What is this study about?
This is an illustration of a model that measures and analyzes the performance of a sales team with respect to customer retention—the ability of the sales team to maintain a successful relationship with customers and make them keep coming back, year after year.
Where to apply?
This model was originally built for a B2B (business-to-business) enterprise where each order has significant value and each client accounts for a large portion of total revenue. Usually, in B2B enterprises, the number of clients is much smaller than in B2C (business-to-consumer) companies. Therefore, it’s more viable to track every single customer. Considering the value of each customer to the entire client’s portfolio, tracking each client is worth the time and resources. However, if it creates value, this model can also be applied to B2C enterprises to track and study each customer.
For the purpose of this study, we assume our company is a B2B enterprise.
How does it work?
This model studies the number of the unique clients who have placed orders with the enterprise in a certain year and tracks those clients to see if they are retained (if there are more orders from the same client for following years) or lost.
Therefore, to track retention, we use a binary system in: order (1) and no order (0). By using this binary system, this model analyzes every unique client’s ordering pattern over the course of the study. The duration of the study is the number of the years (n) this model monitors the clients.
The world of possibilities
If we are looking at any companies to see if they have purchased our service or products over two years, say 2011 and 2012 for example, there are four possible scenarios:
They have had orders in
both 2011 and 2012
2011 but not in 2012
2012 but not in 2011
neither 2011 nor 2012
The following table illustrates these scenarios. Green for order and red for no order:
The first three scenarios are companies who have had business transactions with our company. They are the ones we are trying to study and track using this model. How about scenario number 4? A company that hasn’t had any orders with us at all falls under this category. That can include any companies outside of our client’s portfolio. Hopefully, we are able to narrow this scenario down to a definite number of potential, identifiable clients within our industry and can work with our sales, marketing, and business development teams to learn how we can target and eventually acquire them.
In summary, in an industry with clear boundaries (i.e., a definite number of identifiable clients) scenario 4 is the number of all the possible clients within the industry minus the ones we already acquired (the first 3 scenarios).
Scenario 4 (whatever is left blue) = Industry – Market Share
For the purpose of this model, we just study the first three scenarios out of four possible scenarios (or 2^2 – 1 = 3 scenarios).
What if we decide to look at our client’s list over a course of three years? In this case, if you write down all the possibilities to cover all the existing (or potential) clients, we will end up with eight scenarios. Again, there’s a scenario with no orders at all (exactly like scenario 4). So, we would only be interested in seven different possibilities (or 2^3 – 1 = 7 scenarios).
In general, for any number of years (n), using this model, we will end up with (2^n –1) scenarios.
Table of Possibilities
In this model, we study the client list over four years, 2011 to 2014 for example. Four years seems to be a good time window to capture the cyclical and seasonal nature of all the other businesses.
Any clients over the four years of this study will fall under one of the 15 scenarios (2^4 – 1 = 15). After categorizing clients based on their ordering pattern, a table of possibilities is a good tool to illustrate these scenarios with the actual number of clients:
According to the table of possibilities, the total of the entire client list is 8,385 for last four years. These clients have been divided into fifteen categories.
The last column shows the percentage of the clients in each category in relation to the total number (8,385 in this example).
To understand this table better, let’s go over an example. Scenario 10 accounts for clients with no order in 2011, who were acquired in 2012 and retained for another year but who didn’t have any orders in 2014. According to this table, there are 600 clients or 7% of the total list that fall under this scenario.
At the bottom of the table, there is another category, outside of the fifteen defined categories, for “one-time clients.” This group counts customers that used our company’s services/products only once and never came back for more. To calculate the number of such clients, simply add scenarios 1, 11 and 14 (those who have orders with the company only once in 2011, 2012 and 2013, respectively). According to the table of possibilities, 2,500 clients or 30% of all clients didn’t stay with the company for more than one year.
Scenario 15 also accounts for clients with a one-time order in 2014. But since we don’t have any data for 2015 yet, there is no further information about the ordering pattern of these 1700 clients acquired in 2014. Although scenario 15 doesn’t add any more information to our one-time clients category, it brings up an interesting fact to our attention: 1,700 clients or 20% of our client list are completely new and acquired in 2014.
Another interesting scenario to look at is scenario 8. This scenario counts all the clients that were acquired in 2011 and that have stayed with the company ever since. According to the count in scenario 8, only 800 clients or 10% of the total list have been loyal customers since 2011.
Client Flow Statement
In order to summarize and better utilize the data in the table of possibilities, we can develop a client flow statement. If you are familiar with a cash flow statement, a client flow statement is exactly the same concept but instead of tracking each unit of money, we are tracking the flow of each client. The following is a client flow statement developed using data gathered from the table of possibilities.
The first line represents the beginning number of the clients for the year or the ending number of the clients from the previous year. Since we don’t have any data from year 2010 in this example, we don’t know the beginning number that 2011 started the year with.
The second and third lines count the number of dropouts (or lost customers) and new clients (or acquired customers), respectively.
Line number 4 counts the number of clients returning after not having any orders with the company at least for one year.
The sum of the first four lines gives us the ending number of clients (line number 5) that the company finishes the year with (as long as we consider the number of dropouts a negative number).
Line number 6 is presenting the year-over-year growth percentage in the ending number of clients.
Lines number 7 and 8 show the percentage of new clients and dropouts in relation to the ending number, respectively.
At the bottom of the client flow statement, for the purpose of analyzing the relationship between the client flow and the company revenue, we have presented the total revenue for the year in line 9.
Line 10 shows the year-over-year growth percentage in revenue.
Finally, the last line calculates the average dollar value that each client generates, or the total revenue divided by the ending number of clients.
Data Analysis: What does all this mean?
Observation #1: Obviously, as we expected, the total number of clients at the end of each year has a direct relationship with the number of new clients and has an indirect relationship with the number of dropouts.
Observation #2: In this example, the number of new clients has a high co-relation with the number of dropouts (after using a regression analysis between these two variables, I ended up with an R-square value of 0.94). This means that when the company is doing well in attracting new business, it’s also doing better in retaining those customers. Sometimes, when the company is more focused on new clients and new business, the number of dropouts might increase and vice versa. But in this case, the number of new clients and the rate of dropouts are moving together in the same direction. In examples like this, these two variables can be independent from the performance of the company. They might be a variable of external elements such as economic growth or recession, etc.
Observation #3: As is noticeable, the average rate of dropouts is 40% from 2012 to 2014. Depending on the industry the business is operating in, this can be a lot. If the business is still profitable, it’s at the mercy of our one-time clients stopping by our business for a short while. Again depending on the industry, a 40% dropout rate can be alarming.
Observation #4: As predicted, the number of clients has a high co-relation with the total revenue of the company. An increase in the number of clients has a direct relationship with an increase in the revenue.
Observation #5: From 2011 to 2013, the ending number of clients has a high co-correlation with a decrease in unit revenue per clients (with an R-square value of 0.8). This means attracting more clients might have been a result of a decrease in the unit price of the product or the service while the total revenue has increased. Using an optimization model to calculate the optimal price in relation to the number of clients could improve the performance of the team in attracting new clients and retaining the existing ones, as long as the company stays profitable. The decrease in the unit price of products/services can be introduced in the form of discounts/sales for potential clients or special treatment offered to current, loyal customers.
In 2014, the company has the highest number of clients and revenue unit per client. If not because of external factors, such as a growing economy, etc, the good practices of the sales team can be studied and followed.
What else can you tell from the Client Flow Statement?
This is not an academic paper or a case study. It’s just an illustration of an actual model used to analyze the performance of a sales team with respect to customer retention. Unlike an academic paper, there’s no history, background, literature review, or bibliography to be added here. This study is just an observation and explanation of a model used in an actual job assignment by a financial analyst.
For confidentiality reasons, none of the numbers presented in this illustration are actual or from any existing, operating enterprises. They are solely examples to help with understanding the concept.