Chapter 2: Customer Value Assessment and Valuing

Customers
In this topic, we deal with two related topics: First, we address how to determine the value of
customers associate with the products or offerings that a firm might provide. We call this first
issue customer value assessment. Second, we consider assessments of the value of specific
customers, prospects and customer segments to a firm. This second issue, we refer to as
customer valuation

Recall that the market response model consists of customers and their behaviours. In this topic
we will understand and assess customer responses to actual and possible marketer actions. There
are three important words with vital in order for us to understand.

1. Product (offerings): anything that can satisfy human needs. A need is the difference
between a persons (or organizations) present state and a desired state. Thus, a need
includes a gap, a barrier, a problem or a specific want.

From marketers perspective, a product or offering is a set of designed attributes that satisfy
specific needs of specific customers. Typically, attributes can be organized into as many as
three components: physical, service and perceptual. When buying a BMW, a consumer
obtains a car with all its physical attributes, a warranty, and other service offerings, plus
the perceptions associated with the Ultimate Driving Machine.

We deal primarily with physical products almost all of which also have some service
components) and services (almost all of which have also physical components). Pure
services exist, of course characterized by simultaneous manufacture and consumption; for
example, an airline seat for a specific flight cannot be inventoried or resold if unused. In
addition, physical products may be classified as durables (e.g razors used for a significant
period of tie after purchase) and consumables (e.g. razor blades used just a few times0. We
also focus on products for both consumer uses (ie. personal consumption) and business
consumption (i.e for use as inputs in other products and services). Thus, our products
include consumable and durable physical products and pure services involved in consumer
as well as business-to-business (B2B) use.

2. Market: depending on the perspective (supply or demand), either the competitive arena in
which a particular product or service is sold or the segment of customers who require
particular benefits and solutions from the product or service.

When a market is defined in terms of the needs of customers and the benefits and solutions
to customer problems that the product or services provides, the definition takes a demand
perspective. For example, a market research study by the Polaroid Company in the US
suggested when 18-25 year old customers bought its least expensive instant camera; the
company was competing in an entertainment market. This is due to the camera were
bought to provide fun. If these customers did not purchase cameras, they spent their money

20
 going to the movies or engaging in other forms of entertainment, not on another type of
camera at the same price.

However, for practical reasons associated with the way firms manage their
transaction data bases, they usually adopt a product or supply side view of their market.
Thus, if Polaroid declares itself to be in the camera market, the company has adopted a
supply perspective of its market. This definition traditionally focuses on any competitors
who supply physically similar (versus functionally superior, which would imply a focus on
customer needs) products or services. This perspective also explicitly considers
manufacturing processes, cost structures, technology development and utilization,
marketing and distribution strategies, entry and exit costs, and so on. Therefore, a market
in this context is a competitive arena, which can make it easy to lose sight of customers,
who often become volume or market share rather than opportunity for company to satisfy
needs with products and services.

Consequently, we consider both demand (needs of customers) and supply ( products and
services offered to customers) perspectives of a market.

3. Customers: Persons who interact with the company to purchase and use products and
services.

Marketers must take a broad view of customers. They need to focus on those who are seeking
products and services to satisfy their own personal needs, whereas at other times, they must
consider the customer to be the person buying a gift, buying items for a household, or
participating in a purchase decision for an organization. To explore the broad nature of the
customer concept both consumer and organization buying concepts distinguished buying
centers.

Consumer buying center involves five possible customer roles;

i. Initiator: the person who suggest the idea of buying an offering
ii. Influencer: a person whose advice influences the purchase
iii. Decider: the person who decides what, when, where or how to buy
iv. Buyer: the person who makes the actual purchase
v. User : the person who uses the product or services

Organization buying center basically are having the categories of people who play similar
roles:

i. Influencers: those who influence the buying decision by providing information or advice
or defining the specifications for the product or service
ii. Approvers: people who authorize the purchase, either financially or technically
iii. Deciders: the people who make a (group) purchase decision
iv. Buyer: the person who actually places the order (often a professional purchasing officer or
possibly an automated computer program)
v. Gatekeepers: people who have the power to sanction or block a supplier or prevent the
flow of information among buying center members

21
 vi. User : the people who uses the product or services

In either case, one person any assume multiple roles, and different people often play different
roles in high-involvement purchase situations.

THE CONCEPT OF CUSTOMER VALUE

Value means different things to different people.

1. Value can be measured by what a customer exchanges (in economic terms) for various
options that can satisfy a want or a need.
2. We can measure value in terms of customer intentions or what a customer intends to
exchange or;
3. We might assume that the customer is a rational economic agent and therefore define value
as what a customer should (if he or she had all the facts and the ability to understand them)
be willing to exchange.

All three methods are hidden in the mind of the customers and must be revealed or inferred
through specific measurement. Although the general concept of customer value remains the same
for customers and organizations, the needs that drive value may differ considerably in these two
contexts. For example, individual consumers value products and services for their personal
consumption, whereas organizations value products and services for the competitive advantage
they provide in the marketplace.

For most customers (whether individuals or organizations) value is a function of the (perceived
or expected) benefits relative to the price (to be) paid namely:

Value= Benefits Price.

Both parts of the equations can be examined as follows:

Benefits
functional (e.g. the product does the job or functions well)
psychological (e.g. Nike shoes provide social status)
economic (e.g. AEON offers branded products at everyday low prices)

Price
monetary (e.g. RM)
perceived risk
inconvenience

22
Example 2.1: Consider the development and launch of the Lexus automobile. First, Toyota
identified a segment of potential customers who would like to buy a European brand (such as
BMW or Mercedes) but thought they were overpriced. If they could build a car to the design
standards of a Mercedes and sell it as a significantly lower price, it should appeal to these
customers. With this goal in mind, and this smart buyer as the target segment, Toyota
developed Lexus. Lexus is now sold globally and has become Japan's largest-selling make of
premium cars. The Lexus marque is marketed in over 70 countries and territories worldwide and
has ranked among the ten largest Japanese global brands in market value.

Figure 2.1(a) :The LS 400 flagship sedan launched in 1989, introducing Lexus to the world. (b) The first
Lexus hatchback, the CT 200h, premiered in 2010.

On the benefit side of the value equation, the car sold at a major discount relative to Mercedes.
To reduce the perceived risk of buying an expensive luxury car from Toyota, the car avoided the
Toyota name badge and was not sold through Toyota dealers. Establishing a separate dealer
network and training sales people to sell to this specific target segment increase d the prestige of
the brand (psychological benefit) and reduced the inconvenience of buying and servicing the car.
This great car value thus became an outstanding success in the market.

Value can range from negative to positive numbers, such that customers prefer those products
with high value more than they do those with low value. A negative value score applies to a
product that a customer would not accept for free. In some case like a donors kidney has
immeasurable value to the recipient who wants to stay alive. Many factors determine where
value falls along a scale, including the characteristics of the individuals or organizations need
and the availability (or scarcity) of options to satisfy that need.

Customer Needs and Value

We consider customer value from the perspective of an individual (or a single organization),
where customer value is driven by customer needs and the availability of options. We also can
distinguish between a functional need and a perceived need because what is actually required to
complete a job or perform a function is rarely the same as what the customer perceived as
required.

Once customer perceives a need, three general behavioral possibilities exist: ignore the need,
postpone it or engage in a purchase process to identify options to satisfy it. In Figure 2.2, these
options appear in the behaviors box. The customer searches for options to satisfy the need,
evaluates them, chooses one, purchases it and uses it. As the customer gone through the process,

23
the perceived need remains subject to change (the two way arrow) as the customer gathers
information about various alternatives. The phase of the process that involves an evaluation of
options occurs when the customer formulates more concrete notions of value associated with
satisfying the need.

Figure 2.2 : Customer needs and customer value measurement

A customers needs determine the value a market offering possesses in terms of satisfying them,
such that the closer the fit between an identified offering and needs, the greater is the value of the
offering to a potential customer. Therefore, confronted with an array of offerings, a potential
customer generally chooses the option with the greatest (positive) value. That consideration is
why customer value is so critical to marketers-it relates directly to customer choice. Of course,
actual purchases are subject to various constraints, including availability, the customers budget,
and other factors that might redirect the customer from a preferred option to a secondary one.

24
Thus, market measurements and analysis based on actual behavior may provide different results
than those based on perceptions of needs.

Exercise 2.1:
Figures a and b show the price and details of two packaged products from Gillette. Fusion is
highly demanded by customers despite its high price, as the blades are six times more durable
compared to Custom Plus and 50 percent more comfortable. How do you think that the packaged
value has been decided to be marketed?

Gillettes Custom Plus Gillettes Fusion

Disposable 8 Cartridges
Amazons price of 3 Pack with Cost/cartridge: $2.12
10 razors/pack: $10.61

Understanding Customer Needs

Understanding needs is central to an understanding of customer value. A need can be

characterized by its subject and importance, its temporal aspects, and its information
requirements.

1. Subject and importance of need

The subject and importance of a need are the primary bases for a persons perception of that
need. Figure 2.3 visualizes Maslows hierarchy of needs which vary in terms of their importance
to the individuals survival; physiological, security, social, status and self actualization. Needs
higher up in the hierarchy are more crucial than those father down. Example of application is
when NEC positioned its mobile phone as a safety device (especially for women) to place its
importance higher in Maslow hierarchy, often is a very effective positioning. And many soap
manufacturer emphasize how their product enhances the appearance and beauty of the user (a
higher-order need) rather than cleanliness (presumably, a lower order need) which is the primary
function of most soaps.

25
Figure 2.3: An interpretation of Maslow's hierarchy of needs, represented as a pyramid with the more
basic needs at the bottom

Customers processing of such internal and external stimuli generates motivation to fulfill
various need states. This motivational connection not only influences customer perception of
value but also reveals the importance of external communication (e.g. advertising, sales force,
sales promotion) in helping position a brand or offering in their minds.

In consumer language, the object of a need can be defined by levels ranging from very general
(e.g thirst) to a very specific (e.g. thirst to reduce body temperature or replenish lost fluids). A
need such as thirst also can be described in terms of physical, social, status, and other
dimensions. Physical dimensions might include factors such as sweetness, temperature, alcoholic
content, smoothness, carbonation and flavor; social dimensions recognize that thirst can be
satisfied within a family setting at a restaurant and so on. These factors thus characterize
customer needs and the kinds of offerings that would satisfy them. Both the characteristics and
importance of the specific need serve as the bases for customer judgments of the value of an
offering that quenches thirst.

2. Temporal aspects of need

The urgency, frequency and duration of a need, define its temporal dimensions. The
urgency of a need refers to the perceived amount of time within which the consumer
believes the need must be resolved. People who get severe headaches if they do not eat
when they feel hungry perceive such a need as urgent. Similarly, an organization that
depends on reliable responses to customer telephone calls may deem upgrading or
obtaining a new switching system as urgent once it reaches its calling capacity. The
frequency of a need pertains to how often the same need occurs. Thirst may occur several
times in a day, and an organizations need for a new telephone switching system may

26
 occur every five years. Finally the duration of a need refers to how long the state of
tension associated with that need lasts. Hunger or thirst may last for a few minutes before
being satisfied, whereas the need for a new telephone switching system may exist for
several months before the upgraded system is ready.

3. Information requirements associated with a need

The newness, complexity and clarity of a need help define the customers information
requirements associated with having that need satisfied. New needs generally require
more learning to make a purchase decision. Complex needs, with more characteristics
require more learning, involve longer buying processes, invite consultation with others,
and require consumer consideration of more alternatives. The clarity of a need is the
certainty with which a customer perceives the need and its characteristics. When needs
area unclear, consumers may engage in a significant amount of information search,
including evaluation and trial of the product, if possible, to better define the need and
how well alternative offerings satisfy it. Customer with unclear needs also may be more
susceptible to the marketing information they receive (e.g. salespersons advice).

Taken together, the subject, temporal and information characteristics of a need help determine
both the process the customer uses to satisfy the need and the value of various options available
to satisfy the need and the value of various options available to satisfy it. For example, the more
important and urgent a need, the more the customer values offerings that satisfy it immediately.
Customers who require significant information may favor offerings that provide necessary
communication, even if they are not objectively superior. Thus, the value of various options to
satisfy a need are driven by customer perceptions of that need, combined with how well
marketers can tailor the 4Ps of the marketing mix to satisfy it (with a Product), accommodate
the customers budget (in Price) and make the search process easy (through Promotion and
Place).

APPROACHES TO MEASURING CUSTOMER VALUE

Figure 2.4 illustrates several approaches to measuring customer value.

27
Objective Customer Value: Should-Do Measures

Objective value measures attempt to estimate the objective or true customer value of a product or
service offering. Thus it is not the perception of value by the consumer or other market
stakeholders (despite the importance of these perceptions) but rather the objective worth of the
product in terms of satisfying the consumers set of needs. If this value can be determined, it
provides an important reference point for a better understanding of how perceptions can enhance
or reduce the value estimates that consumers use to make choices.

Firms report using three types of objective customer value measures: internal engineering
assessment, indirect survey questions and field value-in-use assessments.

1. Internal engineering assessment

Some estimate of customer value rely on evaluations by the selling firms own managers and
engineers based on laboratory test, use of the product by the firms own employees (alpha test)
or computer simulations. Although the firm might not contact customer directly to ascertain the

28
value they ascribe to the offerings, it must obtain a very good understanding of the usage
situation and, in the B2B markets. The success of this method depends on how well the firm
translates this understanding into economic estimates of customer value. In situations where a
firm cannot interrogate its customer directly or secrecy is crucial, such method may be the only
viable means to assess customer value.

2. Indirect survey question

To overcome the limitation of little or no customer input in internal engineering
assessment, firms often query customers about the value they place on satisfying a need or
resolving problem. With this approach, sales people may ask company personnel about the effect
of one or more changes in existing offerings on certain aspects of their needs or problems. This
technique is indirect, in that the questions focus on the changes rather than the worth of the entire
product offering. These answers, typically combined in some way with other unknown
information, can create estimates of the customer value of each change to eth product offering.
For example, a potential customer of inkjet cartridges might simply be asked:If we
increase the life of the cartridges that you are currently using by 20%, how much would it be
worth to you? The success of this method depends on how well the questions are framed, as
well as how willing the customer is to provide valid and reliable responses.

3. Field value-in-use (VIU) assessment

This third method requires the customer and the supplier to conduct a joint value
assessment. Although most appropriate for B2B markets, it can be used effectively with
individual customers. For a suppliers current product, VIU is defined as the price that would
make a customer indifferent (e.g. economic breakeven) between continuing to use the current
product versus switching to another option. For a new product, VIU is the maximum amount the
customer would be willing to pay for the new product, given the extra benefit (beyond those
provided by the current product) that it offers.

To arrive at VIU price, marketers must develop a complete list of costs element associated
with the use of the incumbent product compared with the new product. They then can calculate
the values or costs associated with each element to arrive at VIU estimate of the product offering
for that application, usually on a cost per unit basis. The success of VIU approach depends on
identifying various cost elements, as well as the assumptions the firm uses to estimate costs and
in B2B setting, its ability to understand the impact of these cost elements on customers ability to
create value for their own end customers.
Determining the VIU of an offering may be the measure that comes closest to true
customer value, because it represents what a consumer should be willing to pay. Value-in-use
also requires the closest cooperation with the customer to determine how the product will be
used. Joint value assessments by the supplier and the customer often reveal surprising and
unanticipated cost savings. In this context, a VIU analysis offers the opportunity to train the
customer to understand the value that may be realized from smart purchase decisions. Many-
supplier customer collaborative arrangements have started with a VIU assessment.

29
Example 2.2

Suppose a chemical plant uses 200 O-rings to seal valves on pipe that carry corrosive materials.
The plant pays RM5 for each O-ring and must change them during regular maintenance every
two months. However, a new product has twice the resistance to corrosive materials. The
supplier of the new product can determine the VIU of the material according to different
methods.

Solution 1

1. Annual cost of incumbent product = 200 O-rings x 6 changes per year x RM5 per O-rings
= RM6000
2. For the new product, it must be changed only half as often or 3 times, hence the annual
cost of new product = 200 O-rings x 3 changes x VIU= RM6000.
3. Therefore the VIU = 2 x RM 5 per O rings = RM10.

Solution 2

The new material allows a longer time between maintenance shutdowns four months versus
two months- and the cost for a shutdown is RM5000. Therefore,

(200 x 6 x 5) + (5000 x 6) = (200 x 3 x VIU) + (5000 x 3)

Shutdown cost
Equipmen
Equipment cost Shutdown cost Equipment cost

And solve for the VIU, it is equal to RM35.

For a VIU calculation, it is extremely important to include all costs. For example, the equations
fail to include the switching costs, if any, the costs of changes to ordering policies and new risk
factors. Other cost to consider in such VIU calculations include,

Purchase cost
Fabrication cost
Finishing cost
Inventory cost
Maintenance or service cost
Scrap adjustment cost
Level of requirement adjustment cost
Changeover cost

Perceptual Customer Value: Plan-to-Do Measures

Even with limited information, customers can evaluate one or more offerings to satisfy their
needs. The form of perceived customer value often differs from an objective assessment of

30
customer value, as discussed previously. The difference is driven by psychological
considerations, such as risk perceptions and the particular information a customer has about the
different offerings. Assessment of perceived customer value generally employs a series of
customer measurements that question customers about their perceptions of and preferences for
various offerings as well as the attributes and benefits of these offerings. Such questions appear
as part of a questionnaire in a survey of customers. As shown in Figure 2.4, perceptual measures
can involve both constrained and unconstrained measures of customer value. Unconstrained
measure place few boundaries on customer value assessment (everything can be rated as very
important). Constrained measures establish parameters for customer value assessments (customer
rank alternatives or allocate fixed number of points across them), which means respondents have
to give up benefits along some dimensions to obtain gain on others.

Unconstrained question measures

a. Focus group
In a focus group, five to ten (potential) customers convene for a several hour discussion with a
trained moderator about the perceptions, attitudes, preferences and usage of a (usually new)
product or service. The moderator presents participants with potential product offerings or
concepts and asks them about the value or worth of various options to them (or their firm). The
goal of the focus group is to obtain a better understanding of the common or shared perceptions
of a well-defined target group of customers.
For example, suppose a carpet manufacturer wants to know if potential customers will
find value in a carpet with customizable color patterns. As a first indication of whether such
value exists, the company would ask focus groups of potential carpet customers to discuss and
evaluate alternative types of carpets for their homes. At a suitable point in the discussion, each
customer might individually write down his or her answer to a question such as, What would
you be willing to pay for this carpet with color patterns customized for you (or your firm)?
Customers then discuss their answers, and the moderator probes for information about factors
that influenced these valuations.
However, focus groups are limited for assessing customer value. In particular, the
potential; for groupthink can often lead to homogenous results. Consequently, focus group
results should be viewed as exploratory. Nevertheless, they often provide valuable insights about
various components of customer value, as well as the relative ranges of worth of an offering.
Perhaps more important, with the results from a focus group, marketers can gain better
understanding of the language that customers use to describe the value they derive from a
product category or offering.

b. Direct survey questions

Because focus group results are exploratory, they often suggest hypotheses about customer value
that firms can test with a broader-based market survey. Such a survey usually involves a sample
of customers who agree to complete a questionnaire that includes a description of one or more
potential product offerings or concepts. Respondents indicate the value or worth of these
concepts to them or their firm. A typical question in this kind of value assessment would be,
Given the situation, how much more than a standard carpet would you be willing to pay for a
carpet with customizable colors?

31
 As is the case with any survey data, appropriate respondents must represent the buying
population, have the knowledge to answer direct questions about the value for an offering, and
be willing to share that knowledge. The wording of questions their placement in the survey,
whether the survey occurs over the telephone or by mail, and other such considerations all can
introduce error into value measurement. Therefore, marketers must follow basic market research
data collection procedures to carefully ensure that the resultant data are reliable and valid.

c. Important ratings
Important ratings are among the most popular approaches to measuring customer value.
Typically, in a field research survey, respondents receive a set of attributes that describe a
product offering and rate (or rank) them according to their importance to the respondent or the
firm. Respondents also usually rate (or rank) competitive products on their performance
pertaining to each of the attributes, which offers the firm a sense of the customer value provided
by competitors.
This highly popular approach is easy to develop and execute and provides a simple, detailed
assessment of the importance of the various components of customer value. However, it also has
many critical weaknesses. First, respondents might not be able to discriminate effectively among
the various benefits, and because the ratings are unconstrained, some respondents will give top
scores to all the benefits listed. Second, the approach does not provide an easy means to
determine willingness to pay for these important options, because there is no indication of
relative value or any trade off in the level of performance of one benefit in relation to another.
Third such surveys offer no direct link to behaviour, because the items that survey respondents
rank highly often do not drive choice behavior in the market place.

Example 2.3

Suppose the customizable color carpet focus group identified several product attributes that
deliver customer benefits, that is, components of overall customer value. The carpet
manufacturer then asks customers to rate the importance of these identified benefits.

How important is each of the following benefits of personally customizable color carpet
to you? Please circle the number that best represents your rating, where "1" is not at
all important, "5" is very important, and the numbers in between range of importance
to you.
1 2 3 4 5
Allows me to express my self
1 2 3 4 5
Easy to mix and match designs
1 2 3 4 5
Will be attractive to my friends

Example 2.4

32
In a study of buying behavior for industrial heating and cooling equipment, all respondents,
including technicians who worked with the systems (and thus were important buying center
members) rated reliability as a critical buying criterion. However, closer inspection showed
that those technicians preferred newer technologies-state of the art of the equipment)- that they
themselves rated as unreliable. They answered the survey in a way they thought was best for the
company, which conflicted with their need for personal job satisfaction, which was not included
in the survey. As it turns out, they were excited and challenged by the opportunity to work with
new technologies- a need that drove their overall preference and product valuation.

Constrained question measures

1. Conjoint analysis
Conjoint analysis, one of the most widely used customer value measurement approaches,
employs a field research survey to ask respondents to provide their overall ratings for each of a
set of potential offerings. A set of attributes options or features describes each offering, and the
levels of these options vary systematically. Then, statistical analyses decompose these resultant
ratings into the value that the respondent attributes to each option. The value can be recombined
to describe new product offerings, such that each option of each attribute receives a valuation for
each customer studied.

2. Benchmarking
In a field research survey, respondents receive descriptions of product offering that represents the
best available competitive product or service, which thus serves as a benchmark. Respondents
indicate how much more they (or their firm) would be willing to pay for added product tributes
or features to this benchmark offering. Similarly, they might indicate how much less they would
expect to pay if selected attributes or features disappeared from the offering. The benchmarking
approach attempts to determine customers willingness to pay for each attribute or feature.
Although it is easier to implement than conjoint analysis, it lacks rigor for systematically
examining trade-offs among attributes.

Behavioral Customer Value: Have-Done Measures

A major limitation of objective and perceived customer value measures is that they represent
estimates based on information obtained or judgments made before a purchase. Thus, they lack
the credibility associated with measures derived from actual observations of purchase behavior.
Therefore, measures of behavioral customer value use observations of actual past consumer
behavior as a basis for estimating value.

1. Data mining

Many organizations keep extensive records of customer purchases in a form that leads itself to
statistical analysis. Sometimes these data can be cross-matched with other data pertaining to
customer characteristics that come from commercial sources, channel partners, or information
provided when the customer opened an account with the organization. By updating these data
periodically, organizations can analyze the information to produce segments according to

33
customer profitability, amount spent, lifetime value, the range of products and services required
and so forth. This type of customer analysis goes under the name of data mining.

2. Choice model

Choice models use past behavior (often along with associated attribute ratings) to infer or
estimate the value (or utilities) of product characteristics that might bets explain or predict actual
behavior.

Although marketers can ask people about what they claim they would do, actual purchase
behavior can provide greater insight into the true drivers of behavior-the customer values we
seek to measure herein. In recent years, the increasing availability of retail scanner data, internet
transaction data, and other customer purchase databases have provided increasing opportunities
to use choice models to infer customer values. The output of choice model is an estimate of
importance weights, along with the purchase probabilities of each market alternative for each
customer. A simple and powerful choice model is the multinomial logit model that states that a
customers purchase probability for brand A equals the (weighted) utility of A divided by the
sum of the customers utilities for alternative brands.

Brand Choice

Related to customer value is consumer behavior. Behavior can be predicted in terms of brand
choice. Brand choice models can usually be distinguished by how they deal with

(a) population heterogeneity

(b) purchase-event feedback
(c) exogenous market factors

As mentioned above, market factors are the third determinant of brand choice behavior and are
most frequently accounted for in one of two ways:
(1) Explicitly including the influence of specific market factors in the model
(2) postulating that the effect of such forces can be accounted for with a time trend or
structural shift in the model.

The specific functional form most commonly used to characterize individual choice behavior is
called multinomial logit model.

Aij
e
Pij (2.1)
Aij
e
j

Where Aij = attractiveness of product j for individual i

34
 = wk bijk
k

b = individual is evaluation of product j on product attribute k (product quality for

ijk
example), where the summation is over all products that individual i is considering purchasing;
and wk =importance weight associated with attribute k in forming product preferences.

Equation (2.1) gives the probability of individual choosing brand 1, or analogous equations may
be specified for the probabilities of individual i choosing the other brands. We can estimate the
importance of weights wk in a number of different ways, depending on whether we have
information on likelihood of purchase measures or whether we have observations of actual recent
purchase events. In either case the weights are often called : revealed importance weights. This
is because they are revealed by an analysis of the past behavior (e.g. choice) of consumers rather
than by directly asking consumers.

Example 2.5:

Suppose that someone performed a survey of shoppers in an area to understand their shopping
habits and to determine the share of shoppers that a new store might attract. The respondents
rated three existing stores and one proposed store on a number of dimensions: (1) variety, (2)
quality, (3) parking and (4) value for money. By fitting shoppers choices of existing stores to
their ratings through logit model, we can estimate the coefficients wk .

A j w1b j1 w2b j 2 ... w b (2.2)

k jk

Where

A j = attractiveness of store j
b jk = rating of store j on dimension k ; k=1..K; and
wk = importance weight for dimension k
Table 2.1 shows the attitude ratings by store from a group of similar customers and its
importance weights.

Attribute ratings by store

Store Variety Quality Parking Value for money
1 0.7 0.5 0.7 0.7
2 0.3 0.4 0.2 0.8
3 0.6 0.8 0.7 0.4
4 (new) 0.6 0.4 0.8 0.5
Importance weight 2.0 1.7 1.3 2.2

35
 a) Calculate the attribute value of each store
b) What are the market shares of existing store?
c) How much shares proportion will be drawn by the new store?

Solutions:

Store (a) (b) ( c) Share estimate (d) Share estimate Draw

Ai wk b jk e Ai without new store with new store (c-d)
1 4.70 109.9 0.512 0.407 0.105
2 3.30 27.1 0.126 0.100 0.026
3 4.35 77.5 0.362 0.287 0.075
4 4.02 55.7 0.206

Models whose parameters do not change over time are referred to as stationary-in-
parameters models; such models do not necessarily include stationary purchase probabilities,
which may change because of purchase feedback. Much of the information used in a number of
popular brand choice models can be extracted from a brand-switching matrix. The elements in
this matrix are obtained from purchase information on two choice occasions separated in time,
where the time between the two purchases may different considerably for different consumers.
There are several models that can be used to generate the brand-switching matrix.

Example 2.6- Based on purchase behaviour

The following table gives some data on the brand-switching habits of 513 consumers over two
purchase occasions. In this table, parts (b) and (c) show transition probabilities expressed as joint
and conditional probabilities, respectively.

Table 2.2(a): purchases of brands A,B and C over two purchases occasions: 137 of 513 people
bought brand A on both purchase occasions

Brand bought on occasion 2

A B C Total
Brand bought on

A 137 47 19 203
occasion 1

B 41 179 12 232

C 22 10 46 78

Total 200 236 77 513

36
Table 2.2(b): joint probability matrix: 137 of 513 or 0.267 of the population bought brand A on
both purchase occasions.

Brand bought on occasion 2

A B C Total
Brand bought on

A 0.267 0.092 0.037 0.396

occasion 1

B 0.080 0.349 0.023 0.452

C 0.043 0.019 0.090 0.152

Total 0.390 0.460 0.150 1.00

Table 2.2 (c): conditional probability matrix , of the 203 people who bought brand A on the first
occasion, 137 or 0.673 (137/203) bought it on the second occasion.

Brand bought on occasion 2

A B C Total
Brand bought on

A 0.674 0.232 0.094 1.00

occasion 1

B 0.177 0.772 0.051 1.00

C 0.283 0.125 0.592 1.00

Let, p( i ) = conditional probability that a consumer will purchase brand i on the second
j
purchase occasion given that brand j was purchased on the first occasion, then p(i,j) represents
joint probability that a consumer will purchase brand i on the second purchase and brand j on the
first purchase occasion

These purchase probabilities are related as follows:

p(i, j )
p (i / j ) (2.3)
p( j )

37
where p(j)= probability of purchasing brand j on the first purchase occasion; so p ( j ) m j , the
market share of brand j. It also follows from the definition of joint and conditional probabilities
that p(i. j ) m
j
i and p(i / j) 1 .
i
In this example in table, the joint and conditional

probability measures are obtained from information in part (a).

Example 2.7: Zero Order Model

A simple multiple brand model (Ehrenberg, 1972) is given by the joint probability of a consumer
purchasing brand i and j on successive purchase occasions as follows:

p(i, j ) kmi m j (2.4)

and p(i, i) mi kmi (1 mi ) (2.5)

where mi is the market shares of the respective brands, and k is calculated as such:

1 p (i, i )
k (2.6)
1 mi2
i

1 p(i, i ) 0.294
Hence, from Example 2.6, k 0.479
1 m 2
i 0.614
i

and the theoretical switching values can be calculated as follows:

Brand bought on occasion 2

A B C
Brand bought on

A 0.275 0.086 0.028

occasion 1

B 0.086 0.341 0.033

C 0.028 0.033 0.089

Note that the values are very close the values exhibit in Observed values Example 2.6.

38
Example 2.8: Markov Model
Given current market share, a Markov model can be used to predict how market shares change
over time. Suppose we know m it , the market share of brand i at time t; then market share for all
brands at time t+1 can be calculated as follows:

n
m j ,t 1 p ij mit for j=1,n (2.7)
i 1

Consider a two brand example, A and B with following switching

matrix
t+1
Where m At 0.5 and m Bt 0.5
A B

t A 0.7 0.3 Hence market share for Brand A at time t+1 is

(0.5)(0.7)+(0.5)(0.5)=0.6
B 0.5 0.5
Hence market share for Brand b at time (t+ 1) = 1-0.6=0.4

We can calculate the market share up till some infinite value, where m A 0.625 =long-run
market share.

Example 2.9: Learning Model

Learning Model to express the reinforcement effects of past brand choices, the following
applied learning model, developed by Bush and Mosteller (1955) to a consumer choice problem
by Kuehn (1962). Learning model are based on the idea that, at the individual level, Each
purchase of a given brand enhances the likelihood of future purchases of the brand.

To understand the model in its simplest form, consider a two-brand market where

1, if the brand of interest is purchased on occassion t

Yt
0, otherwise
And
pt probabilit y of purchasing the brand on occassion t [i.e., p(Yt 1)]

The basic equations of the simple linear learning model are a pair of operators called the
acceptance operator and the rejection operator:

pt 1 1 pt 1 (2.8) if brand i is purchased at t (acceptance operator)

pt 2 2 pt 1 (2.9) if brand i is NOT purchased at t (rejection operator)

39
The updating process of the model is displayed as follows:

The horizontal axis represents the probability of choosing brand j in period t and the vertical axis
represents the probability of choosing brand j in period t+1. The figure contains a positively
slope 45 0 line as a norm. The figure also contains two positively sloped lines representing the
acceptance and rejection operators from equations (2.8 and 2.9).
For example, suppose the probability that a consumer purchases brand j this period is
0.60. Suppose it is actually what he buys. The probability that the buyer will buy brand j again,
assuming that he is satisfied, is found by running a dashed line up from the horizontal axis at
0.60 to the purchase-operator line (because brand j was purchased) and going across the vertical
axis and reading the new probability. In this illustration the new probability is 0.78. If the buyer
had not purchased j, the dashed line from 0.60 would have been run up to the rejection operator
and been read on the vertical axis. In that case the probability of the person buying j next time
would have fallen from 0.60 to 0.31.
If the consumer continues to buy brand j, the probability of buying brand j approaches
0.87 as a limit. This upper limit, given by the intersection of the purchase operator and the 45 0
lines, represent a phenomenon known as incomplete habit formation. No matter how much
brand j is bought, some probability still remains that the consumer may buy another brand. On
the other hand, is the consumer does not buy brand j for a long time, the probability of buying
this brand falls continuously but never to zero. This is the phenomenon of incomplete habit
extinction. There is always some positive probability that a consumer may buy a previously
neglected brand. The incomplete habit formation aspect can been rewritten as

40
 1
p pu upper limit p
1 1

2
And p pL lower limit.
1 2

Exercise 2.2

The brand choice can be represented as a multinomial logit model in form of:

A
k/ j
e
P (2.10)
k/ j A
l/ j
e
l

Interpret the meaning of terms in the model, Pk / j , Ak / j in particular.

Exercise 2.3

Mrs Smith just moved into town, and on her first trip to the supermarket she enters the store with
0.4 probabilities of buying brand B. Suppose she buys brand B this time and also on her second
trip. Assume that her learning operators are:
pt 1 0.3 0.6 pt for a purchase
pt 1 0.1 0.6 pt for a rejection
a) What is the probability that she will buy brand B on her third trip to the store?
b) If she keeps buying brand B, what is her probability of buying brand B in the limit?

Exercise 2.4

a) Explain why customer is considered as an asset in marketing.

b) Explain the five stages in consumer behavior in making choices.
c) Observed that a customers purchase behavior can be summarized as in the following table:

Last Purchase Current Purchases

Brand A Brand B
Brand A 42.5% 12.5%
Brand B 12.5% 32.5%

i. What is the probability that the customer will buy Brand B on her third trip to the store?
ii. If she keeps buying Brand B, what is her probability of buying Brand B in the limit?

41
VALUING CUSTOMERS AND CUSTOMER LIFETIME VALUE
Recall from above, there are five stages of consumer behavior:

Need arousal through internal or external stimuli (first: personal drive say
hunger, second: an advertisement)
Information search-depending on the intensity of the stored need. It is in the
memory and it requires satisfaction or second state of active information search
where the consumer seeks information from personal, commercial or public
sources
Evaluation
Perception opinion developed based of product attributes
Preference-comparison of product
Purchase- consumers act
Post Purchase- feelings

Table 2.3: Consumer behavior models

STAGES MODELS

Need arousal Stochastic models of purchase incidence

Information search Individual awareness models

Consideration models
Information integration models

Evaluation (c) Perceptual mapping

(a) Perception (d) Attitude models
(b) Preference Noncompensatory
compensatory

Purchase Multinomial discrete choice models

Markov models

Post Purchase Variety seeking models

Satisfaction models
Communications and network models

The lifetime value of a customer generally equals the total profit a firm can expect to earn from
that customer during the time the firm continues to maintain an ongoing relationship with the
customer. These profits include both transactions between the customer and the firm and
referrals and other indirect sources of profit that can be attributed to a specific customer.

The value of a customer to a firm is the expected sum of discounted future earnings, which is
based on key assumptions concerning retention rate and profit margin. It can be determined by
the acquisition rate and cost of acquiring new customers.

42
Example 2.10

If long term value of a customer can be estimated by the lifetime value of framework, then the
growth in number of customers can be forecasted. Thus, it is easy to value the current and future
customer based company. The value of a firms customer base is the sum of the lifetime value of
its current and future customers.

First: a simple scenario where a customer generates margin mt for each period t, the
discount rate is i and retention rate is 100%. In this case, the lifetime value is simply the
present value of future income stream.
Lifetime value of a customer can be calculated as:

mt
LV (2.11)
t 0 (1 i )t

If we account for customer retention rate, r, the formulation is modified as follows:

mt r t
LV (2.12)
t 0 (1 i )t

Here, the infinite horizon is used due to:

No specific number of years that a customer going to stay with the company
The retention rate accounts that over time the chances of customer stay loyal will
go down
The estimation will be significant and simpler
Both retention and discount rate ensure that earnings from distant future
contribute significant less to lifetime value

The lifetime value with constant margin can be calculated as follows:

r r
LV m with margin multiple is equal to m (2.13)
1 i r 1 i r

Table 2.4 shows the changes in retention rate and discount rate for ABC Company. Calculate the
changes in customer value over time for the company.

Retention Discount Rate (i)

Rate (r ) 10% 12% 14% 16%
60% 1.20 1.15 1.11 1.07
70% 1.75 1.67 1.59 1.52
80% 2.67 2.50 2.35 2.22
90% 4.50 4.09 3.75 3.46
43
Exercise 2.5

Birchbox is one of the hottest subscription services out there in the market. A Birchbox
subscription is RM10 per month for women and RM20 dollars per month for men. In return for
every RM10, customers receive a box of goodies from various companies hoping to find lifelong
customers. According to Birchbox, on average each customer will become loyal for around 3
years. This, of course, is only the basic service Birchbox offers.
However, there is a 15 percent cancellation after a one year subscription. This results in the
rate of Birchbox losing its income by 15 percent per year.

a) To maintain the customer, Birchbox give a free month of service for those who complete a
subscription for a year. What will be the lifetime value (LV) per person?
b) Birchbox plans to invest in identifying people who are at-risk of canceling and nurture
them. What would be the maximum cost if the customers worth is maintained?
c) Another competing company emerges and has a 1:3 attraction towards Birchbox.
i) How much market share has been drawn from Birchbox?
ii) What would be the chance of a random male purchases Birchbox in this scenario?

44