Predictive RFM?

Predictive RFM?

Does the RFM model using event probability data make sense?

Proper customer segmentation for any communication should be the first step in building a scenario. It determines the time, the channel and, most important, the content of the message. One of the basic segmentation models used by businesses is RFM(E) which tries to provide information on which stage of the interaction (funnel) the customer is at.?On the other hand, there are predictive models available that answer specific questions related to the probability of purchasing or performing a certain task (subscribing to the newsletter) and churn.

Will the combination of these two segmentation approaches into one analytics called Predictive RFM is really possible?

Let's start with the input data and mapping the classic RFM dimensions to the new ones resulting from the prediction.

R - Recency -> Propensity to buy prediction (predictions window infinite)

F - Frequency -> Number PV prediction (prediction window 90 days)

M - Monetary -> Monetary prediction (prediction window 90 days)

In the next step, let's divide the users based on the scoring from the prediction into equal segments. As in the classic model, we get 27 segments that I suggest grouping into 7 larger sets and focusing on:

Skip the line -> UX

Frequent buyers -> Loyalty Club

Buyers -> Price & Product

Hunters -> Price

Viewers -> Product

Random users -> Marketing Consents

Discount kings -> Discount & Promotion

No alt text provided for this image

Don’t forget, each business is different and in each of them, you can look for different segmentation criteria.

Are you interested in this topic; do you want to learn how to use Synerise smarter? Join us at the next Inspiration Session!


(Un)lucky number

No alt text provided for this image

this is how many segments we have divided users in the predictive RFM analysis



Use (This) Case

No alt text provided for this image
No alt text provided for this image











No alt text provided for this image



An old Chinese adage says...

No alt text provided for this image


"Not every user is a customer, not every visit is a transaction."

~Yì píng pí jiu



Are you looking for an idea for new scenarios, do you want to improve existing ones, or do you need an impulse to know what and where to look for? The Inspirations newsletter is a supplement to the Synerise Inspirations Sessions prepared exclusively for Clients and Partners. In this open to everyone form, we invite you to the world of digital marketing in which AI plays the leading role.

Business Value Services Director,?Dominik Krolikowski

Many people in retail use RFM. And that practise can ease the transition to using predictive RFM with CDPs.

要查看或添加评论,请登录

Synerise的更多文章

社区洞察

其他会员也浏览了