In feature engineering, what does the RFM model stand for and what does it measure?

• A. Randomized Feature Metric; measures random noise impact.
• B. Recency, Frequency, Monetary; used to rank customers by engagement and value.
• C. Rate, Frequency, Margin; used for pricing optimization.
• D. Readiness, Familiarity, Memorability; used for creative testing.

Answer: (B)

RFM is a simple customer-segmentation approach based on three behavioral signals: Recency, Frequency, and Monetary value. Recency measures how recently a customer made a purchase or engaged, which helps identify active versus dormant customers. Frequency captures how often the customer transacts in a given period, indicating their level of engagement. Monetary reflects how much money the customer has spent, signaling their value to the business. By combining these three dimensions, you can rank customers and group them into meaningful segments (such as high-value frequent purchasers who recently engaged, or at-risk customers who haven’t bought in a while). This enables targeted marketing actions, better resource allocation, and optimized campaigns. The other options don’t fit standard marketing analytics: a randomized feature metric isn’t a recognized customer-behavior measure; rate, frequency, margin mixes pricing-related metrics with engagement; and readiness, familiarity, memorability relate to ad recall or creative testing rather than customer purchase behavior.