Artificial Intelligence in Retail - Untitled Project

AI in Marketing – Personalized Shopping Experiences, at Home or Offline

Using technology, data, and analytics to personalize shopping experiences holds great marketing potential, yet it is still an intimidating concept for many retailers. According to a recent McKinsey survey, only 15% of senior marketing leaders think their company is on the right track with personalization. Yet techniques like singular channel product recommendations and triggered communications have been found to drive 5 to 15% increases in revenue and 10 to 30% increases in marketing-spend efficiency. Three major shifts in the development of personalization should help marketers adopt AI for marketing:

Digitization of Physical Spaces

Under 10% of the companies surveyed by McKinsey currently use personalization in their “offline” experience. Yet, 44% of CMOs announce that advanced analytics can provide key insights to allow employees to offer a personalized offering. Similarly, 40% think that personal shoppers could improve service by using AI-enabled tools, and 37% mention the rise in the use of tools such as facial recognition, location recognition, and biometric sensors.

While personalization practices are still rudimentary, some retailers are taking steps to embrace the future of in-person shopping. Covergirl’s new flagship store, aided by Google’s conversational platform Dialogflow, now guides customers through an AI program. At the same time, augmented-reality technology allows them to virtually “try-on” products—reflecting pictures of themselves as if the products have been applied. Going further, AI platforms will be able to generate recommendations based on customers’ skin tone, features, and product response. Of course, these virtual experiences still need a human touch, with stylists offering their own advice. But the potential of in-store personalization is boundless, as AI keeps developing and recalibrating customer experience (think trying hiking boots on a “virtual mountain” for example).

Scaling Empathy

Emotions, and the ability to understand and relate to them, is still the basis of relationships, especially with branded content. There is still work to be done to understand emotional cues and respond to them, both digitally and at scale. Machine learning is working towards this need to read and react to social cues, with ever-more complex algorithms now being able to extrapolate emotions from visual and auditory data. Among others, Amazon’s new Echo device could detect when someone is ill by analyzing nasal tones and offer personalized recommendations like medicine and recipes, which could then be purchased on the platform and delivered at home. Similarly, MIT-born Affectiva will be able to decipher emotions such as joy, fear, anger, or disgust from facial expressions based on emotion-recognition algorithms. Soon, marketers will be able to communicate with customers depending on their own moods, for example, curating particular product recommendations based on the emotions they can detect from their users.

End-to-End Personalization Ecosystems

At the moment, there is a multitude of providers involved in a customer’s shopping experience, from the mall operator to the brand product itself. Personalization would allow marketers to create partner ecosystems to provide a consistently branded decision journey. With AI predicting consumer needs, personalization programs could allow for seamless transitions across journeys (think turning up the house heating when the car departs from the customer’s workplace). Predictions indicate that the share of global sales moving through the ecosystems should grow from less than 10% to nearly 30% by 2025. There is particular potential within the home in making devices work together. Consumer goods, automobiles, and smart home devices will soon need to provide a seamless experience or risk being completely shut out from the line of consumption.