Machine Learning and AI: If Only My Computer Had a Brain Wired for Business

Machine learning and artificial intelligence pave the way for preemptively meeting needs and fixing customer pain points—before customers even realize the pain is there

Providing amazing experiences to customers sounds great, but making those experiences happen has some marketers scratching their heads while trying to figure out how to step up their game. To make the leap from concept to delivery, you’ll need to embrace the emerging technologies of machine learning and artificial intelligence (AI).

Despite being used interchangeably at times, these two technologies are different. Machine learning consists of algorithms that can learn from data without explicit rules. These algorithms operate by building a model to make data-driven predictions or decisions rather than following strictly static program instruction. Artificial intelligence, or AI, is defined as applications that undertake tasks that would usually require human cognition (i.e., logic, language perception/translation, image recognition, etc.).

These technologies take the level of one-on-one interaction a customer could receive under ideal circumstances and make that personalization available at scale and in real time. In essence, machine learning and AI pave the way for preemptively meeting needs and fixing customer pain points — before customers even realize the pain is there. The net effect elevates brand experiences to a level of luxury that earns customer respect and loyalty—and repeat business.

Mirror, mirror, on the wall

The MemoryMirror might sound like something out of Snow White’s story, but it’s actually an innovation designed with AI and machine learning. It isn’t, in fact, really a mirror at all. Instead, the MemoryMirror is a large display with an integrated camera that mimics the functionality of a mirror by showing a customer their reflection. But, it also has a brain wired for business.

“AI is, effectively, a set of complex programming instructions that deliver human-like intelligence.”

MemoryMirror can augment the reality of the viewer by adding information and elements to their reflection that are beyond the scope of real life, including making a customer look younger, removing wrinkles or improving skin tone. The result is a whole new level of customer experience that drives sales by helping customers virtually create the look they want. It has applications for clothes and shoes, hair color and style, makeup, and skin care and dental products, to name a few. The MemoryMirror acts as a virtual “fitting room with an opinion,” providing smart and personalized recommendations to customers based on overall purchase data and comparisons to customers with similar tastes.

The holy grail of personalization

Personalization is the “holy grail” of customer loyalty, but the key is to be able to deliver it in real time and at scale. Doing so requires vast amounts of data and the ability to instantly surface insights from that data. It’s a task that we humans simply aren’t suited for—particularly not in real time. Computers, on the other hand, are perfect for this role.

The concept of AI is fairly simple. It’s an interface that taps into a form of intelligence created to perform a task. AI is, effectively, a set of complex programming instructions that deliver human-like intelligence. Humans have long built machines to perform difficult tasks. Consider the huge machinery in an automotive factory, the giant cranes used in construction, or the massive plants used to generate power. AI is an evolution on that theme as we move further from the manufacturing age and deeper into the intelligence age.

AI helps retailers by serving as an adaptive, automated interface for customer interaction. Similar to a human interaction, this interface can work with customers to resolve issues, route deeper concerns to the right people, and offer personalized recommendations. This is because AI can act on real-time insights supplied from databases that house a user’s browsing history, past purchases, and demographics. Understanding this data opens opportunities for more personalized targeting, and AI can adapt automated approaches in real time to turn shoppers into buyers. For example, global revenue that results from AI technologies just for retail companies is expected to reach $36.8 billion by 2025, growing from $643.7 million in 2016.

And because the technology becomes smarter and more intuitive as it ingests more data, AI also can play a valuable role in automating the content creation process. It offers capabilities for marketers that range from choosing the best image for a campaign or optimizing the content in a creative based on real-time user interactions. For example, from a content creation perspective, this allows the ability to understand the focal—or sellable—point of hero images, and then to auto-crop them for best performance based on an understanding of millions of assets with similar meta-data. In this way, AI enhances creativity and enables a level of responsiveness and efficiency that until very recently was unachievable for marketers.

Machine learning surfaces stellar insights

On its own, AI is a powerful customer-interface tool, but the subset of machine learning takes it to a new level. Machine learning is a series of algorithms that brings data to the AI-application layer—which is where the customer interacts with your content—and makes it possible for AI to update and adapt without needing someone to repeatedly reprogram it. In effect, machine learning enables AI to grow in its intelligence.

Let’s take a standard A/B test, for example. In the past we needed to wait for statistical confidence and someone monitoring the reports until we found the “winner.” With machine learning, we’re now able to let data slowly push more traffic to the “experience” as the data begins to indication which of the multiple experience appears to be the winner—and well before we had statistical confidence when this was run by a straight business rule. Additionally, all the data on one buyer, combined with all the data on all buyers, can come together to identify one customer’s tastes and needs, and then predict their future actions based on the most common next steps that others in a similar situation took—all in real time and at scale.

This underlying ability to access the insights of machine learning means that AI can draw on synergies between everything from customer relationship management (CRM) information to point of sale (POS) data. The challenge of all of this is the volume of data that is being processed, and how intelligent machine learning is today. The algorithms are still evolving. Machine learning develops over time with continual data inputs, so the good news is that it will continue to improve as the data continues to grow.

Beyond the hall of mirrors

While we’ve used the case of MemoryMirror as an illustration of how AI technologies are applied for retailers, we certainly aren’t limited to the fitting room—virtual or otherwise. Retailers also are using AI to track customers.

Fifty-nine percent of fashion retailers in the U.K. are using facial recognition, an AI technology in that it’s a machine undertaking a task that would otherwise require human cognition, to identify V.I.P clients and provide them with special service. The technology also enables retailers to track customer sentiment and gauge how customers respond to in-store displays, how long they spend in the store and traffic flow in each of their retail locations.

But that’s not the only way retailers are taking advantage of facial recognition and its AI technology. They’re using the technology, which is typically employed in airports, for added security. Saks, for example, has leveraged facial recognition technology to match the faces of shoppers caught on security cameras with that of past shoplifters. From this perspective, AI can serve the dual purpose of preventing losses while improving the customer experience—and that ultimately helps retailers boost sales.

Amazon’s Alexa is another popular example of an AI application. The complex AI running the device turns conversation into productivity with more than 10,000 skills ranging from shopping, notifications, virtual assistant capabilities, and more. And now that the machine learning database informing Alexa can share data with the database for Cortana—Microsoft’s data-driven AI—both systems will be that much smarter and give consumers better experiences by drawing on customer history and preferences to deliver personalized experiences easily, especially with virtual personal assistants and social chat.

Machine learning in the future

As with simple machines, these new technologies take over some of the burdensome tasks that people used to perform—like the manual processes involved in campaign execution and creating and updating content—simplifying and expediting those tasks to unheard of levels. The added benefit is that offloading these tasks frees up marketers to look for new strategies to interact with customers, all while allowing personalization at a scale never before possible.

The popularity of voice-enabled personal assistants already demonstrates the power of AI to connect customers with automated “bot personas.” These examples of artificial intelligence are capable of using AI-driven insights about user preferences to make special offers and product recommendations. In some cases, personalized recommendations can literally appear right before your eyes, and that’s a great way to impress customers and win future business, all thanks to the behind-the-scenes work of AI and machine learning.

This post originally appeared in Internet Retailer.

Adobe provides web analytics technology to 253 of the Top 1000 online retailers in North America, according to Internet Retailer’s

Hear more stories from top brands and experience makers, and read more in our #ExperienceBusinessTech series.