Targeted Marketing with Customer Purchase Probability

CHALLENGE

Commerce companies with a non-contract customer base have the challenge of predicting future income. 

RESULTS

A machine learning data application was created to predict the probability and amount that customers will spend within the next 90 days.

Having a purchase probability score for each customer helps increase revenue and reduce churn.  The data application was created to answer key questions:

  1. Which customers have the highest spend probability in the next 90 days?
    This allows the company to target specific customers for a higher probability of purchase.

  2. Which customers recently purchased but are unlikely to buy?
    This helps reduce customer churn by not targeting recently purchased customers with sales email.

  3. Which customers are predicted to purchase but did not (missed opportunities)?
    This allows the company to micro-target customers for a higher probability of purchase.
CONSULTING SERVICES

• Data Application
• Machine Learning