Page 57 - DMGT308_CUSTOMER_RELATIONSHIP_MANAGEMENT

Customer Relationship Management

Notes 5. Building the loyalty programme: The structure, payout levels and reward systems associated
with the loyalty programme may then be constructed; the intention being to drive desired
behaviour from the organisation’s most profitable customers.
6. Estimating the costs to establish proposed loyalty programmes: The costs to build, maintain
and improve the loyalty programme should be ascertained.
7. ROI Model: This model should record investments, incremental profit expected,
unredeemed rewards, and liability projections over time. By measuring ROI on an ongoing
basis, organisations can ensure that their loyalty programmes continue to reward profitable
behaviour. As a rule of thumb a well-crafted loyalty programme should break even in its
first year and cover the set up/implementation costs. It is estimated that a well-designed
programme should cover internal rates of return in the second year, and continue to
improve in the third year and beyond.

8. Testing: Prior to the full implementation of the programme it is advisable to carry out
testing using a select group of customers or special focus groups.

Data Mining

Data mining can provide a better understanding of customer behaviour and provide insights
into ways of reducing customer defections and churn rates. Gordon Linoff (2004), founder and
principal of Data Miners, describes graphical techniques for plotting “hazard probabilities”
which reveal the patterns underlying customer churn rates for subscription paying customers.
By correctly stratifying the available data, the clarity of the output was greatly improved. Linoff
described, as an example, a scenario where it became evident that customers who paid by credit
card were the most likely to continue on the books of the organisation. Other patterns revealed
were:

1. Initial spikes in the dropout rate of customers who were due to matters such as poor
customer information being gathered at the point of sale, or perhaps by buyer’s remorse.
2. After 60 days there was a very strong peak related “forced churn” due to non payment
action being taken.
3. After 90 days there was a significant peak due to customers leaving at the end of promotions
when the full fees were applied.

4. After 120 days the probability of customer loss gradually continued to decline which
underscored an important facet of customer loyalty i.e., the longer customers remained
with the company, the less likely they were to leave. The long-term decline in hazard
probability was therefore a powerful indicator of customer loyalty.

Grading

John I. Coffey and Gene Palm (2005), principals of Profit Resources consulting company, give
examples of CPM in a banking environment. To effectively assess the value of particular
customers, decisions need to be made concerning how these should be graded e.g. as individual
customers or by household.
The standard by which to assess a customer’s value also needs to be ascertained. Customers can
then be ranked according to profitability, and placed into quartile ranges, with the “A” range
being the top group and “D” being the lowest group. It is beneficial to grade customer behaviour
in a number of ways, and a wide variety of variables can be used to achieve this e.g.:

1. Total profits;
2. Total deposits;

52 LOVELY PROFESSIONAL UNIVERSITY