Understanding what consumers want and need — ideally, before they even do — is an ongoing imperative for marketers. Artificial intelligence (AI) could make that job much easier, especially with the emergence of deep learning. As technology and AI evolves so must marketers and the strategies they execute. It’s already automating processes to allow marketers to make smarter decisions faster. AI will be able to better anticipate customer needs because the data set it’s relying on is so vast. Your campaigns will be more successful because of this. Human error will be more of a thing of the past as AI continues to evolve and the strategic decisions we make as marketers will be better informed. Let’s take a look at its capabilities and potential uses below.
A subset of AI, deep learning has the potential to transform the future of marketing by helping businesses to predict consumer behavior. It’s a machine learning method that uses layered or “deep” neural networks, similar to those found in biological brains, to learn skills and solve complex problems faster than people can. It helps computers (or robots) handle “human” tasks, such as perceiving objects, recognizing voices, and translating languages.
Deep learning provides a way to train AI to predict outputs, given a set of inputs. Sound easy, but it’s not: While it requires less data preprocessing by humans than conventional machine learning techniques, deep learning requires a large data set and a whole lot of computational power. However, if a deep learning system has access to those key elements, it can learn to predict human behavior pretty accurately.
Consider the “Predictive Vision” experiment. Researchers with MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) trained the deep learning system to predict whether characters in scenes from shows like “The Office” would hug, kiss, shake hands or high-five. After observing more than 600 hours of YouTube videos, the system was able to predict the action 43 percent of the time. According to the CSAIL researchers, existing algorithms could only do that 36 percent of the time.
Another well-known example of deep learning’s ability to predict human behavior involves self-driving cars. Researchers at Cornell University and Stanford University developed a “Brains4Cars” system, which includes cameras, sensors and wearable devices, that monitors a driver’s body language and traffic around the vehicle. The system sounds an alert when the driver appears to be on the fast track to a car accident. The system’s algorithm can anticipate driver behavior about 3.5 seconds in advance.
These experiments have compelling results, but what’s the upshot for marketers? As deep learning technology continues to evolve and improve, businesses can finally put to work all the massive amounts of data that they’re gathering about current, previous, and would-be customers across an array of online and offline channels. And deep learning will become an even more important tool for marketers as the Internet of Things continues to grow, and even more data about consumer behavior is generated and collected from a wide array of devices.
Marrying big data with deep learning will help businesses to create personalized marketing approaches that will appeal to anyone who might buy their product. (To understand this opportunity better, check out this article by Jeremy Fain, CEO and co-founder of neural network technology provider Cognitiv. His hot dog stand analogy lays out how a deep neural network algorithm based on existing customer and other data could help identify future customers.)
Deep learning has “the potential to find patterns inside of patterns” in data to help businesses understand what customers really want, says digital marketing expert Reshu Rathi, in an article about the technology’s role in marketing automation. She explains that deep learning opens the door to hyper-personalization of marketing messages and the customer experience because it takes a customer’s intent into account, and not just their transactional or interaction history. For example, researchers with Renmin University of China found that information about consumers’ hobbies and work situations, when used as inputs for a deep learning method, can help predict the automobile purchase intent and preferences of different groups of consumers.
The ability to predict a customer’s needs, and get it right, is pure gold for marketers. And with the help of well-trained AI, marketers can rely less on assumptions and guesswork and more on data-driven insights to predict customer behavior more accurately — and even well in the future. As Steve King, CEO of Black Swan Data, explains in this video from Deloitte, advances in AI and analytics are shaping a new era of “social prediction.” Marketers can take social data, like customer sentiment data gathered by social media listening tools, to identify patterns that can help them forecast consumer behavior months in advance.
King also recommends that marketers start experimenting with AI, so they can understand these technologies and — most importantly — figure out how they can use them to create value. They don’t have a lot of time to figure that out, either: A new global research study from consulting firm Protiviti and research firm ESI ThoughtLab notes that while “most companies are still at the starting gate” with AI, a “sizable majority of companies are fast-tracking AI applications and expecting to see significant gains in profitability, productivity, revenue, and shareholder value in as little as two years.” The research suggests that most businesses will be applying advanced AI to almost every function, including marketing and customer experience.
Plenty of leading companies are already using advanced AI to deliver more personalized experiences to their customers — and anticipate what they want or need.
Netflix, for instance, has an AI-driven focus on personalization, and its recommendation system influences about 80 percent of the content that its subscribers watch. The company also estimates that its algorithms help it to save $1 billion annually in value from customer retention. And Amazon, which is famous for its product recommendation engine, also uses AI to speed its deliveries by predicting exactly where to stock products so that they are as close as possible to the people who will buy them. (Amazon got a patent for “anticipatory shipping” back in 2013.)
Look for deep learning apps and tools to proliferate soon — and become more accessible, too. Google is so keen on helping to forward the development of deep learning technology that it’s providing cloud software which provides everything you need to get your deep learning project started on Google Cloud. Facebook is also at the forefront of deep learning research and has developed some powerful applications, including a face verification app called DeepFace, which can recognize people in photos with near-perfect accuracy. The social media giant is now working to create a unified deep learning framework that will be accessible to its developer community — and that effort is generating a lot of buzz.
All of this investment and advancement in deep learning will no doubt lead to exciting opportunities for marketers. In fact, thought leaders with McKinsey predict that most of AI’s use cases for business will fall into two areas: supply chain management and manufacturing and marketing and sales. They estimate that use cases in these areas will account for two-thirds of the entire AI opportunity. The AI opportunity for marketing and sales, according to their research, translates to $1.4-$2.6 trillion of value across the world’s businesses. They also estimate that 40 percent of the potential value that can be created by analytics today will come from … you guessed it … AI techniques that fall under the deep learning umbrella.
Deep learning is still in its infancy, but that won’t be for long given the rapid speed of technology evolution today. The ability to predict consumer behavior with deep learning-trained AI — consistently and with high accuracy — is not a far-in-the-future prospect. Deep learning is already changing marketing tactics and techniques, and not keeping pace with that change would be a mistake for any marketer.
AI has a multitude of capabilities. It can assist marketers to be more efficient by streamlining processes like content creation, campaign design, segmentation, etc. (Hope it doesn’t get good enough to start writing blogs, or your author will need to find another line of work!) Because the information it relies on is constantly changing and evolving, AI itself is able to adapt dynamically so the content it is creating is based on the most up to date trends and data. With that, and because there is SO much information available, analysis and insights are readily available to assist with decision making.
It’s doubtful that AI will replace actual human support entirely, but it can be relied on to cover all those repeated mundane requests. The FAQ section of your website or interactive chat bots can be automated to answer the most popularly asked questions about your products and divert that traffic away from your customer support centers. This would allow your customer support to concentrate on the more complicated questions that require more time or resources. You’re now “killing two birds with one stone” while providing excellent support for your customers.
With all its capabilities, the possibilities with AI for marketers could be almost endless! It’s doubtful that AI would replace people entirely. It has no sense of humor, creativity, or the all-important sarcasm; these are all necessary for creating content to captivate your audience. It can, however, create a foundation for marketers to build from.
Manual A/B testing will eventually become a process of the past as AI is capable of gathering data from multiple variables instead of just one and providing results in real-time. We currently have conversational AI like Alexa from Amazon or Siri from Apple that we ask questions to or make requests from, right? Now, imagine a more interactive marketing AI technology that TELLS you, the marketer, what’s trending, results of your initiatives, and suggested adjustments to your current campaigns?!
Even though AI is being used to assist with creating content, the “human touch” is still needed to check for grammar errors, improve SEO, or properly cite references like this. The question you SHOULD be asking yourself though is “Did AI help write this article about AI?” We’ll never tell…
Want to learn how marketers use artificial intelligence to capture insights from their inbound phone calls? Check out our Ultimate Guide to Conversation Intelligence.