receives compensation from some of the companies listed on this page. Advertising Disclosure


How AI Will Transform the Retail Customer Experience

Tom Treanor
Jan 19, 2021

Among other things, retailers will have to meld an ever-growing variety of data and data sources into a single comprehensive, accurate real-time view of every customer.

If you’ve tried to crack a Rubik’s Cube, you know how tricky it can be. In the past, it was a challenge only humans could learn to solve. No longer. A computer at the University of California, Irvine, figured out all the right moves in just over a second.

The machine wasn’t specifically programmed to decipher the cube. Instead, it taught itself to do so, using an artificial neural net to reason about the challenge and make its own decisions.

For retailers, intelligent machines solve a different kind of puzzle

Retailers, like virtually everyone else, are trying to understand what such advances in artificial intelligence (AI) mean for the future. Machine learning (ML) and other applications of AI have already begun to change how major retail companies understand and interact with their customers. And these efforts are picking up steam as high-tech competitors such as Amazon force more and more retail firms to rethink the customer experience.  

With further developments in AI, retailers will gain the ability to provide more effortless, more satisfying outcomes at every phase in a customer’s journey, from product discovery to customer service. Advanced ML and data analytics will be essential to powering such experiences at scale.

Here’s a look at how the future is taking shape and what is likely to happen next.

AI will take personalization to the next level.

Thanks to Spotify and Netflix, customers have come to expect digital experiences tailored to their individual tastes, wants and needs. With AI, retailers dealing in physical products also have a growing ability to deliver such one-to-one personalization at scale.

The hyper-personalized future of retail is already emerging, especially in sectors such as fashion and cosmetics. Stitch Fix, for instance, employs ML algorithms and data analytics to learn shoppers’ tastes in clothing through personal style profiles they fill out online. Stitch Fix’s system then provides recommendations to help human stylists make personalized choices for customers while using feedback from the stylists to improve its own suggestions over time. Similarly, Shiseido used ML and Treasure Data to model preferences and personalize its loyalty program – for example, by recommending different beauty products to customers as they move through different stages of their lives.

Future AI will only become more adept at discerning patterns in unstructured data such as images and video, unearthing individual preferences and making real-time decisions based on these calculations. And it will need less and less human assistance to do so. Today, automated product recommendations often feel predictable, but tomorrow’s AI, using techniques such as deep learning, could supply that missing note of serendipity.   

In-store shopping, meanwhile, will catch up to online commerce when it comes to providing personalized customer journeys.

Eventually, an AI system may recognize a preferred customer as soon as he or she walks through the door. The system could then send an alert to an in-store human representative, along with the customer’s up-to-date individual profile. Such a profile could include style preferences and customized recommendations, drawn from real-time data on everything from past shopping behavior to the weather outside.

Predictive retail will give customers what they want, as soon as they want it – or sooner. 

Predictive analytics powered by ML is already enabling many companies to foresee customers’ actions and make proactive decisions. A customer data platform (CDP), for example, uses predictive AI built on Apache Hivemall to score customers by criteria such as churn affinity or upsell potential. Marketers can then use these scores to target customers with specialized campaigns.

Here, retailers can take inspiration from what companies are doing in industries beyond their own. A leading game maker, for instance, recently used ML, predictive analytics and data unification with Treasure Data’s CDP to predict which types of in-game rewards were most likely to keep users playing.

Imagine a store chain trying to prevent its most valuable customers from leaving. AI-based predictive scoring could identify and flag the customers most liable to stop buying. Predictive AI could also determine exactly which offers or interactions were most likely to pull at-risk customers back from the brink and learn how to predict such outcomes more effectively over time.

Alternatively, AI-based analytics might be able to predict which traits and sequences of behaviors suggest a customer is almost ready to buy. When the system identified a customer with the right profile, it could then take the action most likely to lead to a sale – for example, sending out a personalized invitation for an in-store demo.

Meanwhile, the spread of predictive AI into back-office operations will impact the customer experience. For example, inventory management systems will grow in their ability to seek out predictive patterns and decide when and where to ship items – as Amazon is already seeking to do with AI-based techniques such as anticipatory shipping. For customers, this trend will mean better in-store selection and less time waiting for items. For brick-and-mortar retailers, it could provide better tools to counter online competitors by making services such as in-store pickup faster, easier, and more satisfying.

Along the way, we may see the democratization of predictive AI, as new automated ML tools put such capabilities within the reach of small and mid-sized businesses as well as the retail giants. One example is Northstar, an interactive prediction tool developed by researchers at MIT and Brown University. With Northstar’s drag-and-drop graphic interface, a user can import datasets and generate predictive ML models on any touchscreen device, using a drag-and-drop graphic interface. Such tools could allow companies of any size to benefit from predictive AI without extensive in-house expertise in data science. For example, a small business owner could use Northstar to predict sales from historical data and decide which products to stock, with only limited manual input.

The in-store experience will become intelligent.

Brick-and-mortar retailers have a pressing need to create in-store customer journeys that can rival online shopping in speed, convenience and personalization. Forrester’s 2018 white paper, “The State of the Digital Store“, summed up the situation: “In the age of the customer, the retail store must now deliver a relevant and differentiated customer experience or face obsolescence.”

A new generation of software systems, enabled by the internet of things, is emerging to help retailers close the gap by creating digitally enhanced stores that provide personalized experiences to shoppers.

The combination of powerful data solutions, from both online and offline sources, with more and more intelligent systems, will help retail stores emulate the seamless, effortless journeys that customers expect online. What’s more, next-generation AI will help brick-and-mortar retailers build on the unique advantages of in-person shopping. AI systems can improve at learning shoppers’ individual preferences, providing suggestions and guiding them to faster, more satisfying choices. And AI-powered conversational interfaces may make such interactions feel more effortless and natural over time. 

To optimize future in-store experiences, brick-and-mortar retailers will need detailed, real-time data collection and analytics similar to what we see in e-commerce. This goal will come within reach as AI-driven systems hone their ability to interpret data from cameras and audio sensors, observe human behavior, and capture insights from the physical environment.

For example, an intelligent system combining ML and advanced video analytics may be able to find and interpret otherwise elusive patterns in how customers are picking up and abandoning different items, just as e-commerce systems currently track and analyze online cart abandonment. Such applications of AI could enable retailers to fine-tune the paths customers take through their physical locations. Likewise, AI could use direct observations from video and audio to provide new insights into how a store’s human staff interacts with customers. This could lead to better decisions about employee development and training, and enhance the human touch that makes for a special trip to the store.

Customers will learn to love talking with robots.

Current chatbots and voice assistants have only a limited ability to converse with retail customers and respond to their wants or needs. Companies in industries from auto rental to satellite TV are already using AI-powered chatbots to handle customer service questions, but such systems struggle to handle the more complex interactions involved in other stages of the customer journey, such as sales.

This may change in the not-too-distant future. More advanced conversational AI could affect every phase of the retail customer experience. Such systems would interact with customers in more humanlike ways and resolve an ever-growing range of questions or requests in real time.

Customers will likely interact with conversational systems at many touchpoints, from smart speakers, apps and call centers to in-store interfaces. Over time, retail customers will become accustomed to machines that ease the burden of communication and can meet an ever-growing range of requests. They will grow used to accessing help on demand, 24/7. And activities such as searching for products, shopping in stores or calling for support could all become more effortless and enjoyable.   

Will human workers disappear from the picture?

Probably not. Instead, you can expect to see heightened collaboration between people and machines, as humans are freed up to focus on tasks that call for nuanced empathy and judgment.

For example, an automated customer service agent might be able to judge emotional states from a speaker’s language and tone of voice. When it detects anger or impatience, it could route the customer and relevant information to a human representative, who may turn to another conversational system to help find a solution. 

Is the retail singularity at hand?

Some futurists invoke the idea of the AI singularity: a revolutionary moment at which computers will transcend human intelligence. AI still appears to be far from that point, and so is the retail industry.

Nevertheless, the retail customer experience is likely to evolve at an accelerating pace in the years ahead. To compete in this new era, retailers will have to master two essential resources: data and people.

An AI-powered customer experience requires high-quality data to produce customer-focused insights and decisions. Among other things, retailers will have to meld an ever-growing variety of data and data sources into a single comprehensive, accurate real-time view of every customer.

At the same time, retail companies face the choice of building their own AI solutions or working with outside partners or vendors to acquire tools with AI under the hood. An enterprise customer data platform can help firms meet such challenges.

The human element is just as crucial. New customer experiences will require technical teams skilled in deploying AI systems. They will also require customer-facing personnel with updated skills for roles transformed by automation. In addition, retailers will benefit from building organizational structures and cultures that welcome AI-driven innovation.

Above all, the most successful retailers will stay focused on understanding what experiences their human customers want and delivering the goods. That’s something even the most intelligent machines will (probably) never change.

Image Credit:

Viorel Kurnosov / Getty Images

Tom Treanor
As the CMO for Treasure Data, Tom drives the marketing strategy and execution for Treasure Data's CDP (Customer Data Platform) solutions. Previously he helped define the product roadmap for Alexa Internet's SaaS marketing and analytics tools. Before that, he was the Director of Content Marketing, Social Media and SEO for Wrike, a leading project management and collaboration software solution. He has an MBA from the Wharton School of Business, as well as a Master of Arts in International Studies from the University of Pennsylvania.With a unique blend of business experience, technical skills and creativity, he has been able to make a meaningful impact on many companies. Others: SMM, Content Creation, Business Blogging, SEO, Public Speaking, Online Marketing