May 26, 2021
If you haven’t seen the writing on the wall, here’s the gist: AI will have a profound impact on marketing, especially in predicting customer behavior and market shifts and enabling hyper-relevant personalization across channels at scale. It’s best to get ahead of this once-in-a-lifetime marketing disruption before competitors leave you in the dust.
“There’s a potentially powerful upside in the application of machine learning in marketing, yet there’s been slow adoption among marketers,” says Jim Lecinski, clinical associate professor of marketing at Northwestern University’s Kellogg School of Management and co-author of The AI Marketing Canvas. “Half or less are applying machine learning in any meaningful way.”
Through thought-leadership blogs and research, the CMO Council will stay abreast of AI’s emergence and impact on marketing. But if you feel you’re a step behind the AI learning curve, we highly recommend The AI Marketing Canvas to get up to speed. While the book geeks out a bit in the fields of data science and statistics, nodes and networks, it’s chock-full of real-world case studies showing AI in action. (For more, check out our blog post, Are You Mastering AI? Or Will AI Master You?)
Today, we’re highlighting some basic AI concepts as described in The AI Marketing Canvas and our interview with Lecinski. We cannot stress enough that this is just a run-through. If you’re really interested in learning more, you’ll have to read the book.
Let’s start with AI in marketing, which is really about using underlying machine learning to make predictions. It’s not about native language processing, robotics and other forms of AI. Machine learning makes predictions based on data, learns from successes and failures, and improves its predictions on the next go-around. Marketers already intuitively predict customer behavior, but machine learning brings speed, scale and accuracy to the process.
Here’s how The AI Marketing Canvas defines it: “Machine learning occurs when a data set is loaded into a machine which then processes it by applying one or more computer algorithms to arrive at a series of predictions.”
So, for our purposes, AI and machine learning equal predictions.
Next up, artificial neural networks. Don’t get scared off by the sci-fi term. This is basically a set of algorithms modeled after the human brain and designed to recognize patterns in data, such as images, according to The AI Marketing Canvas. It helps marketers cluster and classify data.
Not only does this have tremendous marketing potential, it’s already being used to great effect. Just ask Airbnb, which was able to classify images of places to stay — no small feat — and then present photos that would appeal to specific customers based on their profiles and feedback.
Machine learning and neural networks often work in concert with deep learning. This is the phase where AI develops its own kind of intuition. The AI Marketing Canvas describes deep learning as the application of machine learning’s neural networks to complex problems, allowing the machine to learn from its mistakes and to assess its own probability of reaching a correct result.
A prime example of this is IBM Watson, the machine made famous by beating a couple of Jeopardy!’s greatest champions. According to The AI Marketing Canvas, Watson’s machine learning used algorithms to analyze ways a Jeopardy! question could be interpreted and then searched internal data for plausible answers. Watson used a second set of algorithms — that is, deep learning — to find additional evidence in order to rank these answers and achieve a high level of confidence in the final response.
The cornerstone characteristic of AI in marketing is that machine learning trains on the latest data, even real-time data, to constantly improve its answers and predictions. The system isn’t a hard-coded, rules-based program relying solely on historical data to make predictions. This is an important distinction, because historical data isn’t very useful after a black swan event, such as the pandemic, renders past behavior obsolete.
“There has been a huge reset on the data applicable for AI for all the brands,” Lecinski says. “This is an opportunity for brands starting their AI journey to catch up. ... But without some guidance, brands will either remain paralyzed and do nothing or aim too high and do too much involving tactics only, which can lead to catastrophic failure.”
Tom Kaneshige is the Chief Content Officer at the CMO Council. He creates all forms of digital thought leadership content that helps growth and revenue officers, line of business leaders, and chief marketers succeed in their rapidly evolving roles. You can reach him at email@example.com.
No comments yet.