Google began as a humble research project out of a garage in 1996. Today, that same garage experiment is on a mission to organize the world’s data. By commanding ~70 percent of the search engine market share, Google has managed to collect immense amounts of data.

Google is quite hush-hush when it comes to how much data they’ve collected on us, the people. Experts have estimated the amount to be 10-15 exabytes. What is an exabyte? Here’s a simple data measurement comparison to help put an exabyte into perspective:

• Bit – Single binary digit
• Byte – 8 Bits
• Kilobyte (KB) – 1,024 Bytes
• Megabyte (MB) – 1,024 Kilobytes
• Gigabytes (GB) – 1,024 Megabytes, your average laptop is 500 GB
• Terabyte (TB) – 1,024 Gigabytes
• Petabyte (PB) – 1,024 Terabytes
• Exabyte (EB) – 1,024 Petabytes, Google’s estimated storage is ~15 EB

To sum it all up in a neat little package, 15 Exabyte is 1.5^10 Gigabyte, or 30 million laptops.

Google’s data collection will only expand in the years to come, and with it their Search Engine Optimization (SEO) must mature as well. To manage the big data influx efficiently, Google has moved away from using only basic keywords in SEO.

Now, don’t get me wrong, keywords are still important. But now more than ever, quality content on your site is driving Google’s search priorities your way. How so? Well, in September 2013 Google celebrated their 15-year anniversary with a new search algorithm named Hummingbird.

 What is Google Hummingbird?

Hummingbird is the name given to Google’s latest algorithm change. The name comes from the new algorithm’s ability to move quickly and accurately, much like a Hummingbird. More importantly, along with its speed, the algorithm gives greater prominence to Natural Language searches.

Natural language searches consider context over individual keywords. Context is increasingly important as conversational searches become the norm as people search more often with the Mic (speaking) function or by asking questions in a search engine instead of searching with direct keywords.

An example of this would be my latest conversational search using the Google Mic.

What I said: “Where can I buy pipes for my motorcycle?”

A traditional keyword search would target “Pipes + Motorcycle,” providing me with hundreds of results for both pipes and motorcycles.

The new Hummingbird algorithm would recognize this as something now called a ‘Long-tail keyword’ and will search for a site containing relevant content, providing me with options on where I can ‘buy pipes that are for a motorcycle.’

With long-tail keywords, I can get even more descriptive to narrow my search, such as “Where can I buy used pipers for my Suzuki motorcycle?” Now Hummingbird will search for a smaller target of relevant content, placing the top results in my search with the sites having content that matches my natural language (conversational) search.

This could be a blog post about fitting used pipes to a Suzuki, which would match me to the site I need.

How does this affect my SEO?

If you’ve been filling your homepage with relevant keywords to your site, that is still helpful. But now, as Hummingbird searches for related content quality and quantity, it is important your site has a collection of related information on your product or service.

Blogs are a fantastic way to increase your site’s content, create backlinks and internal links, and to position yourself as an expert in your field.

So, boost that content! Either start writing or hire a freelancer to build content for you.

Beyond Keywords

Hummingbird is not only a new way to quickly scan enormous amounts of data. Google is doing more with its big data than scanning it to help your searches. They are using machine learning combined with natural language search data to build an AI assistant called Google Duplex.

Google Duplex is in its prototype stages and is already making simple test calls as an assistant. Right now, it is only booking restaurants and haircuts, but the potential is incredible. The reservations I have would be that our data is being used to study us, and in turn teaching a machine to speak like us. All in an effort to sound natural and actually trick us into thinking we are speaking to a human.

I would like to see some robotic influence in this AI assistant or something to alert the person on the call that they are speaking to a machine. It seems unfair to not know you are speaking to a computer, but I suppose we will all need to adjust with the future.