Propelled by the need for speed, the desire to enhance customer experiences and mandates to reduce operating expenses, artificial intelligence (AI) is poised to disrupt industries and companies around the world.
Machine learning, a subset of AI, analyzes numerous data points at superhuman speed to create algorithms for a growing range of applications, such as detecting fraud in real-time, facilitating personalization, improving patient outcomes and optimizing inventory.
Although not perfect, AI has greater speed and accuracy contest than its human counterparts. It standardizes recommendations, thereby reducing the risk of biased decisions, and has no problem taking on the most mundane of tasks.
Large advances are occurring in the area of voice recognition, as well. The rising popularity of voice assistants and rapid growth in skills they offer will pave the way for widespread consumer interaction with robots and chatbots. Meanwhile, rapid headway is also occurring in visual systems, including biometrics recognition and systems used to detect objects in self-driving cars.
The possibilities seem only to be limited by the creativity that humans bring to AI.
Artificial Intelligence and Loyalty
The same factors – the need for speed, customer relevancy and expense reduction – that drive AI disruption overall are also at the heart of the surge in interest in applying AI to the loyalty industry.
Customer loyalty data represents some of the richest data sources available to feed AI algorithms. AI can comb through reams of customer touchpoints to portray a data story of one’s personal shopping behavior. This, in turn, helps companies improve customer journeys online and at brick-and-mortar locations. Al can recommend and help customers locate items that hold specific appeal; it also can increase cross-sell opportunities and enable more targeted communications.
HSBC’s card program, using Maritz Motivation Solutions, recently ran a promotional campaign to 75,000 loyalty rewards customers. The goal was to be able to predict the percentages of program members who would redeem their points in four different redemption categories, based on AI recommendations. Each cardholder was sent an email with one recommendation and the results were quite impressive – an open rate of 40 percent for the emails and 70 percent redemption in the recommended category among those receiving the targeted messages and redeeming.
AI can help retailers develop programs, including loyalty rewards, to be more reflective of what individuals want, such as more personalized retail loyalty programs with more control over how rewards are earned and spent.
Creepy or Relevant?
Recent revelations regarding Facebook’s use of individuals’ data has sensitized consumers to what’s being collected about them. While the consensus is that Facebook crossed the red line of privacy, the question of how much and what kind of data consumers are willing to trade for more relevant interaction with a brand remains.
Some companies have made the decision not to include healthcare-related data in their cardholder data sets, but where the red line falls beyond that remains hazy. What’s clear is that brands must become transparent about the data they collect and how it will be used. What has been presented as a black and white issue – a consumer either opts in or out of data sharing – fails to consider the greater specificity that consumers will demand regarding what they share, with whom and for what purpose.
We need to fix that creepiness factor. That means using caution about what variables are included in data sets. It also means putting consumers in charge of decisions about what happens to data.
It’s Still the Wild, Wild West
Companies are racing to embrace it, but we’re still very much in the early days of AI. Although the outlook for AI is bullish, it will take time for companies – merchants, issuers, and processors – to determine how best to leverage this powerful tool in the contexts of their businesses.
An overly ambitious AI strategy is likely to result in lack of support. Instead, companies should start with small projects upon which early successes can be built. Learn through iterations. Assess what works. Hone the algorithms. Make necessary course corrections. And remember that human behavior will never be 100 percent predictable.