The following post is part of WTWH Media Marketing Lab’s ongoing blog series from MozCon 2018.
On the last day of MozCon 2018, Britney Muller, senior SEO scientist at Moz presented one of the most interesting and endearing talks of the conference. Muller’s discussion helped define machine learning, and how we can apply it to SEO. Muller explained how this technology can solve many SEO problems and even write meta descriptions for you. She also explained the common models of machine learning, how it is being used currently, and provided resources on how to get started.
What is Machine Learning?
Machine learning is a subset of A.I. that combines statistics and programming to give computers the ability to “learn” without explicitly being programmed. This technology combines statistics and programming and has the ability to pull data automatically. However, a machine learning model is only as good as it’s training data. As Muller explained, if machine learning was a car, data would be the fuel.
How can you apply Machine Learning to SEO?
Marketers and SEO professionals don’t realize that machine learning is already affecting the work that we’re doing. The first glimpse for SEO professionals is keyword research for voice. More consumers are using smart speakers to ask questions and search by using voice commands. Dialogflow’s Small Talk tool is the first computer interaction technology for keyword research using voice search. This tool will help SEO professionals optimize for voice search and understand the most common questions or searches on voice systems.
SEO pros can also use recommendation models (such as what Netflix does) and apply them to real visitors on your site. Recommendation models in real-time will help you understand what your users are searching for, what pages they have already visited, and what content they don’t need to see again. When you understand what your users viewed when they last visited your site, you can recommend new content and help visitors get more value from your site.
Machine learning can also assist your SEO efforts by automating meta descriptions. Algorithmia, an advanced content summarizer can help you create auto-generated meta descriptions. SEO pros provide a URL and Algorithmia will pull text from the URL and generate a meta description. Imagine what you can do with large sites when your meta descriptions are automated! Machine learning for SEO is just getting started and many opportunities are beginning to arise. Marketers will soon be able to use machine learning to find ranking opportunities, optimize title tags, find keyword gaps, create client reports, find common questions, create content and more.
How can you build your first Machine Learning model?
The hardest part of building your machine learning model is pulling enough data, cleaning it, and preparing it to learn. There are lines of code available on CodeLabs, Google’s machine learning crash course. To begin, collect and clean your dataset, build your model, train your model, evaluate and predict.
Machine learning will help scale SEO tasks and allow us to evolve with higher level thinking as SEO professionals. By embracing the upcoming opportunities that machine learning provides, this will only help us grow as marketers and position our content to rank on Google. We are only going to see more of machine learning in SEO, so why not get ahead of the game?