What Artificial Intelligence Can Teach Us About Marketing

We’ve seen it a million times, or at least it seems that way: The summer blockbuster set in a dystopian future dominated by machines that have gained consciousness and rendered humankind useless. Likewise, we’ve seen movies where artificial intelligence (AI) did not go the route of world domination; instead, we saw the kinder and gentler side of cognitive computing. Although these fantasies almost always take place in a far-distant future, AI is here now.

In fact, AI has infiltrated our day-to-day lives.


Case in point, consider for a moment these technologies that you may not have previously recognized as AI:

  • Recommendation sections on sites such as YouTube, Facebook and Netflix
  • Common language voice search in smartphones, such as Apple’s Siri and Windows’ Cortana
  • Translation software such as Google Translate
  • ATMs that allow you to deposit handwritten checks
  • Controllerless video games that receive commands through video cameras, such as Kinect for Xbox One
  • Search prediction and completion on search engines such as Google

As marketers, how can we apply AI to improve the way we communicate our brand messages to our target audiences? We need to change how we think about AI and the implications for consumers everywhere.

Whereas consumers tend to think of AI through the lens of Hollywood when, in reality—as the above list demonstrates—AI surrounds us every day. On a very basic level, AI is the science of programming a computer to complete the fundamental and mundane actions that humans perform every day, moving objects from point A to B, for example.

The ultimate goal of AI, according to Robin Hanson of the Future of Humanity Institute at Oxford University, is to create a machine that is as capable and flexible as humans are. In a brief PBS video, Hanson explained the real challenge in AI development is in emulating how the human brain works. While at first AI was thought to be accomplished by programming computers with thousands of rules on how the world around us works, it is now understood that the best way to create AI is to create a machine that is capable of learning in the same way that we do.

The best example of this can be seen in IBM’s supercomputer known as “Watson.” This cognitive computing machine is capable of learning, which is astounding. What is most important about Watson for marketers, though, is understanding how IBM engineers accomplished this feat.

Watson is fed “learning algorithms” that program the machine to complete learning tasks in a specific way. Once an algorithm is set, Watson goes about learning as much as it can under the new parameters by searching the Internet’s treasure trove of information. When Watson is asked a question, it traces back through the material it has read, makes connections throughout the information, and decides on the best response.

AIHere’s what marketers can learn from this process:

  • Authenticity: Since Watson needs to parse through billions of bytes of information, it has to also be able to determine what qualifies as a credible source. Likewise, consumers have begun to learn—on their own and in schools—what credible sources are as well. If marketing campaigns aren’t based on credible material, or seem to be dishonest in one way or another, consumers will turn a blind eye toward it for something more authentic.
  • Familiarity: Watson makes connections in the information it reads. For instance, if a fact continually pops up in source after source, the computer recognizes it as familiar and supposes that the information is accurate. This is also how Watson learns about culture: through understanding familiarities in social media and throughout the Web. Similarly, this is how memes, or culturally significant ideas and behaviors that spread from person to person, are formed. If marketers can identify memes in a culture, they can use those to their advantage and provide content that is familiar to consumers.
  • Accuracy: For Watson to be successful at answering questions accurately, it needs to understand the question completely. Therefore, Watson does an equal amount of work processing questions as it does finding answers. Marketers, like Watson, need to completely understand their audience before they can engage with them.

By breaking down the basics of how we learn, and how we’ve been able to teach computers to learn, we can also understand the basics of successful marketing. While we tend to move away from the basics as we become more successful in marketing, it’s important to see that the basics (as with AI) can have powerful impact when used properly.