CRM Analytics
CRM Analytics
CRM Analytics
Customer Relationship
Management Analytics
Introduction
This chapter introduces the role and benefits of customer relationship
management (CRM) and the historical development of the CRM
concept. In addition, it discusses how to measure and manage customer
value and customer lifetime value (CLV) by presenting various approaches
and principles. Besides, it introduces two types of strategies, namely,
“across-customer” strategies and “within-customer” strategies, which help
companies maximize CLV. The chapter also discusses the importance of
CLV-based marketing initiatives, which channel firms’ efforts toward cus-
tomer-centric marketing and away from product orientation. After intro-
ducing CRM solution providers, the chapter concludes with two cases
that illustrate how CRM technologies can help firms grow their business
and play a key role in achieving long-term success by transforming the
whole company.
The former Union Trust Bank in San Francisco, now Wells Fargo.
Firms such as Wells-Fargo Bank have invested millions of dollars for
hardware as well as employee training to effectively manage their
customer relationships in order to maximize their profits.
Source: Wikimedia by Kjteil ree.
The CRM process begins with customer selection (based on the mission
statement of the firm), developing the value proposition to serve them,
monitoring nonfinancial metrics of performance, and finally making sure
that these metrics are ultimately reflected in financial performance, that
is, customer profitability. However, in many cases, switching from mass to
relationship marketing may also require a transformation of the culture of
the firm from the top all the way to the bottom. The technical aspects of
CRM, although expensive, are relatively easy to handle. It is the human
105
unfounded: (a) the costs of serving long-life customers are less, (b) they
pay higher price premiums, and (c) they spend more. Therefore, behav-
iorally loyal customers are not necessarily more profitable than other cus-
tomers.10 Firms need to be able to differentiate their customers by their
(current and potential, long term) value and implement CRM initiatives
that are aligned with providing the customer better value.
Small volume customers can be profitably reached via direct market-
ing campaigns. A well-crafted direct marketing campaign is not wasteful
but can be very effective. For example, Dell considers their direct mar-
keting skills as a key competitive advantage, more so than their products
and manufacturing.11 Customer value is not built one transaction at a
time but rather through a series of mutually successful episodes. Some
episodes will inevitably include failures. The way the firm handles these
failures and recovers from them can create even more loyal and devoted
customers. Yet, roughly half of the customers with complaints are not
happy with the way companies handle their complaints.12 In the airline
industry, it has been shown that customer complaints can destroy firm
value more than satisfaction contributes to it.13
Point to Ponder: Why do you think that CRM strategies are successful?
Various metrics are available for measuring the value of the customer
and implementing the CRM strategies. Some traditionally used metrics are
recency–frequency–monetary (RFM) value, share of wallet (SOW), and
past customer value (PCV). These existing metrics have several limitations
and do not link the value of a customer to future profitability. On the other
hand, CLV is an advanced metric that is widely gaining popularity across
all sectors and industries. The forward-looking CLV metric considers the
future value of a customer to the firm and aids in designing and implement-
ing marketing strategies for the present, thus maximizing profitability.15
RFM Approach
On the basis of the assumption that past purchase behavior of custom-
ers better predicts their future purchases than demographic data, RFM
scoring models have been used in marketing for more than four decades.
Historically, the initial applications were in business-to-consumer (B2C)
contexts: direct marketing, insurance, banking, telecommunications, and
so on. The customers are ranked based on their R, F, and M score. Recent
purchasers, frequent purchasers and big spenders, on average, receive
higher scores. The higher the score is, the more profitable is the customer,
which is expected to continue in the future. A basic ranking, within par-
ent cell ranking, or weighed cell ranking can be used.
In basic RFM, the customers are scored on each factor separately. That
is, the customers are ranked based on the recency of their purchases and
then split into subgroups (e.g., quintiles). As a result, there could be five
quintiles where the top quintile represents the 20 percent that has made
the most recent purchases. The procedure is then repeated for frequency
and total money spent and reveals 5 5 5 = 125 cells from highest to
lowest ranking.
Within parent cell ranking, the cells are grouped based on recency,
similar to the basic approach. However, each group is then subgrouped
based on frequency (rather than separately). Each of the resulting 25 cells
is finally ranked by monetary value. This approach requires more sorting
than the former.
In weighed cell ranking, weights are assigned to the metrics and sorted
accordingly. These weights can be equal; however, an often used weight
scheme is: (3 × R) + (2 × F) + (1 × M). The weights used depend on the
data and specific application. For example, based on the response rate, M
may turn out to be the most important factor in some cases.16
Share of Wallet
SOW refers to the share of a company of a customer’s expenditure for
a given offering category. Although market share has been shown to
uniquely impact the financial performance of the firms based on their
strategy types,17 and market share and SOW are linked, targeted efforts
to increase the share of large wallet customers can have more direct and
positive bottom line impact. For example, major casinos “comp” their
high rollers (or “whales”) with the hope to attract them to gamble in
their facilities whenever they visit, that is, they hope to attract close to
100 percent share of gambling wallet. In other cases, firms target much
smaller shares. A local restaurant would be happy to consume $50 a week
of the dining wallet of a customer who spends $600 a month ($50 × 4
weeks/$600 = roughly 33% share of dining wallet). The data is typically
acquired through primary surveys, which is then combined with other
predictors of behavior in a database (behavioral and demographic vari-
ables) and share-of-wallet is then modeled as the dependent variable.
Sj
Share-of-Wallet:
j =1 Sj
J
where S is the actual sales to the focal customer, J refers to firm in cate-
gory, and the denominator represents the value of sales made by all firms
to the focal customer.
PCV = t =1GCit * (1 + d )
T t
Repeating the exercise for all the customers in the portfolio would
enable prioritization of all customers and allocation of marketing
resources based on their values.
A Primer on CLV
A main advantage of CLV approach is that it incorporates not only the
future expected value of a customer but also the probability of that customer
remaining active in future. In contrast with the former methods, CLV
enables optimal allocation of resources at the individual customer level.
( pt − ct )rt − AC
CLV = t =1
T
(1 + d )t
where
The margin (m) has been found to be relatively stable over time.22
However, we can also incorporate a growth rate to the equation (assuming
1 + i is larger than r(1 + g)):
r
CLV = m
1 + i − r(1 + g )
(S i,t +k )
− ci,t +k − Bi,r ,t +k − Bi, AO,t +k
1 + d
where
where
(1 + d )n
where
of the customer remaining active is 50 percent for the next month and
35 percent for the month after that, we can calculate the NPV of EGC
as follows:
where
Customer Selection
Strangers Barnacles
• Ensure profit on each • Improve profitability by investing
Low
transaction; if not outsource for share of wallet when wallet is
large
Short-term Long-term
Customers
customer via phone call for high-CLV customers and e-mail contact
to low-CLV customers may be more justified. However, other choices
may not be as straightforward, for example, a middle-CLV category may
by-and-large prefer e-mail over direct mail because of their Internet com-
petency. This approach can also be used to develop customer profiles.
For example, Figure 7.2 categorizes customers into four groups based on
122 rELATIONSHIP MArKETING rE-IMAGINED
Some advanced models even allow for heterogeneity at the customer level.
For example, one particular high-CLV customer (whale) may refuse to be
contacted by phone at all times and only visit Las Vegas during a cer-
tain season and so on. Recent developments in modeling have enabled
marketers to estimate optimal marketing spending at the individual level.
There is a need to estimate the interpurchase duration of each customer,
cash flows from each customer, and then maximize profits accordingly.
Then the firm would know what the optimal time interval to contact a cus-
tomer is, as well as the best mode of communication with him or her. More
contact is not only costly but also not always better and can create tedium.
Given the tremendous amount of data collected by retailers, it is pos-
sible to correlate a customer’s purchases with those who have bought the
same or similar products and predict what he or she might buy next.
The more relevant the communication is, the higher is the likelihood of
the customer to respond to the communication with a purchase. There-
fore, the sequence of purchases, timing, and value of the next purchase
needs to be estimated.31
and customer equity and increase sales, market share, cash flow, ROI,
and ultimately the value of the firm as measured by market capitalization
(shareholder value).35
Implementing CLV
Implementation of CLV to B2C contexts is harder, as gathering the
necessary data from all customers may be cost-prohibitive and time-
consuming. Furthermore, the firms that rely on channel intermediaries
may not be in direct contact with their customers. In fact, developments
in CRM technology and availability of computing power provide oppor-
tunities (through sampling if not for the entire customer base) for large as
well as small firms for both B2B and B2C contexts. Therefore, the context
to consider and plan for becomes B2B2C. The challenges of legacy CRM
systems and the rationale to upgrade them include their high mainte-
nance costs, minimal flexibility, limited functionality, poor integration,
and minimal analytical capabilities. For example, one European carrier
reportedly spends $2M to manage its 500,000 members using a legacy
system, that is, $4 per member every year spent just to maintain the
system. Moreover, it takes six months for the carrier to simply change
a tier requirement. The same feat takes only a day with contemporary
systems. Nevertheless, a primary reason for not successfully implement-
ing CRM is viewing it as solely a technology initiative.36
Point to ponder: Can you think of cases where the customer does not
want to build a relationship with the marketers or brand? What are the
societal implications of abandoning relationship marketing in these
cases? Should building relationships be-all and end-all of marketing?
Point to Ponder: Do you think the state of B2B or B2C CRM is more
advanced? Why?
for almost a century. Over time, its magazine subscriber database has
become a valuable asset for the company. When the company segmented
its subscribers by level of usage, it found that, the more products a cus-
tomer has bought, the higher is the likelihood of further purchases. In
this case, the length and breadth of the relationship predicted future sales
much better than traditional sociodemographic segmentation variables.
Today, Trusted Media Brands, Inc. (TMBI) has offices in 42 countries and
sells books, magazines, music, video, and educational products to a cus-
tomer database of more than 140 million in 76 countries through direct
marketing. It publishes 75 magazines (nine with more than a million
circulations), including 49 editions of Reader’s Digest, the world’s larg-
est paid-circulation magazine, and sells approximately 30 million books,
music, and video products across the world each year. It also owns the
largest food web site in the world, Allrecipes.com, which has launched 15
web sites serving 20 countries. However, direct marketing was not enough
to save a highly leveraged RDA in a post-Internet world where print media
readership including that for magazines has been steadily declining. The
company filed for bankruptcy during the global recession of 2009 but
reemerged in February 2010. It is trying to reduce its debt substantially.42
Point to Ponder: How do you think TMBI should make the tran-
sition from direct marketing to CRM? What steps should the new
management take?
Key Takeaways
• Loyal customers are not necessarily profitable customers.
There is a need to segment the customers according to their
profit potential.
• CRM is a comprehensive strategy and process of acquiring,
retaining, and partnering with selective customers to create
superior value for the company and the customer.
• CRM has numerous benefits: it enables the marketer to
measure and manage to optimize the cost to serve customers;
profitability metrics can be tracked at the individual level;
it empowers the firm to determine which customers are not
130 rELATIONSHIP MArKETING rE-IMAGINED