A primer on how FCRM (Frequency, Consistency, Recency, and Monetary Value) analysis can help Industrial OEMs predict buying behavior.
Customers expect OEMs to understand their needs and expectations and hence they have to come up with the right marketing strategies that work for their customers. Successful organizations do not make product, service, and marketing decisions based on their ideas, but they take into account things such as what products/services customers purchase, how often and regularly they purchase, what makes their customers engage with them, etc. Hence, understanding customer behavior at scale becomes paramount for OEMs in such cases. This is where FCRM (Frequency, Consistency, Recency, Monetary Value) analysis could help.
An analysis of customer FCRM data can help OEMs understand customer purchasing patterns, identify potential problems, and improve customer satisfaction. FCRM analysis help marketers better strategize on their campaigns, fine-tune their efforts and improve their results. Thus, FCRM plays a vital role in learning and gaining insights from the vast amount of customers and their transactional data.
But, What is FCRM Analysis for Industrial OEMs?
FCRM analysis is a 4-dimensional framework that ranks customers based on their FCRM scores. The framework is a behavioral approach to target customers a company can focus on to maintain a healthy business. The analysis help OEMs know who are their best customers and which customers need their attention.
But what exactly is FCRM?
F for Frequency: How often do the customers engage/purchase?
C for Consistency: How regularly they engage/purchase?
R for Recency: How recently did they engage/purchase?
M for Monetary Value: How much did they spend?
Let’s understand these scores with examples. The below table contains the transaction history of few customers between 2015-01-01 and 2017-04-30. These scores can be calculated at different time durations (time ordinal), it can be day, week, month or year. The table below contains time ordinal in weeks and hence contains data for 121 weeks.
| Customer | Order Date | Time Ordinal (Week) | Order Amount ($) |
| 1 | 2015-01-02 | 1 | 631 |
| 1 | 2015-02-15 | 7 | 261 |
| 1 | 2015-04-12 | 15 | 565 |
| 1 | 2015-09-08 | 36 | 330 |
| 1 | 2017-04-11 | 119 | 793 |
| 1 | 2017-04-24 | 121 | 990 |
| 2 | 2015-01-01 | 1 | 83 |
| 2 | 2015-01-08 | 2 | 511 |
| 2 | 2015-01-16 | 3 | 522 |
| 2 | 2015-01-25 | 4 | 535 |
| 2 | 2015-02-02 | 5 | 768 |
| 2 | 2015-02-05 | 6 | 903 |
| 3 | 2015-03-10 | 10 | 826 |
| 3 | 2015-05-18 | 20 | 654 |
| 3 | 2015-07-26 | 30 | 310 |
| 3 | 2015-10-02 | 40 | 494 |
| 3 | 2015-12-15 | 50 | 675 |
| 3 | 2016-02-21 | 60 | 476 |
| 3 | 2016-04-30 | 70 | 418 |
| 3 | 2016-07-08 | 80 | 675 |
| 3 | 2016-09-16 | 90 | 85 |
| 3 | 2016-11-26 | 100 | 16 |
| 3 | 2017-02-03 | 110 | 535 |
| 3 | 2017-04-14 | 120 | 962 |
| 4 | 2015-01-04 | 1 | 876 |
| 4 | 2015-01-19 | 3 | 603 |
| 4 | 2015-02-05 | 6 | 320 |
| 4 | 2015-03-03 | 9 | 322 |
| 4 | 2015-03-22 | 12 | 211 |
| 4 | 2015-04-09 | 15 | 287 |
| 4 | 2016-02-23 | 60 | 628 |
| 4 | 2016-07-09 | 80 | 890 |
Below table contains FCRM scores derived using customers’ above transaction history which are indicative of customers purchasing behavior. Each FCR score contains value between 0 to 1 where 0 is worst and 1 being best and Monetary Value is the total amount spent by customers on orders.
| Customer | Frequency | Consistency | Recency | Monetary Value |
| 1 | 0.09 | 0.25 | 1 | 3653 |
| 2 | 1 | 1 | 0.085 | 3239 |
| 3 | 0.1 | 1 | 1 | 6126 |
| 4 | 0.18 | 0.3 | 0.75 | 4137 |
As one can see, Customer 2 is making purchases often, hence, their frequency score is reflective of that. Customer 3 is not making purchases very often, but on a regular basis which reflects its low frequency and high consistency scores and the high score of recency also tells about the recency of its transactions compared to other customers’ last transactions in the same period.
By analyzing the above scores, OEMs can figure out which customers are most valued and can take necessary steps to improve relations with those customers. The below image shows a few examples of types of customers identified using FCRM scores.
| Cluster | Description | Frequency | Consistency | Recency | Monetary Value |
| Champions | Purchase recently, often, and regularly, and spend the most | Medium-High | Medium-High | Medium-High | Medium-High |
| Loyal | Purchase regular and spend the most | Low-High | Medium-High | Low-High | Medium-High |
| At Risk | Spent most and often, but a long time ago | Medium-High | Low-High | Low-Medium | Medium-High |
| Lost | Lowest FCRM Value | Low | Low | Low | Low |
Focusing on one type of customer can often lead to neglecting customers of another type. Unsatisfied customers can be less frequent, less valuable, and older than a company’s best customers. Such unsatisfied customers may never complain, but by understanding their behavior, a company can work on improving overall customer satisfaction.
When considering other customer attributes along with FCRM scores, such as the type of product or service they purchase or the industry they belong to, business unit, etc., It’s helpful to divide customers into different groups of similar types. This way, you can more easily conduct exploratory data analysis of each customer group and create customer profiles. These profiles can then be used to identify the needs, requirements, and pain points of certain customer types. From there, you can tailor your marketing efforts to better meet their needs.