Posted by Dr. Andrew Jennings, FICO Chief Research Officer and Head of FICO Labs
No one doubts that more data and more relevant data leads to better models. Winners, however, won’t just use that data to build better models—they will use it to ask better questions.
Take for example the way response and propensity modeling has traditionally been done in marketing. Often what you’re actually modeling is the offer you just made—e.g., “How many customers in this new target population will respond to and accept this existing offer?”
A smarter use of analytics is to model the offer you are about to make. That means asking questions about which products individual customers are most likely to buy next, and when they are most likely to make the purchase. If customers buy a particular product, which related products are they most likely to buy within a specific range of time? Which channels does this customer use most, and does that pattern vary by season or by day of the week?
By using analytics to answer these kinds of questions, you can predict an individual customer’s sensitivity to the specific attributes of an offer (including packaging, pricing, delivery channel and timing). You can also automatically generate population segments with similar sensitivity. This is the key insight on which differentiation can be built. We’ve seen this kind of precision boost response rates to 25%, with 85% of the responding population converting to new customers.
Better questions need to be asked in the credit risk area too. Lenders can now use analytics to predict not only “What is the current risk that this consumer will become delinquent?” but “What is the future risk if they take on additional debt?” Answering questions like these, which address credit capacity, is essential to fully meeting requirements for assessing a borrower’s ability to pay. Lenders should also be addressing questions of credit serviceability—”How are this customer’s usage and payment behavior likely to change as interest rates shift?”
Finding such answers is immensely more difficult when you have silos of data and product-oriented views of the customer than when you have shared data and a holistic view. So while the effort required to overcome data and organizational differences is not trivial, it can pay back handsomely in opportunities for analytically driven competitive differentiation.
