By Steve Zisk
Your primary care provider calls you to let you know to stop taking a recently prescribed medication because it causes a reaction when paired with another medication you’re taking. A hotel provides you with a virtual tour on a mobile app to let you choose which room to reserve. When you head to the water park, the app guides you to a vacant chaise lounge that meets your preferences for shade and proximity to a lifeguard station. At a wholesale grocery store, an associate finds an item you’re looking for by scanning the entire warehouse — all while standing by your side.
Once a niche technology in engineering and manufacturing, digital twins are taking industries by storm, and the real-world examples above from healthcare, hospitality and retail demonstrate how digital twins are increasingly being used to enhance customer experience.
A Mirror Image: Digital Twins and IoT
Gartner defines a digital twin as a digital representation of a real-world entity or system, with implementation of the entity or system as an “encapsulated software object or model that mirrors a unique physical object, process, organization, person or other abstraction.”
As a concept, digital twinning originated as a method to test and monitor a physical object without having close proximity to it — think rocket ships. In its infancy, use cases were mostly tied to building and testing engines and other complex physical objects. Soon enough, digital twins expanded to entire factory floors, supply chains, city grids and other processes and systems, used mainly to identity existing or future problems with the non-digital object or system, such as predicting engine failure.
There is often an Internet of Things (IoT) component, with sensors connecting the physical and digital objects and providing a virtual representation. A classic example might be multiple pieces of equipment on an assembly line, where sensors on the machinery connect to a visual model that identifies potential fail points. IoT devices also represent several familiar customer experience use cases, such as a smart thermostat or a connected appliance that adjust to a customer’s preferences. Using sensors to create a virtual model of real-time inventory might enhance a personalized customer experience by helping a customer fill an online shopping cart based on availability at their favourite location.
Predictive Digital Twins
A predictive digital twin takes the concept a step further and allows digital twins to not just measure but also predict real-world behaviors. Once a digital representation of a real-world entity or system has collected enough sensor data, it might then be programmed to predict how the measured system will react or respond to external forces or situations. A healthcare provider advising you on medication adherence is not, for example, referencing an actual digital model of your body but a statistical model based on past experience.
Similarly, a virtual downtown traffic grid, fed enough data, might predict high volume patterns which are then used by your hotel the night before your check-in to notify you of the shortest route from the airport at the anticipated check-in time.
Another digital twinning use case for enhancing customer experience combines a digital model with an interactive component. One example is a virtual dressing room, where an augmented reality (AR) app allows a customer to virtually try on items before purchase. Similarly, a customer might “place” a digital representation of a piece of furniture in their living room, seeing for example whether a couch will fit in a certain corner, or how the fabric matches the paint colour.
Digital models can even encompass machine learning models. Using sensors in a store, a brand might collect traffic patterns for an extended period of time. Once data is gathered and the model trained, a machine learning model can run simulations with a goal of optimizing floor layout on a store-by-store basis. Similarly, a model might use a digital representation of an audience to analyze or predict a propensity to purchase, product affinity, likelihood to churn or another business metric. By feeding the results of how an actual audience behaves into the digital model, a brand closes the testing loop and improves the accuracy of the predictive digital twin. Conceptually, a digital model of a customer might represent a composite of real-world behavior with what’s happening in the virtual model, providing a brand with an even more detailed roadmap or insight into customer journeys.
Still a relatively new concept, first gaining widespread recognition about 20 years ago, the use of digital twins is quickly making inroads in customer experience across all verticals. Gartner recently predicted that digital twins of customers have the potential to transform how enterprises deliver experiences by simulating and anticipating customer behavior, much like how engineers first used them for predictive maintenance.
Redpoint is a firm believer that delivering a real-time, omnichannel personalized customer experience requires matching the cadence of a customer as a customer journey unfolds. At its core, the use of digital twins in customer experience is all about staying close to a customer without encroaching on their experience — being relevant but not creepy. A digital representation of a customer that models close proximity adds another tool to a marketer’s arsenal, helping a brand to deliver memorable moments that resonate with customers.
Steve Zisk is a seasoned technology professional with more than 35 years of expertise in software engineering and product marketing. As senior product marketing manager at Redpoint Global, Steve is tasked with developing messaging and marketplace positioning for Redpoint’s customer engagement platforms.