RFM stands for Recency, Frequency, and Monetary value. Within our platform, each representing the following;
Recency: How recently a customer has made a certain purchase
Frequency: How often a customer has made a purchase
Monetary value: How much (money) a customer spends on average
These three metrics are key to get a better understanding on a customer's behaviour, for instance;
If you want to know more about a customer's lifetime value, you can check the Frequency and Monetary value.
If you want to know more about a customer's engagement with your website, you can check the Recency, which affects the retention of a customer.
Now that we have explained what RFM stands for, the next step is to provide some information on how we measure RFM;
Once a week, we measure all (RFM) statistics off all your customers who made a purchase in the 365 days prior to the date of the measurement.
Important to note is that an RFM score is based on segmentation, where in our case, we split up eacht metric in 5 segments (see table below);
With all of your customers (eligable for analysis) being placed in these segments, you can get a clear picture of what type of customer an individual is.
For instance; a person who places a big order of 300 EUR once every season, will have a lower F-Score and a higher M-Score than a person who places an order of 50 EUR once every month.
To get a better understanding on how exactly an RFM analysis work, we will take a look at the following customer set:
Recency (R-Score)
To determine the R-score of the customer set above we can sort the recency numbers, rank these and them split them up in 5 segments. The customers who most recently bought something will receive the highest R-Score. As shown in the table below:
Frequency (F-Score)
To determine the F-Score, we apply the same principle as the R-Score:
Monetary Value (M-Score)
To determine the M-Score, we apply the same principle as the R-Score and F- score:
RFM Score
Now that we have the individual metrics in order, we can determine the RFM Score:
Conclusion
Since we have all of our numbers in place, we can now draw several conclusions from the analysis, for instance;
Customer ID 10 is one of our best customers. This customer is a heavy spender, and shops quite often. It's been a while though, since this customer's last order. So it might be a good idea to engage with this customer and recommend some products (via email for instance).
Customer ID 1 is a new customer, you can tell by the fact that this customer has very recently bought something, and has only bought something once.
Customer ID 5 is a customer at risk, since you can tell that this customer hasn't made a purchase in a long time, but has frequently made purchases.