6c. READING

Question 1

In RFM, which factor is considered most important?

Answer 1

Recency

Question 2

In a 5-bin system, how many combined RFM scores are possible?

Answer 2

125 possible scores

Question 3

What variables are needed for RFM from transaction data?

Answer 3

A customer identifier, a transaction date, and a transaction monetary value.

Question 4

What is the difference between nested and independent binning?

Answer 4

Nested binning: Frequency is ranked within recency groups, and monetary within frequency groups. It gives a more even distribution of combined scores, but scores are harder to compare across customers. (This is also called SEQUENTIAL)

Independent binning: Recency, frequency, and monetary are ranked separately. Scores are easier to interpret, but combined scores may be less evenly distributed.

(this is just: independent RFM /// sequential RFM)

Question 5

What is the difference between a cluster and an RFM analysis?

Answer 5

• In an RFM analysis you get two segments, most valuable customers, and a huge segment of non-value customers. - So filtering out valuable from the non-valuable.
• If you run a cluster analysis with RFM variables, you would get 3 or 4 clusters who will differ according to R-F-M, but they wouldnt tell me which are more valuable or not.. You might get a cluster with high recency but low monetary value… it sorts them according to the patterns it sees - NOT according to their VALUE.

Question 6

Can you run an RFM analysis on this data like this?

Answer 6

No.
This is at the transaction level, not the customer level.

You first have to AGGREGATE the dataset by ID.