Understanding what RFM segmentation is
RFM segmentation is a classical way in marketing to segment a customer database depending on purchasing behaviors. The goal is to define several customer groups and to address them with specific campaigns. RFM segmentation is always calculated on a predefined period, and is re-calculated every day.
The customer groups are defined by three indicators:
The last purchase date - Recency
The purchases regularity during the period - Frequency
The amount of the purchases - Monetary
And here are the RFM customer groups :
1 VIP
2 Very good clients
3 Good clients
4 Average clients
5 Occasional clients
6 New very good clients
7 New occasional clients
8 New clients
9 Inactive clients
10 New prospects
11 N/A: This segment is for clients not present in the database at the calculation date
Time range and Recency Frequency calculation
We have chosen to use an RFM bucket calculation method with four time periods: by default, we calculate the RFM segments over a period of the last 360 days, divided into four 90 days "buckets".
For each bucket, we check:
If the customer bought: "1” in the schema
If the customers didn’t buy whereas they were already customers: ”0” in the schema
If the client was not yet a customer at the time: ”X” in the schema. This is computed based on the first touchpoint we have for the client (registration date/first email sent/pageview/purchase…)
The first bucket is on the left, and the last is on the right on the schema.
Based on these three indicators, we calculate the Recency Frequency segment.
Amount calculation
The next step is to check, for the total period of the buckets (360 days by default) how much the customer spent in comparison to the others.
We divide the customers into four equal populations:
The 25% who purchased the least: revenue class --
The 25% who purchased the most: revenue class ++
And we split into two equal halves the rest of the customers, still based on their amount: revenue class - and revenue class +
Attributing a RFM segment to each customer
The table linked to this page shows you the details of each RFM segment attribution.
Using the RFM segmentation
In Custom Audience Filter, you can use several attributes to take advantage of the RFM segmentation calculated on your database:
RFM segmentation: this will allow you to select customers who belong today to a specific RFM segment, such as VIP.
Previous RFM Segmentation: this enables the selection of the previous different RFM segments for customers. This will apply only if the RFM segment has changed.
RFM update date: This is the date when the RFM segment has changed for the last time.
Historical RFM 3 months (on demand): This custom attribute displays the RFM segment customers belonged to 3 months ago (X needs to be defined during the setup).
Historical RFM N months (custom)