Professional Documents
Culture Documents
373-380
313
1994. ’Ihe Economic Society of Australia. ISSN 0013-0249.
374 ECONOMIC RECORD DECEMBER
repeat purchase for these goods is such that the II The Model
present value to the firm of a reputation built on Time is divided into three periods. In period
repeat buying alone is likely to be small and rel- zero a single firm develops a new product x. The
atively ineffective in supporting signalling. More- new product is high quality with probability po
over, the set of consumers who are in the market c[O,I) and low quality with probability 1 - po.
for these goods at any particular time will gener- These probabilities are common knowledge. A
ally be different across time. This means that highquality product provides a discounted stream
much of the demand for a new product. even in of utility Y - p to a consumer with valuation v
periods subsequent to its introduction. will stem who pays price p for it. A low-quality product
from consumers who have no previous experience yields no service and so provides utility - p .
with it. These consumers will rely on communi- Quality is not observable prior to purchase and a
cation with other consumers to obtain information warranty cannot be offered.2 The product is pro-
on product qua1ity.l duced at a constant marginal cost c that is inde-
Word-of-mouth communication is in some pendent of quality. I make this assumption to
respects functionally equivalent to repeat buying in eliminate the possibility of signalling via the cost
terms of providing a mechanism for signalling. In differential mechanism outlined in the inuoduc-
both cases there is a transfer of information about tion; the focus of the paper is the reputational
quality from a consumer in one period to aconsumer channel for signalling. Moreover, it is a reitsona-
in the next period. In the case of repeat buying the ble assumption in many instances. The quality of
consumer has the same identity in both periods but a product is often largely determined by the
this is largely irrelevant from the firm's perspective. quality of research that has gone into its devel-
However, there is a key difference between word- opment. For example, the quality of a software
of-mouth communication and repeat buying that package is primarily a function of the develop-
can have important implications for introductory ment process, and this is quite independent of the
price equilibria. This difference relates to the cost marginal cost of production.
of information tnnsmission. A repeat buyer auto- There are two over-lapping generations of con-
matically acquires information about product sumers who have a potential demand for the
quality as a result of her initial purchase. In contrast, product. The first generation is young i n period
word-of-mouth communication between consum- zero and old in period one, while the second gen-
ers requires a deliberate action and that action may eration is young in period one and old in period
be costly. In a separating equilibrium in which the two. Consumers have a demand for the product
firm has signalled its quality there can be no benefit only when they are old and each consumer has
to the inexperienced consumer from seeking infor- use for at most one unit. This differentiation
mation, and so if communicationis costly they will between consumers on the basis of 'age' is
seek no information.This can undermine the mech- designed to capture the fact that different consum-
anism that supports the signalling. ers will be active in the market for a product at
The rest of this paper examines in more detail different times. Each generation comprises a con-
signalling equilibria in a market with word-of- tinuum of consumes distributed uniformly over
mouth communication, and highlights the differ- the interval [O,B] with mass B.
ence between word-of-mouth communication and Information about the quality of the new
repeat buying. Section 11 presents a simple model. product can pass between generations via word-
Section HI characterizes the signalling equilibria of-mouth communication. The probability r a[O. I )
when word-of-mouth communication is costless.
Section IV then examines the implications of costly
communication. Section V concludes. An appendix Then arr IWO reasons why full warranties can nrely
contains those proofs not included in the main text. be offered. Fint. then is a moral hazard associated with
guaranteeing a product against failure when the pruba-
bility of failure is a function of the consumer's trcatmcnt
of the product Second, it is often difficult to verify
pmdwt failure and measure its associated utility cost.
I 'Word-of-mouth communication' will be wd to In p&cular, product 'failure' is sometimes a very sub-
describe any exchange of information between con- jective matts: a consumer's dissatisfaction with a
sumcn. m g n g from casual discussion with friends product may simply be due to a mismatch between the
or neighbours to nfercnce to consumer survey product's c h t t r i s t i c s and the consumer's idiosyn-
publications. cratic tastes.
1994 WORD-OF-MOUTH COMMUNICATION 375
with which a second-generation consumer r(xl))V. If the product is low quality then none of
receives a report from a first-generation consumer the informed consumers will make a purchase and
is increasing and concave in the volume of first demand will stem only from those uninformed
period sales x , . That is: r'(xI) > 0 and r"(xI) < 0. consumers for whom v > pJp2. Hence, second-
This implies that the likelihood with which a con- period demand for a low-quality product will be
sumer receives information on a product increases
with the number of people who have bought the x: = ( 1 - r(x1)) (V - P,/PZ)
product in the past. if p2 > 0 and p 2 =sp2V
The introduction of the new product is modelled = 0 otherwise (2)
as an extensive form game. Play proceeds as If the product is high quality then all of those
follows. The product is developed in period zero. informed consumen for whom v 3 p z , together
At the beginning of period one the firm introduces with those uninformed consumers for whom v 3
the new product at a price pI. Consumers revise pJp,. will make a purchase and so demand is
their beliefs about quality on the basis of this given by
announced price. This posterior belief is denoted
pI t(0.11. First-generation consumers then make g = r(x,)(V - p J + ( I - r ( X , ) ) ( S - pJpJ
their purchase decisions to maximize expected for p2 > 0 and pz G p,V
utility. A consumer who makes a purchase imme- = r(x,)(V - p 2 ) otherwise (3)
diately observes whether the product is high or To simplify the analysis, consideration is confined
low quality. At the beginning of the second period
to the case where word-of-mouth communication
the firm announces a second-period price p z .
is not so effective as to ever make it profitable for
Second-generation consumers revise their beliefs
about quality on the basis of this price. This pos- the firm to sell exclusively to informed consum-
ers. Hence. attention is restricted to the case where
terior belief is denoted pz c[O.l]. Some of these
second-generation consumers may then receive. pr p2si.
via word-of-mouth communication with first-
generation consumers, a report on the quality of 111 Signalling with Costless Communication
the product and further revise their beliefs accord- Attention is confined to perfect Bayesian equi-
ingly. Second-period consumers then make their libria in pure strategies.' A perfect Bayesian equi-
purchase decisions and the game ends. librium is a vector of strategies and il system of
Consider the demand functions associated with beliefs such that (i) each player's strategy maxi-
this model. A consumer will make a purchase if mizes her expected payoff (starting from any node
and only if the expected utility from a purchase is in the game) given her beliefs and the equilibrium
non-negative. If the announced price in the first strategies of the other players; and (ii) beliefs are
period is pI then the expected utility to a first- computed from strategies using Bayes's rule when
generation consumer from a purchase is p,vl - applicable. More formally:
pI. The marginal first-generation buyer is there-
fore defined by v; = pIIpI and all consumers for Definition. Let ui @,, pz. pI, p2) be the discounted
whom v 3 v; make a purchase. First-period profit stream from a product of quality i. given
demand is therefore given by prices (pl.p2] and beliefs (p,. p2).That is,
xI = V -pI/pIif pI > 0 andp, d pI Ti U'@P P2r PI. P J = @I - Cbl (PI9 PI)
= 0 otherwise (1)
+ s @ 2 - crx'z @I. P 2 r PI1 P 2 )
It will never be profitable for the fiyn to set p, where k(0.1) is the firm's discount factor and the
> pI V so hereafter attention is restricted to pI 4 demand functions are as defined in (1) through
pI V. Second-generation consumers will be ran- (3). A pegect Bayesian equilibrium is a set of
domly partitioned into two subsets according to prices {fl,H,&,&] and posterior beliefs
whether or not they become informed through (p, (pl),p2 (PI, pJ) satisfying individual ration-
word-of-mouth communication. Recall that the ality conditions:
probability of becoming informed is dx,), so the
set of informed second-generation consumers will 'In signalling games such as this one, perfect Baye-
have mass r(xl)V while the set of uninformed sian equilibria coincide with the Kreps and Wilson
second-generation consumers will have mass ( 1 - (1982) notion of sequential equilibria.
376 ECONOMIC RECORD DECLVBER
Proposition 1 If an intuitive separating equilib- quality type u+,, j2, I, 1) and the equ$ibrium
rium exists then it is unique, In this equilibrium. payoff to the highquality type u”@,, p2, 1, 1)
deviate increasingly as the introductory pnce falls.
This deviation reflects the decline in second-
+ (6/4)( 1 - r(V - p,))(V - c)* < 0) period profit to the lowquality type as first-period
sales rise and more second-generation consumers
and become informed of its true quality. The full
g = z; - -
= argmax ( P ~ c ) ( ~ p~ = (V + c)/z information first-period price is illustrated as p ; .
Existence of the separating equilibrium requires
p1
that the candidate equilibrium prices in
Proof See the appendix. proposition 1 be individually rational for the high-
To see why this must be so consider first the quality type. That is.
second-period price. The choice of a separating
price in the first period informs second-generation GI - c)(V - j,)+ 6(V - c)V4 > 0. ( I 1)
consumers that the product is high quality, and in This condition is equivalent to condition (9). so if
equilibrium they receive no reports to the contrary. the candidate equilibrium exists then both incen-
The firm is then able to charge the profit-maxi- tive compatibility constraints are satisfied.
mizing full iqformation price in the second period. Whether or not ( 1 I), is likely to be satisfied
p i s price is p 2 .Next consider the first-period price depends on how low pI must be in order to be a
p , . This is the maximum price satisfying incentive separating price. This in turn depends on the effi-
compatibility condition ( 10). No sandidate equilib- cacy of the Word-of-mouth communication mech-
rium price less than or eqial to p , can satisfy the anism. If there is a high degree of information
intuitive criterion becausep, is unprofitable for the flow between generations then the introductory
low-quality type even if pI = I, 9 all p , < p , discount required to generate separation will be
yield a strictly lower gayoff than p , for the high- relatively small and separation is more likely to
quality firm when p, @,) = I. be profitable for the high-quality type. The sepa-
Note that no word-of-mouth communication rating equilibrium is therefore more likely to exist.
actually takes place in the separating equilibrium. Conversely, if word-ofcommunication is rela-
There is no need for it. Second-generation con- tively ineffective at transmitting information then
sumers become informed by the separating price separation is less likely.
and so they have no incentive to engage in com-
munication. However, the threat of communica-
tion should separation not take place is sufficient IV The Implications of Costly Communication
to ensure that the lowquality type does not mimic The foregoing analysis assumed that reports
the separating price. about product quality pass frcely between genera-
Proposition 2 In the intuitive separating equilib- tions. This section considers the implications of
rium, the firm signals high quality with a low costly communication. A second-generation con-
introductory price (relative to the full information sumer must now incur a utility cost s (independent
price). Price then rises over time. of her valuation of the product) in order to com-
municate with first-generation consumers. Com-
Proof See the appendix. munication is a binary action; it is either undertaken
This result has the following intuition. A low or it is not. It may be helpful to interpret the cost
introductory price boosts first-period sales and of communication as the cost of purchasing a con-
thereby increases the likelihood with which second- sumer report magazine. As in the previous section,
period consumers become informed through word- the probability rc[O.l) with which communication
of-mouth communication. This is to the f m ’ s is successful in locating a reliable report on the
advantage if the product is high quality and to its product’s quality is an increasing function of the
detriment if the product is low quality. This asym- volume of fust-period sales.
metry allows a low introductory price to signal high A second-generation consumer will choose to
quality. Once separation has occurred. the firm can undertake communication if and only if (i) the
charge the full information price in the second expected utility from doing so is non-negative;
period. Thus, price rises over time. and (ii) the expected utility from doing SO is at
The fmt-period equilibrium is illustrated in least as great as from making an uninformed pur-
Figure 1. The ‘mimicking’ payoff to the low- chase. Communication is successful in locating a
378 ECONOMIC RECORD DECEMBER
FIGURE1
P1
report with probability dx,) and this report will Proof The proof is by contradiction. Suppose sep
be favourable with probability p2. If the report is antion occurs in period one. Then pz c(0. 1 ) and
unfavounble then no purchase will be made. tf by (12) and (13) E u < E d Vv if s > 0. Therefore,
no report is obtained then a purchase will be made no second-generation consumers will undertake
if and only if p2v 2 pl. Hence, the expected utility communication and the asymmetry between types
from undertaking communication is required to support separation is lost. Conse-
quently, no separating equilibrium exists.’
Err = r(.c,j max 10, p2 ( v - pl)l This result arises because second-generation
+ (1 - r ( x , ) ] max (0, p2 v - p2 1 - s (12) consumers have a strict incentive not to undertake
communication once sepmtion has occurred.
while the expected utility from not undertaking Recall that when communication is costless no
communication is communication actually takes place but the threat
EUO = max [O. p2 v - p 2 ] .(13) of communication is sufficient to support sepm-
tion. That ‘threat’ is no longer credible when com-
Communication will generally not be optimal for munication is costly and so the separating
all consumers when s > 0. This will tend to equilibrium breaks down.’ This fragility of the
reduce second-period demand for the highquality separating equilibrium is the key distinction
product relative to the costless communication between an environment with word-of-mouth
case because fewer second-generation consumers communication and one with repeat buying. It
become informed about the true quality of the suggests that word-of-mouth communication
product. This effect on demand is continuous in s cannot be relied upon to support signalling in
and one might expect the equilibrium of the markets where repeat buying does not do so.
altered game to also be continuous in s. This is in
fact not so:
’Grossman and StigliU (1980) have argued that this
Proposition 3 If s > 0 then no separating equi- type of paradox in general renders informationally e f i -
librium exists. cient markets impossible.
1994 WORD-OF-MOUTH COMMUNICATION 379