If you’ve bought anything online recently, chances are you’ve experienced something like this:

You visit a store’s website to buy a sci-fi movie. The website makes recommendations of other movies for you to purchase. The next day, you get a follow-up email recommending other similar movies, and even some similar books. Intrigued by one of the books on the list, you decide to buy it as well.

All of those recommendations are powered by artificial intelligence.

But even though you’ve probably experienced AI interactions as a customer, the majority of companies aren’t integrating AI across their business. In fact, recent research done by the McKinsey Global Institute found that only 20% of surveyed companies said they had adopted AI and another 41% said they were uncertain of AI’s benefits. The 2018 CMO Survey found similar results with respondents rating the implementation of AI in their marketing toolkit as very low.

However, if we know AI can increase revenue because we ourselves have made purchases based on AI-driven recommendations—not to mention research like Forrester’s 2018 report that found an estimated 30% growth for companies when using AI—then why aren’t we adopting AI in our marketing strategies?

Chances are the lack of adoption is related to AI’s newness and the fact that many AI softwares and implementation techniques are still being developed. Of course, pop culture stereotypes about AI, like Skynet and Terminator probably don’t help.

Harnessing Big Data with Machine Learning and Deep Learning

Before we jump into what AI can do for CMOs, we first need to explore 2 key terms often used with AI: machine learning and deep learning.

Both machine learning and deep learning AI can be used to do everything from producing automatic social media updates, to creating chatbots with the ability to answer nuanced questions, to analyzing big data about cross-channel and cross-device promotions.

Machine Learning Applications

Machine learning is the algorithm which enables a computer to learn desired behaviors without specific directions, like the recommended product emails you get after making an Amazon purchase….