We’ve been thinking.


Confusion Matrices Hold the Key to a High Performing Predictive Model

Most of us probably realize that there is incredible power in predictive models and machine learning models. These models can help predict our customers’ future behavior before it happens, create individualized customer experiences, and help us connect to our customers on a deeper level.

However, did you know that every predictive model comes with its own set of trade-offs? That’s right, many times, by maximizing predictive accuracy in one area, you can actually increase your error in another area.

Read on to learn how to balance these trade-offs and take advantage of predictive modeling.

Accidental Beneficiaries Ate My Lunch

One of the key challenges in designing a loyalty program or any type of promotion is to avoid rewarding behavior that would have occurred anyway. We use the phrase “accidental beneficiaries” to describe rewarding existing behaviors. The cost of rewarding existing behavior is easily underestimated or overlooked in designing and assessing promotional initiatives.