major innovations to Google’s algorithm: the aforementioned focus on conversational search, and the integration of Knowledge Graph to provide results based on relationships, rather than mere association.

A Closer Look at Conversational Search

To understand what this kind of change looks like in practice, it’s best to use examples.

In the screenshots below, I consulted the Googlenets for help in putting together some hypothetical (read: wishful) travel plans for the next couple weeks. Here’s what they showed me:

My search query didn’t include a specific date, nor very particular language about how one might “see” this “phish” I spoke of. But Google knows Phish is a musical artist, with a touring schedule, and so the first three results direct me to their website – specifically, the pages referring to an upcoming fall tour, complete with a list of venues. In other words, exactly what I was looking for.

Of course, there’s a little SEO at play here. The group’s domain name is the same as their real name, and it’s not surprising that Google ranks them first for their own branded term. But compare those results to this more traditional, throw-keywords-at-the-wall style search:

These four results are, for the purposes of my search, much less relevant. Note that, even though the domain is the same for the first three results in either search, the pages displayed, meta-descriptions, and bolded keywords are all very different – and much less helpful for answering my original query.

Conversational search is here to stay – and, for both users and companies alike, that’s a good thing.

The other important characteristic of the approach embodied by Hummingbird is that it focuses considerably on relationships as signs of authority.

Links have always functioned in this way, as direct endorsements from one site to another. But links can be manipulated, and it’s been that way since SEOs first discovered their value in Google’s algorithm.

These days, there are other ways to observe relationships between websites, brands, and even individuals. Google’s ever-growing Knowledge Graph strives to collect and organize this information on a massive scale, using it to provide more and more relevant results to the user by synthesizing what it knows from a vast variety of sources on the web.

In practice, this should mean Google will produce results that have much more to do with its understanding of your keywords and what they mean on the web than whether or not a given result has the most links pointing to it.

In many cases, a query might allude to a topic, event, individual, or organization that Google knows is related to a particular field. Google will then locate the domains, pages, or even individuals that it knows to be authoritative in that field, and prioritize those authorities in the results for that query. It’ll be a while before we see just how much better this type of search works, but the rather brainy folks who dreamed it up seem to be fairly confident it’ll reduce spam and produce better results overall.

But now, the million-dollar question:

What Does It All Mean For SEO?

Let’s look at what we know so far:

• Hummingbird places an emphasis on a semantics-driven approach to search, as Google focuses more and more on user intent and the meaning behind a particular query.
• Context is becoming more and more important in determining results: not just what the keywords themselves mean, but how they form a specific query when strung together
• Authority within a field now influences rankings even more: not just who you are, but who your connections are, and whether those signal that you’re a relevant source within a given field
• The rise of specific, long-tail search queries, especially in mobile search, means that local businesses stand to benefit heavily from Hummingbird, provided they optimize correctly.

We can use these findings to arrive at a few reasonable conclusions about the world of SEO post-Hummingbird:

Relevance matters. This shouldn’t be news, but I still see pages on the web littered with popular keywords in what looks like a hackneyed attempt to somehow rank better in Google – even on sites that clearly have little or nothing to do with those keywords. If Panda and Penguin weren’t enough to dissuade practices like this, Hummingbird should be!

In this era of the web, Google will do everything in its (substantial) power to reward only sites that demonstrate a high degree of relevance to a given subject area – not determined by mere keyword density, but by the context in which those keywords are used, and by the site’s overall relevance according to Google’s own information about it. And in that vein:

Quality content is key. If you already produce content solely related to your industry, at a high level of quality and accessibility, the chances are good that Hummingbird will help and not hurt your search marketing efforts. Quality content is one of the best ways to signal to Google that you are an authoritative source in a given field, and as that type of context-specific authority becomes a more and more important factor in the algorithm, sites with that kind of content will only continue to do better.

Effective brands make connections. Again, nothing revolutionary. But there are many, many agencies out there who will push link building strategies aimed at quantity over quality of links, disregard the importance of social because it rarely passes link value, and encourage content outreach with sites that are popular rather than those that are relevant.

The problem with these strategies, as has been made clear by the last set of updates, is that they are strictly short-term. A link from a low-value directory isn’t the signal of a lasting, credible relationship, and so it its value as an endorsement from Google – as links have always been – can be expected to diminish in coming years. On the other hand, if you consistently seek out relationships, be it via social, targeted link building, or even by earning brand mentions across the web, Google will take note and reward you for them.

So what are your thoughts about the Hummingbird? Has it improved or hurt your positioning on the web? We’d love to hear what you think in the comments below!