CRM Analytics — Customer Segmentation with Using RFM Model
For companies, the customers to whom they sell products or services are of great importance. Companies that do not want to apply the same strategy to each customer prefer to divide their customers into groups according to various characteristics and develop a group-specific strategy. In terms of grouping customers, the data of the customer and the processing and analysis of this data are critical. One of these methods is the RFM model under the topic of CRM Analytics.
In this article, I tried to explain CRM analytics, customer segmentation and RFM model. You can access the codes of my case study with the GitHub link at the end of the article.
What is CRM?
CRM stands for Customer Relationship Management. It works on how to classify customers and how to get the most efficient result from customer relations. CRM analytics come into play to determine whether there is a similarity between customers and how customers should be classified.
What is CRM Analytics?
CRM analytics is trying to analyze customer relationships through data. CRM analytics measures customer purchasing behavior through various metrics and tries to measure how successful the company is selling the right product to the customer. These metrics are generally called KPI (Key Performance Indicator). To give a few examples:
Customer Churn Rate: The customer stops buying products from company X and starts to prefer another company.
Customer Acquisition Rate: The percentage of customers gained in a given time.
Customer Retention Rate: It is the percentage of customers who continue to buy / visit the company’s site for a certain period of time after customer acquisition.
Useful Information: When the user makes his/her first purchase, companies generally do not accept him/her as a customer. But if the user makes his/her second purchase from the same place, they now accept him/her as a customer. In this way, Customer Churn Rate and Customer Retention Rate values can be more accurate.
Useful Information: For the firm, acquiring a customer is costlier than retaining one. The cost spent on marketing activities through various channels to gain customers is quite high. However, a small discount or campaign to retain a customer can be done at a very low cost.
RFM Analysis
RFM Analysis is a simple rule-based technique used for customer segmentation in CRM analytics. In the RFM analysis, customers are grouped based on the sales data of the customers. RFM; It consists of the initials of the words Recency, Frequency, and Monetray. Let’s take a look at these terms:
Receny: It means the time elapsed between the analysis date and the customer’s last purchase date. In CRM analytics, the analysis date is usually chosen close to the date the data was collected. If the calculated recency value for a customer is small, it means that the customer has recently made a purchase. Customers with low recency are more likely to remember the company and the product they bought, while customers with high recency are more likely to forget the company. Different incentive plans can be made for different customers based on this value alone.
Frequency: It refers to the customer’s purchasing frequency. The customer with high frequency value is loyal to the company. Loyalty programs can be applied to such customers. In addition, by analyzing how many purchases the customer has made in which time period over the customer’s frequency value, it can be predicted when the customer will make a purchase in the future.
Monetary: It is the total money earned by the customer as a result of their purchase.
When R, F, and M values are calculated, comparing these values with each other is complicated at first. Because the F value can be expressed with smaller numbers, while the M value can be larger. When we standardize these 3 metrics, we can make the data more understandable. We can create an RFM score by scoring all 3 metrics between values 1–5. After making these scorings, we can complete the customer segmentation. After making the scores, if we show the Recency values on the X axis and Frequency values on the Y axis in 2-dimensional space, we can get a table like this:
When we examine the table, you may be wondering why the Monetary value is not included. The reason for this is that there may be errors in customer segmentation if some customers have a very high monetary value and only shop once. Therefore, the Monetary value is not included in the table. It is possible to identify customers who buy more frequently and remember the company over Recency and Frequency values, and carry out special campaigns and marketing activities for them.
If we included the Monetary value in the segmentation, it would be necessary to make a classification over 125 (5x5x5) different RFM values. We created 25 (5x5) classes using only the Recency and Frequency values, and we can evaluate these values in 10 different segments. To summarize these 10 segments:
· Champions: It is the best customer group for the company. Both the top shoppers and the most recent purchase dates are very recent.
· Loyal Customers: They shop frequently and it has not been long since their last purchase date. Champions and Loyal Customers group is the most valuable group for the company. Loyalty program can be applied to these groups as they are the groups that the company makes the most money. These customer groups should be made to feel special.
· Potential Loyalists: Purchased recently and the frequency of purchases is fine.
· Promising: It is a group of customers who are not considered very new customers, but some time has passed since the last purchase date.
· New Customers: He/She has shopped very recently and his/her purchase frequency is not high. It is probably the customer group that has made the first purchase specific to the relevant company. The company should remind itself to this customer group by methods such as discount messages and promotional e-mails.
· Need Attention: It is located in the middle of the table. If a study is not done for this customer group, it is the customer group that is likely to move to the regions of the table that are considered bad for the company.
· About to Sleep: It is a customer group that does not shop very often and it has been a while since their last purchase.
· Hibernating: It is a group of customers who are not affiliated with the company, maybe only shop for a product and then do not return.
· At Risk: It is the group of customers who have made significant purchases, if not the most, and have forgotten the company for a long time.
· Can’t Loose Them: It is a group of customers who have made frequent purchases in their time but have stopped shopping for some reason. A critical problem for the company is discovered if the reason is investigated by paying special attention to At risk and Can’t Loose Them groups.
You can access the codes of the project I made using the RFM method in CRM analytics from my GitHub profile.
www.linkedin.com/in/onurhalityenice
onurhalityenice@gmail.com
Thank you for your interest. See you in my next post!
