CLTV Model

Use

A CLTV model forms part of the customer lifetime value analysis (CLTV analysis) and describes the customers of a segment across the customer life cycle. It uses the following key figures:

·        Profit per customer

·        Customer retention rate (standardized customer retention rate as well)

·        Customer lifetime value (discounted CLTV as well)

In a CLTV model, you specify attributes such as the characteristic that is to be used when the customer segments are considered and the lifetime periods that a customer lifetime cycle is to be divided into. You can then use the setting you make in the CLTV model as the basis for calculating the key figures or entering them manually.

In a typical scenario, you first calculate the key figures automatically using historic data and then adjust them manually (for example, by adding values for noncalculated lifetime periods based on known values).

You can use a CLTV model as the basis for calculating predictions about the future development of the customer segment during a CLTV prediction.

Integration

You use a single transaction for CLTV modeling to execute all the CLTV analysis functions found in SAP BW.

Prerequisites

·        You need to be assigned to the role Customer Value Analysis ( SAP_BWC_CUSTOMER_VALUE) in SAP BW. You are then authorized to call up the transaction for CLTV modeling from the user menu (by choosing Customer Lifetime Value ® Customer Lifetime Value) and to call up queries for displaying the results.

·        In SAP BW, you can use for data collection data sources that deliver the data at the customer, segmentation characteristic, period, profit, and customer since levels. The "Customer Since" date marks the start of the customer relationship and must exist for each customer for the profit to be assigned in the corresponding lifetime period. The "Customer Since" date should be stored in an attribute for the customer.

Features

You first need to specify some parameters for the calculation by a CLTV model. You can create different models for different purposes.

Settings for the CLTV Model Calculation

Under Segmentation Characteristic, you specify the characteristic to be taken as the basis for the customer segmentation. You can choose any of the InfoObjects in SAP BW. However, the InfoObject you choose must be included in the data source that you assign for the data collection.

If you used the data mining method Clustering to perform customer segmentation by customer behavior, you now have to specify which InfoObject was used to store the determined clusters when the prediction results were uploaded into SAP BW.

The number of Months in a Lifetime Period is heavily influenced by the industry you are involved in. The number of months is usually less for an automobile component supplier than for a construction firm.

To adapt future values for CLTV to inflation, for example, you can discount the CLTV with a discount factor.

Alongside these general settings, you can specify other parameters under Calculation Settings.

You can enter a past period for the calculation (in the field No. of Previous Months to be Read). The months are counted back from the date that you enter as the key date for the calculation in the Calculate tab page. The number of lifetime periods considered in the CLTV model depends on the number of months that you enter for the calculation and on the number of months in a lifetime period.

To calculate the customer retention rate, it is necessary to specify after what length of time a customer is to be considered lost (in the field No. of Inactive Months for a Customer to be Considered Lost). This can vary greatly from one industry to the next.

You can consider all customers in the CLTV analysis or just those that you acquired as new customers at the start of the calculation period. By restricting the analysis to new customers, you can gain a more precise picture of how the lifecycle of specific customers has developed. If, for example, you carried out a campaign in the period under consideration, you can analyze how many of the new customers acquired as a result of the campaign have been retained as customers over time. You determine the exact period in which the customers were acquired. You determine this period in two ways: firstly, by specifying the key date for the calculation and the calculation period, and secondly by entering the number of months preceding the key date and during which the new customers are to be considered. The system first calculates the calculation period preceding and closing with the key date. The date determined for the start of the calculation period is then taken as the starting point of the subsequent number of months (within the calculation period) during which new customers are to be considered.

Entries: Key date = 01/01/2001; calculation period = 12 months; number of months for the consideration of new customer = 3. In this example, the development of all new customers acquired between 01/01/2000 and 03/31/2000 is considered over the entire year 2000.

The data source that you base the CLTV model on should contain information at the customer level including the "Customer Since" date, that is, the date when the customer relationship began (see the section Prerequisites above). Under InfoObject for Customer or "Customer Since" Attribute, you specify the InfoObjects that provide this information. The InfoObject for the "Customer Since" date must be an attribute of the InfoObject for the customer. For the business partner ( 0BPARTNER), such an attribute( 0CRM_CUSSIN) is available from SAP CRM Release 3.0. You can fill this attribute with the date of the first sales order, for example. In cases where there is no "Customer Since" date for a customer, the profit made with the customer is posted to the lifetime period 0. This lifetime period is not displayed in the CLTV model.

You assign the data source used for reading historic data (mentioned above) when you define data collection. For information about how to proceed, choose the question mark symbol  in the Data Collection dialog box. For the calculation of profit per customer and the customer retention rate, you can define either just one data source for data collection or different data sources in each case.

If you have made all the settings for the CLTV model, you can then use the tab page Calculate to schedule the calculation for background execution. Alongside the key date for the calculation, you can also specify whether the data collection should be processed concurrently. This can have a positive effect on the speed of the calculation as well as on resource consumption. It makes sense to select the Read and Process Data Concurrently indicator if you can use several application servers and if several data sources need to be executed. The system then distributes the queries across the available application servers using the server group that you specify, and the queries are then executed in parallel on those servers.

Result of CLTV Model Calculation

If you executed the calculation directly, it can take a while for the result to be displayed in table form and graphically in the tab page CLTV Model. The results table lists the different lifetime periods (LTPs) for each segment determined (technically the calculated values for the segmentation characteristic). For each combination of lifetime period and segment, the determined profit per customer, the customer retention rate, and the standardized retention rate are shown, as well as the CLTV and the discounted CLTV. The standardized retention rate expresses how many customers have been retained since the first lifetime period. The determined results are saved to the ODS object CLTV Models ( 0CRM_OLVM). You can also display the result with the query CRM CLTV Models ( 0CRM_OLVM_Q0001).

You can manually adjust some of the determined values (apart from the CLTV, the discounted CLTV, and the standardized customer retention rate). If no values could be determined for a lifetime period, the corresponding row in the table is then missing. During manual postprocessing, you can insert a row for these lifetime periods and extrapolate the missing values. The column Other Costs is not filled during the calculation. You can use this column for any additional costs that you were unable to determine with the data source. Alternatively, you can simulate the effect that costs for a marketing campaign have on the CLTV of a given segment. This assists you in making investment decisions.

The following table uses a segment to demonstrate how the values for the individual columns are calculated. The discounted CLTV takes into account the discount factor you entered. This factor is not included in the following table.

Example of a CLTV Model

LTP	Profit	Retention Rate	Standardized Retention Rate	CLTV
1	1000	1	1	1000
2	1000	0.5	1 x 0.5 = 0.5	1000 + (0.5 x 1000) = 1500
3	2000	0.5	0.5 x 0.5 = 0.25	1500 + (0.25 x 2000) = 2000
4	1500	0.2	0.25 x 0.2 = 0.05	2000 + (0.05 x 1500) = 2075