ReFUEL4: The Science Behind Knowing What Women (and Men) Want

The Science Behind Knowing What Women (and Men) Want

Imagine you’re at a shopping mall, looking for a new pair of running shoes. You don’t find what you’re looking for, so you move on to look at sweatpants. A few minutes later, a salesperson approaches you, holding a newly-released running shoe in your size and favorite color. He asks if you’d like to try it on. Based on the browsing behavior of other similar shoppers, the salesperson gathered some important information about your needs and preferences, and used it to offer up what you were looking for. This is a little like what predictive analytics is like – except that online, marketers are able to gather much richer data about customers to better inform the tactics they use to connect with people.


Predictive Analysis: A Digital Body Language


Predictive analytics is a concept that has been around for quite some time. From the earliest days of advertising, marketers have tried to use past performance to inform future decisions. Today, you can think of predictive analytics as a kind of digital body language. The introduction of technology and data means that predictive analytics helps us find people who are the right fit for your brand, but who haven’t necessarily shown behavioral signals that would allow you to easily target them.


AI: Building an Artificial Brain


At the core of predictive analytics is artificial intelligence. You’ve probably heard of artificial intelligence, or AI – but many people still don’t know exactly what it means. It helps to start with our brains: the most complex, intelligent, and dynamic organs that have ever existed. Your brain can learn and apply knowledge – like when you touch a hot stove and learn that the result is the searing, stinging pain of a burn. Collectively, our brains have developed art, language, government, technology. There’s no question that they’re impressive machines.


Throughout history, man has been trying to reproduce our brain’s amazing intelligence. We’ve taught machines to think and act like us, and we’ve created responsive expressions of this type of technology – like Siri, who can engage in two-way conversations while exhibiting some semblance of human personality and character in some of its rather philosophical or humorous responses.


Building an artificial brain might sound simple. It might even look simple, thanks to Hollywood. But instead of building a brain that acts as the core of an AI system, scientists and technologists looked deeper – what makes the brain so powerful? They found it’s actually a network of neurons, and the connections between them, that allow our brains to complete complex tasks. AI is modeled similarly, as an enormous network of interconnections that can learn complicated tasks.


AI in Marketing


As machines become more intelligent, they’re able to observe, and through many samples over time, predict future scenarios. As it relates to marketing, this means that a machine can begin to predict a user’s behavior, response, or preference before a campaign is even launched. The enormous amount of both real-time and historical data available about customer behavior means that marketers are using predictive technology to anticipate customer needs instead of just responding to them. The result is an amazingly personalized shopping experience – and a certain path to an improved brand experience and an increase in sales.


Deep Learning: Making Sense of Things


Perhaps the most impressive thing about the predictive analysis model is that it can perform what’s called “deep learning.” Think about a baby. At 3 months old, a baby is exploring its environment and sees a hand holding its bottle, or a shoulder where it rests its head. But at 9 months old, a baby understands that the hand and shoulder are both a part of an entire person – and an important person, at that! Deep learning is the ability to map pieces and functions to higher level concepts and compositions.  In practice, this means that predictive analysis might also help a bottle brand serve ads to woman who has just started browsing maternity clothing.


Making Ads More Intelligent


Predictive analysis is one of the most exciting tools in advertising, because it provides more relevant and personalized experiences for the consumer, and increases engagement and sales for the brand. At ReFUEL4, we’re excited about predictive analysis because it lets us combine insights from our designers’ past work, target audiences, industries, and the relationships between text, images, and video to deliver data-driven ad designs that work for our clients.


HanSheng Chiu

ReFUEL4's resident Data Scientist and AI enthusiast. Hoping to someday create an artificial clone of himself for handling daily affairs so he can devote more time to pursuing natural passions like food, nature, charity and humanity.