In the business world, understanding customer behavior is critical to developing effective conversion and loyalty strategies. A powerful tool to achieve this is RFM analysis.
In this article we will explore how to simplify segments and group-specific actions based on this powerful procedure.
What is RFM analysis?
Once the first purchase has been made, we consider the user as a customer. But when he is already a customer... Do all customers come back with the same needs or can they change their interests? Based on this, the following question arises: how do we get that customer who has trusted us to make their first purchase to continue trusting our brand for their next purchases?
To answer this question, several issues come into play, such as the type of sector of our ecommerce, whether it is a recurring sector or based on specific needs, the monetary value of the product as well as the availability of the user or whether it has solved a specific need without having a second purchase in mind, among others.
Here it is essential to analyze the database of our "already" customers so that, in the same way that we segment users who have not yet made their first purchase to encourage purchase, we get those who have already bought from us and therefore, we consider them as customers, to buy again or even be able to recommend others to do so.
Thus, the RFM analysisThe company's customer loyalty program, as a means of classifying customers who have already made a purchase on the basis of three variables:
● R: Recurrence (time since last purchase),
● F: Frequency (total number of purchases made in a given period of time)
● M: Monetization or monetary value (average amount of purchases made).
How to perform the RFM classification?
Here the technical part comes into play, which is based on the explanation of the functioning of this system based on quintiles, as we can see in the following table. The three variables from 1 to 5, resulting in 15 grids that are put together in 11 groups:
These 11 segments can in turn be grouped into smaller groups in order to increase the number of contacts in each segment. In this case we will use 4 main groups:
● BEST CUSTOMERS: Champions + Customer ﬁeles + Potentially ﬁeles.
● POTENTIAL CUSTOMERS: New customers + Promising
● ALERT CUSTOMERS: Attention + Can't lose them + Falling asleep + At risk
● CUSTOMERS ON A TIGHTROPE: Hibernating + Lost
Based on this classiﬁcation, each of the marketing strategies that we will discuss below will be focused, creating speciﬁc workﬂows for each of the segments.
How is the RFM classification applied within the purchasing process?
If we recall previous articles such as the one on lead nurturing where we talked about the phases of the buying process, this time we will focus on the last stages, which include Loyalty, Retention and Reactivation of dormant customers.
It is thus essential to classify the strategies aimed at existing customers on the basis of these three objectives:
● LoyaltyFocused on customers with high RFM, with the goal of converting them into VIP customers who not only buy from us against the competition, but also recommend us to other users to make their first purchase.
● RetentionFocused on customers with medium RFM, with the objective of getting those who have already made their first purchase with us, but are undecided, to return to make more purchases with us.
● Reactivation: Focused on customers with low RFM, those who once made a purchase on our website or ecommerce, but both recurrence, frequency or even, monetary value, implies a speciﬁc strategy focused on recovering them as customers before they go to the competition.
Where can you find the RFM segments within Connectif?
Within the Connectif tool itself, we can view these RFM segments from the user profile. Once inside, we scroll to a middle area and we will see a basic table with the values of Recency, Frequency and Monetary Value.
As can be seen in this example, this customer has Recency 4, Frequency 5, Monetary Value 3, and is therefore categorized as Potentially Loyal.
In addition to this position, we can play with the values from the Dynamic Segments and Dynamic Plus part itself.
- Dynamic SegmentAs you can see, you can filter for each value individually or also by type of specific segment.
- Dynamic Plus Segmentwith this more complete functionality we can extract multiple variables to add them to the segment or exclude them.
RFM applied to Data First
Now let's get down to what we are really interested in, which is looking at the RFM data within the tool and how to get the most out of it. The most important thing when it comes to planning actions with a customer is evaluate the data, understand it and be able to provide insights to help us improve for future campaigns.
From the Dashboard we can use the RFM metrics to better understand our customers in various aspects of eCommerce:
● Web Content
|● Push notifications
● Web Traffic
Once we see the graph, we can ﬁltrate by several variables:
RFM analysis case studies
- BEST CUSTOMERS - Loyalty - The idea is to offer them a personalized message where we let them know that they are our best customers and some kind of pre-sale/private sale. It could be a campaign where this RFM group is sent the promo 6h before the rest.
- PROMISING CUSTOMERS – Retention - We can provide you with valuable information
or branding, for example in the time it takes to ship and receive a product.
- ALERT CUSTOMERS – Reactivation - To be able to try to make some kind of interactive game where the user can make engagement with the brand. This way you will see that we include news within the account and that we are interested in enhancing the relationship with the user.
- ON A TIGHTROPE - Speciﬁc campaign, ignore segment of all other generic campaigns (either they buy again or they leave).
- CUSTOMERS WHO HAVE NEVER PURCHASED - Boost first purchase - Applying an interesting discount for the first purchase.
Let's develop one of them so you can apply it in your campaigns. In this case we want to do something more interactive so we're going to go with the segment #3for Clients on Alert.
The roulette of luck
As we love to interact with the user and offer him innovative actions where he is a participant, we have developed a lucky roulette based on GIFs and the wonderful "Split" node.
First we create a Dynamic Plus Segment to make our Customer Alert RFM segment (important to note is to add the other RFM values with "o"):
We start from the initial node in which we select our new Clients on Alert" segment. When one of these customers visits the page, we activate the Split node with different odds. Through a pop-up with a roulette GIF, we show different versions with different prizes. That is, we have: 4 prizes = 4 pop-ups = 4 GIFs (each for one prize).
Once we show the roulette and the prize that has been won, we provide a form so that the customer can leave us their email (great for capturing new leads ;). In order to use the coupon, the customer needs a code that the customer receives via email once he/she has left us his/her email.
And it's that simple! You can make similar campaigns adapted to many themes and times of the year. You set the limits for yourself.
Here is an example of the roulette wheel:
We do not want to close the post without first making a small recommendations and essential tips: