Question 1

What is RFM analysis and why do businesses use it?

Accepted Answer

RFM stands for Recency (how recently a customer purchased), Frequency (how often they purchase) and Monetary value (how much they spend). It turns raw transaction history into a compact, three-number fingerprint per customer that correlates with future purchase likelihood and lifetime value, without needing a trained machine-learning model.

Question 2

How are Recency, Frequency and Monetary scores actually calculated in this simulation?

Accepted Answer

Recency is the number of simulated days since a customer's most recent purchase. Frequency is the count of purchases inside the selected time window. Monetary is the total spent inside that window. These are recomputed every tick directly from the live purchase-event stream.

Question 3

How does quantile-based scoring differ from using fixed thresholds?

Accepted Answer

Quantile scoring sorts the entire live customer population by each metric and cuts it into five equal-sized groups (quintiles) at whatever boundary values the current data actually produces, rather than using hard-coded cutoffs, so the cut points genuinely shift as the population's real distribution changes.

Question 4

What do segments like Champions, At Risk and Hibernating mean?

Accepted Answer

Each customer receives an R, F and M score from 1 to 5, and a rule table combines the three into a segment label such as Champions (top scores on all three), Loyal Customers, At Risk (used to buy often but recency has slipped), or Hibernating (low scores across the board).

Question 5

Why do customers move between segments as new purchases stream in?

Accepted Answer

A customer's own history can change (a new purchase improves recency), and the population-wide quantile boundaries also shift because they are recomputed from everyone's current values, so a customer can change segment even without buying anything themselves.