Customer Lifetime Value Essay
Customer Lifetime Value
Marketing Engineering for Excel is a Microsoft Excel add-in. The software runs from within Microsoft Excel and only with data contained in an Excel spreadsheet. After installing the software, simply open Microsoft Excel. A new menu appears, called “MEXL.” This tutorial refers to the “MEXL/Customer Lifetime Value” submenu.
Customer Lifetime Value (CLV) represents a metric of a customer’s value to the organization over the entire span of that customer’s relationship with a firm. Short-term sales influence CLV, but so do overall customer satisfaction, the churn rate in the segment, and the costs to acquire a new customer and retain an existing customer.
The CLV approach helps firms answer such questions as:
How much is my customer base “worth”?
Taking into account observed churn rates, how many of the currently active customers will still be active in a few years?
How much is a customer worth, depending on the segment to which he or she belongs?
If acquiring a new customer costs $150, after how many periods can we recoup this investment?
Customer lifetime value analysis considers your database at a segment level, using the answers you provide to the following questions:
How many segments do you have in your database, and how many customers per segment?
For a given period, how much is a customer worth, on average, in each segment (margins and costs)?
What is the likelihood that a customer in segment A will switch to segment B during the next period?
A CLV analysis allows you to use your own data directly or a template preformatted by the MEXL software.
The next section explains how to create an easy-to-use template to enter your own data.
If you want to run a CLV analysis immediately, open the example file “OfficeStar Data (CLV).xls” and jump to “Step 3: Running analysis” (p. 4). By default, the example files install in “My Documents/My Marketing Engineering/.”
Creating a template
Using the interactive assistant
In Excel, if you click on MEXL CUSTOMER LIFETIME VALUE CREATE TEMPLATE, a dialog box appears. This box represents the first step in creating a template to run the CLV analysis software. The first dialog box prompts you to use an interactive assistant.
Unless you are already familiar with the methodology, you should select “yes.”
The first step of the template generation process requires you to label and list the segments that you want taken into account. Enter the names of segments to which a customer can belong. Press ENTER or click the “Add to list” button to add it to the “List of Segments.”
Note that a segment of “lost customers” always appears in your list. This segment has the following properties:
There is no activity by these customers (margins and costs equal 0).
It entails an absorbing relationship state. As soon as a customer reaches this segment, he or she stays there forever. In other words, there is 100% chance the customer stays in that segment in the next period, and all other transition probabilities will be equal to 0%.
After entering all your segments (at least one), click the OK button to proceed to the next step of the template creation process. Clicking on the OK button generates a template.
Not using the interactive assistant
You may skip this intermediary step and create a blank template. When you are prompted to use the interactive assistant, just click “no.” The following dialog box appears:
When you click OK, you generate a new blank spreadsheet. You must enter the segment labels manually in the spreadsheet.
In this example, if you update the names of the segments in cells B6, B7, and B8, the names of the segments automatically update in the other cells of the spreadsheet.
Entering your data
In this tutorial, we use the example file “OfficeStar Data (CLV).xls,” which in the default conditions appear in “My Documents/My Marketing Engineering/.” To view a proper data format, open that spreadsheet in Excel. A snapshot is reproduced below.
A typical CLV analysis spreadsheet contains:
Number of customers per segment. As of today, how many customers does the company have in each segment?
Gross margins, or the average margins that the company expects from a customer over each period (e.g., year, quarter), on the basis of the segment to which this customer belongs during that period. In the OfficeStar example, a customer who belongs to the “Warm Customer” segment should generate $15 of gross margins on average during the next period (e.g., first quarter).
Marketing costs, or how much money the company plans to spend per customer during the next period, according to the segment to which this customer belongs at the beginning of the period. Typically, active customers are followed more closely, receive more attention (e.g., direct marketing solicitations, sales representatives visits), and cost more to the firm.
Transition matrix, which summarizes the likelihood a customer will switch segments during each period. This matrix should be read horizontally, and each line sums to 100% (because all customers appear in some segment). In the OfficeStar example, an active customer has a 75% likelihood of remaining in the same segment and a 25% chance of switching to the warm customer segment.
A customer’s behavior during the previous period determines into which segment that customer is classified, and his or her segment membership then determines the marketing dollars the company should allocate to that customer in the next period. In the OfficeStar example, a customer who belongs to the “Active Customer” segment generates $90 of gross margins per period (e.g., quarter).
After entering your data in the Excel spreadsheet using the appropriate format, click on MEXL CUSTOMER LIFETIME VALUE RUN ANALYSIS. The dialog box that appears indicates the next steps required to perform a CLV analysis of your data.
CUSTOMER LIFETIME VALUE – V130522
Number of periods: Specify the number of periods for which you want a detailed CLV analysis. Note that this choice does NOT affect the CLV computations, because the value of a customer always gets estimated over an infinite time horizon (though as time passes and discount rates apply, future revenues have less relative impact). The number of periods affects only the level of output.
Discount factor: Indicate the discount rate to apply for the value of a dollar spent or received in the future as compared to the current period. A discount rate of 15% means that $100 profit in the next period is only “worth” $85 in today’s dollars. A greater discount factor reduces the impact of future revenues on CLV computations and thus focuses on shortterm profits. You should increase the discount rate for turbulent or rapidly evolving markets, in which conditions change rapidly and future revenues therefore are highly uncertain.
Setting: Select either Transactional or Contractual depending on the nature of the product or service you are modelling. Contractual models imply the existence of a contract between the transacting parties (e.g., a mobile phone contract between the provider and consumer). Contractual relationships imply continuous transactions and a known end to the contract. Transactional models imply discrete transactions with no implied end to the relationship. For use with our CLV model, the impact of this setting will affect the first period of the analysis. A Contractual setting implies no loss/gain in first month (since the customer is under contract) while the Transactional setting will reflect loss/gain in the first month.
The discount factor gets applied after each period, regardless of how you define a period.
If you define a period as a quarter, a discount factor of 15% translates into an effective yearly discount rate of almost 48% (15% discount rate applied four times per year). Remember to take this multiplicative effect into account when selecting an appropriate discount factor.
After selecting these options, you must select the cells containing the data. First, the software asks for ranges of the current segment sizes and profits and costs for each segment, including a row dedicated to permanently lost customers. If you use a template generated by Marketing Engineering for Excel, it has already pre-selected the cell ranges.
Second, the software asks for a square range that shows the likelihood that a customer in each segment (row) will switch to each segment (column) in the following period.
The newly generated spreadsheet contains the results of your CLV analysis.
Interpreting the results
Customer lifetime value
The last column of the CLV table outputs the expected CLV of a customer who currently belongs to a given segment, determined by summing the stream of all future gross margins, minus all future marketing costs, and taking into account both the discount factor and the likelihood of customers switching from one segment to another.
These figures also appear in the “Lifetime Value” chart, shown below.
A customer with a negative CLV actually means a loss of money for your firm. Number of customers per segment
The next table (and chart) shows how many customers will be in each segment at each period of time in the future. The time horizon displayed on the chart matches the number of periods you specified in the “Run Analysis” options.
Note that the “Lost Customers” segment is not displayed. In most applications, all customers eventually become lost customers, and over
sufficient time, all other segments become empty.
Customer base’s lifetime value
The third table in the “CLV Analysis” sheet, labeled the Customer Base’s Lifetime Value, summarizes the future stream of revenues and marketing costs over a specified number of future periods (whether cumulative or not) at the global level. Some key elements of this table plot in the third (and last) chart of the spreadsheet.
In particular, the Discounted Net Margins (cumulated) provide an answer to the question: “Over the next x periods, how much is my customer base worth?”
The final tables depict the likelihood that a customer will belong to any segment in any period of time in the future, depending on the segment to which he or she currently belongs. There are as many tables as there are segments in the analysis.
In most applications, all customers eventually join the “Lost Customers” segment. The probability of belonging to that segment thus slowly reaches 1 (100%), and the probabilities of belonging to any other segment trend toward 0 over time.