Customer journeys help to increase positive results in all customer-related KPIs, but in large enterprises and public organizations, the rule is if you cannot measure it, you cannot improve it. How do we measure commitment to a brand and products – and how do we score KPIs that need to be improved upon?
A recent report by McKinsey indicates that customer journeys are 30% more predictive of customer satisfaction than measuring individual interactions and that using customer journeys can increase customer satisfaction by up to 20%, leading to significant revenue gains and lower costs. In another study, a premier automotive OEM said that every 1% increase in sales retention translates to a $700 million increase in revenue annually – an average of $150,000 per dealer.
Customer journeys can be understood as a discrete, unevenly sampled time series of customer events that contain both unambiguous signals of commitment—like buying a new car—as well as more ambiguous signals of commitment, such as a monthly series of credit card purchases or contact with customer service.
Machine learning comes into the picture
How do we unleash profitable growth for these companies using customer journeys? Since we are trying to predict an outcome, statistically we need to gather lots and lots of customer journeys to arrive at an answer. A larger data set will most often yield a better result than a smaller one. This is the revenge of the large enterprise because 10 million customers can, ostensibly, produce billions of data points.
However, the sheer size of the data set means nothing. What really matters is how these data points interact with each other. It is simply impossible for a human brain to comprehend all of the patterns that could be discernable in this volume of data. Machine learning solves this problem.
Machine learning models are good for making predictions. The easiest case would be a sale. The average vehicle sold in the U.S. and Canada costs roughly $35,000. That’s easy to see in a data set, but there might be 10 to 20, or even 100 events before a customer finally buys that SUV they always wanted.
The machine learning model not only looks at the customer, but it looks at all customers, especially those with journeys close to the customer in question, to understand what events are most important in driving an outcome. This requires a lot of number crunching. It is not unusual for us to do 5-10 trillion calculations to solve this type of problem. In changing the weights assigned to events in a model in this way, the machine learning model – learns.
Why it matters when measuring customer experience
Cerebri AI has filed numerous patents on how all this works, but why does this matter? Assume we have two buyers of SUVs, one paid $30,000 and one paid $60,000, and each had ten events leading to the actual purchase. In the first case, all ten events are valued adding up to $10,000. In its simplest form, if each event was equally weighted by our models, then each event was worth $3,000 to the end goal, the purchase of an SUV. In the second case, under similar circumstances, each event was worth $6,000, or double.
How does that work? If events lead to a bigger purchase from an enterprise point of view, the events are simply more valuable and more events like these lead to a larger purchase. In other words, if you go to a vehicle OEM website to decide on a car and your purchase ends up being twice the standard price, then your visit is twice as valuable to the vehicle OEM.
It’s simple. Everyone understands money, we use it every day. But that simplicity masks the real power of the system, and that’s why Cerebri AI has applied this approach across markets where customer experience is critical to predicting growth. More than anything else, it’s an approach that is easy to introduce, using technology as a platform, with little disruption in time and training.
We are AI ML Editorial Team. We come up with informative quality articles on AI, Data Science, and Machine Learning. If you also want to contribute, kindly get in touch with us.