As business leaders, we appreciate a well thought-out and defined plan, looking far into the future to ensure our financial stability. So, as you look at the next business intelligence report that comes across your desk this week ask yourself this question: “What is this report really telling me about the future of my business?” The answer, most likely, will be less than shocking.
Anticipating the future needs of even a single customer is nearly impossible without having the ability to properly use existing data. As marketing leaders, we may like to think in terms of trend. When we follow trends we make the assumption that the trend is going to continue, or at least it should continue within some small increment of time. The truth however, is that we never know when the trend is going to end or possibly change direction entirely. Now, I’m not saying let’s fire all of the business analysts. What I am saying though, is that it is time to put down the report and demand better.
In this article we take a deeper dive into wallet share maximization and describe an advanced toolset for providing exceptional customer service and maximizing wallet share by anticipating impulse, sporadic decisions, and uncertainty. Welcome aboard.
The Random Walk
Often the most brilliant and intuitive of solutions come from experiences we have as children. Finding the answer to a problem is often a result of looking at the problem from a different perspective. Let me explain.
I remember a game I used to play as a child with my dad and older sister Christine. Saturday mornings were always spent with dad. The morning would start with all three of us piling into the car. If we were lucky, and dad was feeling especially brave, we would spend the morning playing one of our favorite games, ‘lost’.
The game of ‘lost’ starts with everyone in the car. A coin is flipped. While the coin is in the air, two people in the car call ‘heads’ or ‘tails’ predicting the outcome of the toss. Whoever predicts the toss correctly gets to start the game. Now, imagine yourself in your car driving down the road away from your home. At the first intersection, the person who won the toss gets to decide which way to turn. This starts the game of ‘lost’. At each subsequent intersection, all passengers in the car take turns deciding whether to turn right, go left or continue through the intersection. The game finally ends with a stop for lunch (and directions) often many miles away from home.
Driving the Marketing Machine
Customers begin their relationship with a bank in a variety of ways. Some may open a savings account, require direct deposit checking or shop for an auto loan. With each additional product, the direction of financial needs change. We must anticipate each turn with flexible financial products. Often, the timing of a customer’s financial need can feel like our once favorite childhood game. The next move is often determined only by the last. This type of mathematical model uses a sophisticated statistical toolset called Monte Carlo simulation.
To describe the situation at hand, let us assume that our bank has 4 different financial products. Say a customer begins their relationship with the bank by purchasing product # 2. If the customer has only purchased product # 2 then at this point there are 3 available options. We then use past customer data to first cluster within stage and then build predictive models to determine the probability and timing of a particular customer to purchasing each of the three remaining products. All three models are thus different depending on the cluster of the particular customer. We duplicate the process after each purchase, marketing individual products to customers with custom offers at each stage when they are most likely to purchase. The picture of the problem is analogous to the game of ‘lost’ with each customer represented by an individual car on a street and each turn of each car dependent only on the timing and decision of the last.
Business intelligence has evolved. We are no longer confined to the limitations of reporting post mortem. Servicing the financial needs of our current and future customers requires the ability to anticipate sporadic decisions and impulse. Properly executed, random walk algorithms coupled with traditional statistical methods can provide the required insight, that is, if a customer is going to buy, when it is most likely to happen and why. Data is becoming our single most valuable asset in maximizing wallet share.
In our next series of whitepapers we will begin to take a deeper dive into helping you develop a random walk platform; be able to simulate the decision making process in order to better service the financial needs of your current and future customers. As we progress, we will continue to provide easy to follow steps you can take to maximize marketing efforts to your existing customer base. Enjoy!