Example of customer profitability analysis using accounting data

February 9, 2016

Big data, data mining and business analytics are on the rise these days. Are there ways these can be used by accountants? This article provides an example of using accounting (“big”) data to perform an analysis of customer profitability.

1. About tools and approaches for data analysis

You have probably heard about big data and business analytics.  Different business and non-business areas make use of these concepts and accounting is no exception.  Accounting produces a lot of data that sometimes becomes dormant as soon as it’s created.  However, such data may provide valuable insight in various aspects of running a business.

There are a lot of tools to analyze accounting/business data and software packages which include simple or sophisticated methods and algorithms.  Microsoft Excel (or another spreadsheet application) may represent a relatively simple tool to analyze data.  SAP and SAS include more comprehensive modules for data analysis.

Aside from the various tools, there are also different ways or approaches to analyze data.  Some of them are relatively simple (e.g., calculating an average order amount) and some are complex (e.g., statistical regressions).

2. Example of performing customer profitability analysis

Let’s take a look at a more complex example of performing customer profitability analysis using the recursive partitioning algorithm.  This concept is described in more detail in an article titled “From Bean Counters to Bean Growers: Accountants as Data Analysts - A Customer Profitability Example” Journal of Information Systems 2015.

The article authors discuss the role of accountants in business analytics and specifically talk about using the recursive partitioning algorithm to analyze customer profitability.

This algorithm is typically performed using software packages.  The idea is to analyze all customers and determine what factors impact their profitability and take action based on obtained results.

This analysis assumes that a company has sufficiently detailed cost information to performed needed calculations.  In this case, sufficiently detailed cost information means an income statement by customer as well as data about business transactions related to customers.  Such business transactions are the dimensions across which customers will be analyzed (“partitioned”) and may include the average order size, the number of shipments per order, the number of customer returns and so forth.

Data about customers is used in a software package that performs algorithmic calculations to determine the best way to split customers into subcategories using the mentioned dimensions (criteria).  The created subcategories are then further split into subcategories using different dimensions and so on.  In effect, this process creates a tree with one branch at the top, two branches at the second level, four branches at the third level, etc.

For example, a result of this analysis may indicate that the best way to group all customers initially is by their order size.  Most profitable customers will be included in a subcategory with larger order sizes and most low- or non-profitable customers will be included in the subcategory of smaller order sizes.  Then each of these categories will be split further.  The large order size category may be divided into two more subcategories using the number of shipments per order.  On the other hand, the smaller order size category may be split into two subcategories based on the number of customer returns.  Once the tree has been completed, accountants can see which customers are least profitable and what drives lack of profitability (i.e., smaller order sizes with lots of customer returns).   Using this information accountants can then make recommendations to management about possible actions.

Not a member?
See why people join our
online accounting course:
Lecture Contents:
Free Study Notes
Download free accounting study notes by signing up for our free newsletter (example):
First Name:
E-mail:
We never share or sell your e-mail to third parties.
Ask a Question
Suggest a Topic
Do you have an interesting question or topic?
Suggest it to be answered on Simplestudies.com: