Why Google Uses Machine Learning in Search – Here’s Why #176

Here's Why

Eric has created a new machine learning app to help Mark improve his driving. Can anything perform such a miracle?

In this episode of our popular Here’s Why digital marketing video series, Eric Enge explains what benefits Google tries to get from machine learning to improve the quality of search results.

Don’t miss a single episode of Here’s Why. Click the subscribe button below to be notified via email each time a new video is published.

Subscribe to Here’s Why



Mark: So, Eric, machine learning is an area we hear a lot about these days, and people speculate quite a bit about how Google might be using it. So, what are your thoughts on that?

Eric: Hit me with a big topic, why don’t you? Look, seriously, the key insight is that machine learning is just a tool, all right? It’s not something that magically changes the world that we live in. And just because Google has such a tool, it doesn’t mean that their objectives have changed.

We need to think about what their goals are. So, let’s talk about that. The big goals are:

  • improve user satisfaction
  • maintain market share so they can continue to grow as the market grows.
Covid 19
COVID-19: Digital Insights For Enterprise Action

Access Perficient’s latest insights into how you can leverage digital technologies to not only respond to the pandemic, but drive your operations forward and deliver experiences your customers need.

Get Informed

Mark: You know, that’s a great overview, but I know I’ve heard you share way more specific ideas about how they might be actually applying machine learning.

Where Is Google Currently Applying Machine Learning?

Eric: I think there are two major areas they’re investing in right now, based on what we’ve seen from clients and other sites out there.

  1. improving their understanding of a user’s query intent, i.e., when someone enters in a search query, what is it that they’re actually looking for
  2. finding the best content to match that intent.

These are the areas where I see major shifts happening in the results these days.

Mark: Well, can you expand a bit more on the query intent part?

Machine Learning and Query Intent

Eric: Traditional algos were highly reliant on matching keywords on the pages with the query, and that does remain an important part of the process. Think of it as table stakes to be identified as being potentially relevant to a topic.

But the way it’s being done today is a bit different. It looks to me like Google is doing much more to figure out user intent in two areas:

  1. Looking at past query response history to see what has worked for users and what hasn’t. Based on those query histories, they can see what users have responded to most in the results and then adapt the results for new additions of that query going forward.
  2. Generating more data for query intention through testing different types of search results, i.e., trying one kind of search result, seeing how that performs, then trying another one, and doing that through dynamic testing to see what works. I’ve seen tons of evidence in recent shifts in the results that look like they were clearly driven by sampling of user intent.

Mark: So, what about the way they’re using machine learning to find the best content?

Eric: Well, I have to admit, I’m still working in scoping this out better. But what I see is Google rewarding deep content experiences on sites far more than it used to.

So, sites that go extra deep in addressing the user needs with content that covers many aspects of a topic seem to be doing very, very well today. That may be because Google is getting better at understanding the nuances of content, and at the same time, it’s getting better at understanding the nuances of what content matches the user’s query intent. Perhaps this may also be a result of a lot of active testing on their part.

Don’t miss a single episode of Here’s Why. Click the subscribe button below to be notified via email each time a new video is published.

Subscribe to Here’s Why

See all of our Here’s Why Videos | Subscribe to our YouTube Channel

About the Author

Eric Enge leads the Digital Marketing practice for Perficient. He designs studies and produces industry-related research to help prove, debunk, or evolve assumptions about digital marketing practices and their value. Eric is a writer, blogger, researcher, teacher, and keynote speaker and panelist at major industry conferences. Partnering with several other experts, Eric served as the lead author of The Art of SEO.

More from this Author

Thoughts on “Why Google Uses Machine Learning in Search – Here’s Why #176”

  1. You guys are hilarious. You never stop. I love the driving app. Mark, be careful!! I so enjoy coming here for all my bits of news about search.

  2. You guys are hilarious, this is my first time watching your video. I guess the next step from optimizing websites is to ensure that the content is deep and relevant to the keywords the websites are hoping to be found for on Google. Thanks for the article, I’m really interested in how machine learning is going to take off in other sectors and industries outside of Google.

  3. Mark Traphagen

    Glad you both enjoyed and learned from the video, Brad! Be sure to use one of the subscribe options in the post so you’ll be first to see them each week.

  4. It was great to watch the video. A lot to learn and understand about Google and it algorithms. Thanks for sharing. I sure have a lot more to learn about the algorithms of Google

  5. While as I understand, machine learning in this case is just a very, very complicated split test, it’s amazing what it can do.
    I suppose the best thing is the same as before – making quality content that people love. Even then though, it’s nice to know as much as possible how the machine learning is working, both to know how to optimise your site for it, and also we largely need to do the same things it’s doing (understand what people want to try out best to give it to them).

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Subscribe to the Weekly Blog Digest:

Sign Up