Jesse Wolfersberger, Chief Data Officer with Maritz Motivation Solutions, sat down with Loyalty360 to discuss artificial intelligence in relation to customer loyalty.
In broad terms, what role does AI play in terms of customer loyalty?
Forgetting the gory mathematical details for a moment, AI generally refers decision-making with speed and scale. For example, let’s think about the AI that translates your voice in an Amazon Echo or Google Home. Without AI, you could hire a team of humans to listen to the audio and translate it and they would do a great job. However, there are tens of millions of smart speakers in use today. The work force needed to listen to all this audio would be unwieldy and expensive. The AI enables fast, accurate translation at scale.
In loyalty terms, the AI is being used, not for language translation, but for things like preference prediction.
Picture your favorite corner coffee shop. The barista knows your name when you walk in, starts your favorite drink, and recommends the blueberry muffin. This is a great customer experience that is created by and creates more loyalty with this brand. But it can’t scale. The barista can only remember so many faces, and if you go to another location, your preferences don’t follow you.
AI is helping us get closer to that corner coffee shop experience for every interaction you have with a brand. It can remember your likes and dislikes, greet you at new locations, and predict your preferences for items you’ve never even tried yet. I think this is a very exciting time to be in the loyalty business because we finally have the computing horsepower to create loyalty experiences that would have been considered science fiction a few years ago.
We still have a long way to go, but we recently completed a pilot with our client HSBC, in which we used AI to predict reward redemption preferences for cardholders. We know people often carry a lot of points in their accounts just due to inertia. We used our AI to do the shopping for them and suggest items they would likely have redeemed for themselves. Of the people who redeemed during our test, 70% of them did so in the category that the AI recommended for them. This is a great indicator that AI can add value to a loyalty program.
It is evident that AI can tell us more about data and in turn, loyalty, than a human can. What are some insights that we can get by pairing AI with data collection?
As a data analyst, using AI is fun because the algorithm finds relationships that you simply wouldn’t have found using traditional methods. For example, in our pilot with HSBC, the AI found that the cardholders with the strongest preference to use their points on travel also were likely to have used their card previously on public transportation and drinking establishments. In fact, those were two of the top leading indicators out of thousands.
This was fascinating because it is a relationship that we never would have started out trying to find, but now that the AI found it, we can understand a lot more about these customers. It makes sense, these are not homebodies, they like to get out of their houses be with other people, so of course they want to use their points to travel.
This example is the type of thing we will find the more we use AI. It gets even better the more disparate datasets you can include. You can still do AI using siloed data, but you are missing most of the interesting, unexpected data relationships that way.
How has AI changed loyalty in the last five years?
I’m not sure that it has. My answer is going to be very different five years from now. We are just scratching the surface of what an AI-powered loyalty future looks like. I think the loyalty programs in five or ten years will look unrecognizable from what we have today.
Today, most consumers join loyalty programs for the financial benefit. They are going to shop with a brand and get free products or points in return. I think this is going to change drastically for AI-powered loyalty.
Instead of a financial incentive, the exchange will be data for experiences and conveniences. I often think about Stichfix, the subscription clothing service. That is a case where I give a ton of data – access to my Pinterest, Linkedin, Twitter accounts, and my body measurements to name a few – in exchange for clothes that fit my size and style. I think this is more of what loyalty will begin to look like. The brands that can improve their customers’ lives the most will be able to differentiate themselves and earn more trust and data from their customers.
What is the biggest mistake you see brands making with AI today?
I see some brands making mistakes in their tech stack. No matter what a vendor tells you, AI is not a tool you can just buy and then press a button. The AI will only do what it is instructed to do, so the real hurdle right now is the creativity of how to integrate AI into the customer journey. Getting to AI is about strategy and alignment within an organization. Only after that is accomplished can you pick the right tool to implement.
In fact, the dirty secret that no tech vendors and SaaS tools will tell you is that the world-class, cutting-edge AI algorithms are all open source and free. This is great news, as long as you have a data science team with the knowledge and bandwidth to make use of it. This is a non-trivial point, so I want to be clear – brands in the next five years will rise and fall depending on if they have people and partners who can ride the wave of innovation happening as we speak in the open source community.