Google Hummingbird SEO Tutorial – A Deeper Look Into A Site That Performed Well

In my last post I described what I believe the history of Google’s Hummingbird was, the effects of it, examples of sites that were affected by it and a brief look at what you could do to optimize your content for it.

I’ve received some feedback on that post asking for more details about the exact SEO strategy I used for my example site that performed extremely well as a result of Google’s new algorithm.

But to be clear, I’m going to only go over the ‘Content Related SEO’ that I used when building my site. This post won’t be able building links or internal link structure.

So let’s dig in…

A Brief Review Of A Site That Performed Well In Hummingbird

Let’s look at the traffic chart again:


As you can see from the chart, the traffic to that site began to pick up starting September 13th, but by September 14th a new all time high had been established and kept climbing over the next few weeks.

To go back over the same information shared in my original Hummingbird post, here are some basic facts about the domain:

The domain in the chart has just over 50 pages indexed by Google and is receiving as many as 170 unique views per day as a result of the Hummingbird update.

  • Every page on the domain uses a question as it’s title and in the subheadings of each page.
  • The average document length is roughly 1300 words and as many as 3800 words.
  • The domain is only a few months old and established post-penguin 2.0
  • The only link building done has been a few comments on relevant blogs

Hopefully that gives you most of the basic facts you may need about the site.

But let’s dig in a little deeper yet…

How Keyword Research Was Conducted

I’ve been reading a lot about how Google Hummingbird will ‘kill keyword research‘ and that is laughable at best.

It probably comes from people misunderstanding the importance and roll of keywords with the news that Hummingbird was a move AWAY from ‘Keyword-Driven Search.’

Unfortunately I think that people are confused about what ‘Keyword-Driven Search’ is, so I’ll take a moment to talk about that before digging into how I did my keyword research.

How A Device Affects The Search Query

Search queries change based on the type of devices we use to conduct them. It’s not a good thing or a bad thing – but it is a difference that Google has to adjust for.

Let’s look at two popular examples.

Desktop Search

When you sit down at your desk and open up Google to search for something, you will often use keywords.

cheapbluewidgetsFor instance, if you want to buy a ‘Blue Widget’ you might type ‘BUY BLUE WIDGET’ into the search box. Maybe even ‘CHEAP BLUE WIDGET.’

But you tend to use keywords and not full sentences to conduct the search. The query is typically just 3 or 4 words and then it’s done.

That’s called a ‘Keyword-Driven Query.’ Google had been optimized to serve results based on that criteria in the past because that’s how most searches were conducted.

As time has gone on and smartphones have become more popular, search trends have changed. That required Google to change what type of queries it was able to handle and optimized to serve results for.

‘Conversational’ Search

What do you use your phone for? Play games? Listen to music? Surf the web?

How about having conversations…?

A phone was meant to be used as a means of spoken communication. Because of that, when we are using a phone to conduct searches we will often speak them.

Unless you are a machine, you aren’t going to change the way you talk just to conduct a search. Your speech pattern will likely remain the same as it always has been.

The problem with that for Google is that as a search query becomes longer, the number of results they return will decrease. In addition, the quality of those results degraded as well.

They simply weren’t prepared to serve results for those longtail questions – hence the change to Hummingbird.

A Closer Look

Let’s use the example above.

If you were looking for ‘Blue Widgets’ and using your phone to search, your query might be something along the lines of:

‘Where Can I Buy Blue Widgets?’ or ‘Where can I find the best price on Blue Widgets?’


Do you see the difference between that type of longtail question and the keyword-driven search examples from above?

That’s exactly why Hummingbird was developed and put into place.

So How Does That Affect Keyword Research?

As far as I’m concerned, this change has been  long-time coming.

Over the course of the past 18 months Google has been negatively affecting the ability of marketers to optimize and rank for single keywords.

Penguin damaged the ability for SEO’s to target extremely specific keywords by manipulating anchor text. So that left us using the tools we have available – including the keyword in the title, or subheadings where possible.

Hummingbird is nothing but another evolution in this cycle and again affected how we look at ‘Keywords.’

Now instead of laser targeting one keyword, we have to target the same keyword, but in phrases and questions.

In other words, the longtail search is now the bread and butter of keyword research.

So, with all of that finally covered and put on the table – let’s finally look at keyword research.

Keyword Research Tools

kwtwhatdoidoOnce upon a time, you would put a keyword into Google’s now defunct Keyword Research Tool and hunt down keywords that were 2-5 words long and pound them with links.

NOW – I have found that keyword research done to drive traffic to your site must be a more planned out and organized effort.

That means that tools designed to serve up one short keyword still have their place – but only to serve up seed keywords.

They are no longer the be-all, end-all answer to keyword research.

But not all is lost.

Dig Deep & Help Others While Helping Yourself

To come up with my specific keywords you are going to have to find out what about your keyword people want to know. See what questions they are asking and then turn those questions into document titles and answer them in the content.

If you want to cheat, visit – and popular forums to see what people are asking about. This is where the site: operator comes in handy.

For example, one such search you might use to dig up good questions might be: intitle:”blue widgets”

Find out what people want to know, and then write your content around those questions.

 How To Construct Your Content For Hummingbird

When I was writing my content for the site shown in the chart above, I used a very basic formula – the exact same one I’m about to show you.

This is so incredibly simple that I didn’t really feel the need to include it in my original post on Hummingbird and you will see exactly why I say that.


A simple layout that was used with success in hummingbird


It really is just that simple.

Factors I Used That May Have Nothing To Do With Google’s Hummingbird

When constructing my content, I did some basic things that are just common sense. I can’t say whether or not hummingbird considers them, but they were a part of my game plan so I’ll include them.


I tried to use between 3 and 5 images per post. They were meant to be directly related to the topic and further explain certain points of my answers and explanations.


Again, I would include youtube videos where they made sense. If they helped explain or answer the primary question I would make the video and include it into the post.


Would you trust a research document that didn’t cite it’s sources?

That’s the stance I took with Google. I wanted them to trust my material and therefore linked out to the best sources possible when explaining my answers.

Of course – I was also linking out to my affiliate offers using Cloaked Links.


I avoided big chunks of content where possible and instead opted for small, easily digestible sentences and paragraphs.

My rule of thumb was 2-to-1. Two short sentences for every one long sentence. Sometimes one long sentence would then be followed by a paragraph break.

I feel that it helps avoid reader fatigue & frustration.

Production Schedule

I would normally write 3 articles per day and spent about 2 weeks doing that.

My goal was just to give myself something to test with and therefore I didn’t blow it out into a complete mini-authority site, although I could at any time I wanted to now.

Content Length

The average document was 1300 words long, with some reaching up to 3800 words.

I didn’t aim to produce large documents, but rather it was a side effect of writing the best answers possible.

As with everything above, I have nothing to indicate document size having a relationship to favorable ranking other than the fact the long documents ranked in this example.

SEO For Google Hummingbird Wrap-Up

Hopefully this provides you with all of the essential information you need to understand how I built my site and what factors I considered while doing so.

SEO is simple. You just have to feed the beast.

Think for yourself a little bit and use this only as an example.

You’ve got to think for yourself and think about what is best for your readers. The more you focus on them, the less you have to worry about Google as long as you are giving them what they want at the same time.

Good Luck & Take Care,