Intern. J.

of Research in Marketing 31 (2014) 192206

Contents lists available at ScienceDirect

Intern. J. of Research in Marketing

journal homepage: www.elsevier.com/locate/ijresmar

Full Length Article

Choosing a digital content strategy: How much should be free?

Daniel Halbheer a,, Florian Stahl b, Oded Koenigsberg c, Donald R. Lehmann d
a
University of St. Gallen, Department of Economics, Varnbelstrasse 19, CH-9000 St. Gallen, Switzerland
b
University of Mannheim, Department of Business Administration, L5, 2, D-68131 Mannheim, Germany
c
London Business School, Regent's Park, London, NW1 4SA, United Kingdom
d
Columbia Business School, Uris Hall, 3022 Broadway, New York, NY 10027, United States

a r t i c l e i n f o a b s t r a c t

Article history: Advertising supported content sampling is ubiquitous in online markets for digital information goods. Yet, little is
First received in 27 August 2012 and was under known about the prot impact of sampling when it serves the dual purpose of disclosing content quality and gen-
review for 8 months erating advertising revenue. This paper proposes an analytical framework to study the optimal content strategy
Available online 7 November 2013
for online publishers and shows how it is determined by characteristics of both the content market and the ad-
Area Editor: Kalyan Raman
vertising market. The strategy choice is among a paid content strategy, a sampling strategy, and a free content
Guest Editor: Marnik G. Dekimpe strategy, which follow from the publisher's decisions concerning the size of the sample and the price of the
paid content. We show that a key driver of the strategy choice is how sampling affects the prior expectations
Keywords: of consumers, who learn about content quality from the inspection of the free samples. Surprisingly, we nd
Information goods that it can be optimal for the publisher to generate advertising revenue by offering free samples even when sam-
Content pricing pling reduces both prior quality expectations and content demand. In addition, we show that it can be optimal for
Sampling the publisher to refrain from revealing quality through free samples when advertising effectiveness is low and
Advertising content quality is high. To illustrate, we relate our framework to the newspaper industry, where the sampling
Dorfman-Steiner condition
strategy is known as the metered model.
2013 Elsevier B.V. All rights reserved.

1. Introduction
Digital information goods have been available on the Internet for
almost twenty years. During that time, publishers have developed dif-
ferent strategies to distribute content. Some publishers provide all
their information for free, while some charge consumers for access to
their content. Other publishers employ a hybrid business model, giving
away a portion of their content for free and charging for access to the
rest of their content. Offering free content samples allows publishers
to both disclose their content quality and to generate revenues from
advertisements shown to online visitors. According to Alisa Bowen, gen-
eral manager of The Wall Street Journal Digital Network, working with
advertisers to offer open houses has proven to be one of the most
We thank Asim Ansari, Jean-Pierre Dub, Anthony Dukes, Jacob Goldenberg, Avi
Goldfarb, Raju Hornis, Ulrich Kaiser, Anja Lambrecht, Philipp Renner, Catherine Tucker
and seminar participants at the 11th ZEW ICT Conference (2013), the GEABA 2012
(Graz), the Marketing in Israel Conference 2011 (Tel Aviv), the INFORMS Marketing
Science Conference 2011 (Houston), the University of Hamburg, the HEC Paris, the
University of Passau, the University of Tilburg, and the University of Zurich for helpful
comments and suggestions. Daniel Halbheer gratefully acknowledges support from the
Swiss National Science Foundation through grant PA00P1-129097 and thanks the
Department of Economics at the University of Virginia for its hospitality while some of
this research was being undertaken.
Corresponding author.
E-mail addresses: daniel.halbheer@unisg.ch (D. Halbheer),
orian.stahl@bwl.uni-mannheim.de (F. Stahl), okoenigsberg@london.edu
(O. Koenigsberg), drl2@columbia.edu (D.R. Lehmann).
valuable and efcient ways to expose our premium content to new
readers and potential subscribers (GlobeNewswire, 2012). The main
contribution of this paper is to provide a formal analysis of how pub-
lishers should choose between different digital content strategies.
Information goods such as digital content are experience goods;
hence consumers must have an experience to value them (Shapiro &
Varian, 1998). Offering free samples is a way for publishers to disclose
their content quality, thereby allowing consumers to have actual expe-
rience with the good before purchase. Digital information goods are par-
ticularly suitable for sampling because the costs of providing free
samples are negligible and publishers can include advertisements in
the free samples to generate advertising revenues. These two features
distinguish sampling of information goods from the typical sampling
of perishable or durable goods.
Recently, publishers introduced business models that allow con-
sumers to select samples of their choice within a set sample size. A
prominent example of this is the metered model in the newspaper in-
dustry, where publishers offer a number of articles for free and charge
for access to the rest. Such customer selected sampling differs from
the approach where the publisher chooses not only the sample size
but also the sample content, which allows the rm to strategically
manipulate the sample and creates an environment where customers
are likely to discount the sample quality in estimating actual quality. A
recent study by the Newspaper Association of America (2012) shows
that 62% of the publishers employ a metered model, out of which 95%
offer up to twenty free articles monthly. For example, the New York

0167-8116/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ijresmar.2013.10.004
 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206
Times currently offers access to ten articles for free on its website each
month. Advertising supported sampling is also employed by distribu-
tors of music such as Spotify or Rhapsody. Allowing consumers to choose
which content to sample means that publishers have no control over the
content consumers actually sample. Taking this into account is impor-
tant for publishers when setting the optimal sample size.
The business model where publishers set a sample size and let con-
sumers choose which content to sample differs from versioning or
freemium, where a rm selected low-end version is offered for free
and consumers have to pay for access to the high-end version.1 Such
versioning of information goods is often observed in the software indus-
try (see, for instance, Faugre & Tayi, 2007; Cheng & Tang, 2010). Cus-
tomer selected sampling, in contrast, does not involve quality
differentiation: Within the set sample size, the publisher allows the con-
sumers to sample any of its content for free.
This paper develops an analytical framework to study the optimal
content strategy for online publishers of information goods. Content
strategy consists of two decisions: the size of the sample and the price
of the paid content. The publisher is assumed to receive revenues
from content sales and from advertisements, which are embedded in
the free content. A key feature of sampling is that it serves the dual pur-
pose of generating revenues from advertising and disclosing content
quality. Consumers have prior expectations about content quality,
which they update in a Bayesian fashion through inspection of the
free samples. The information transmitted through the free samples af-
fects the consumers' posterior expectations about content quality,
which in turn inuence expected content sales. Taking the consumers'
quality updating into account, the publisher faces a tradeoff between
an expansion effect (through learning) and a cannibalization effect
(through free offerings) on expected content demand induced by sam-
pling. When the publisher makes its sampling and pricing decisions, it
should take both the effects on content demand and the impact on ad-
vertising revenue into account. We assume that the publisher can
choose among a sampling strategy, a pure paid content strategy, or
a pure free content strategy. Importantly, under a sampling strategy,
consumers are allowed to sample some content for free, while they
have to pay to access the remaining content (which corresponds to
the metered model in the newspaper industry).
We derive several important results. First, we show how the
publisher's advertising-sales revenue ratio and hence its optimal con-
tent strategy is determined by characteristics of both the content mar-
ket and the advertising market. The analysis conrms the insight that
the elasticities of content demand and advertising demand jointly de-
termine the publisher's optimal ratio of advertising revenue to sales rev-
enue (Dorfman & Steiner, 1954). Importantly, the elasticity of
consumers' updated expectations with respect to the sample size
plays a key role in determining the ratio of advertising revenue to
sales revenue. This elasticity is positive if sampling increases consumers'
expectations about content quality. In this case, managers can expect
the advertising-sales revenue ratio to be low due to high revenues
from content sales (resulting from the expansion effect). In contrast, if
sampling reduces consumers' expectations about content quality, man-
agers can expect the advertising-sales revenue to be high due to low
revenues from content sales (resulting from the cannibalization effect).
Second, we describe a Bayesian learning mechanism and derive ex-
pected content demand from model primitives. This allows us to pro-
vide insights into how the effects of pricing and sampling on expected
content demand are intertwined with the consumers' prior beliefs. We
nd that managers must consider two demand regimes, depending on
whether consumers' prior expectations are high or low. In both
cases, expected content demand is decreasing in price and increasing
in expected posterior quality, as expected. Importantly, we show that
sampling can have a demand-enhancing effect through consumers'
1
Bhargava and Choudhary (2008) analyze optimal versioning of information goods.
193
learning when prior expectations are sufciently low (even though
sampling produces a cannibalization effect). We establish the rule of
thumb that sampling increases content demand if the elasticity of con-
sumers' updated expectations exceeds the ratio of sampled to paid
content.
Third, to study the publisher's optimal pricing and sampling deci-
sions, we bring our model of the content market together with a
standard model of the advertising market. When content quality is
common knowledge, we show that a paid content strategy is optimal
for the publisher only if the effectiveness of advertising is sufciently
low. For intermediate levels of advertising effectiveness, the publish-
er should employ a sampling strategy and generate revenues from
both sales and advertising. Once advertising is sufciently effective,
the publisher should switch to a free content strategy. Thus, it can
be optimal for the publisher to offer content samples even if sam-
pling solely cannibalizes content demand.
In the case where consumers are uncertain about actual content
quality and learn about it through inspection of the free samples, the
optimal strategy is determined by the relationship between advertising
effectiveness and quality expectations. As in the benchmark model, two
cut-off values of advertising effectiveness determine the publisher's
optimal content strategy: a lower bound that depends on prior quality
expectations (separating paid from sampling strategies) and an upper
bound that depends on posterior quality expectations (separating sam-
pling from free content strategies). Interestingly, it can be optimal for
the publisher to generate advertising revenue by adopting a sampling
strategy even when sampling reduces both quality expectations and
content demand. In addition, we nd that it can be optimal for the pub-
lisher to adopt a paid content strategy and to refrain from revealing high
quality through free samples.
Finally, we explore three extensions of the model to generate addi-
tional managerial insights. First, when the publisher also includes ad-
vertisements in the paid content, the cut-off values depend not only
on the relationship between advertising effectiveness and quality ex-
pectations, but also on the consumers' attitudes towards advertise-
ments. Second, we provide insight on how the publisher's optimal
content strategy is determined when the willingness to pay for adver-
tisements is related to content quality. Third, we introduce competition
in the content market and analyze how asymmetries in prior beliefs af-
fect the equilibrium content strategies.
This paper is related to two literature streams. The rst stream is on
media strategy in two-sided markets.2 For instance, Kind, Nilssen, and
Srgard (2009) analyze how competition, captured by the number of
media platforms and content differentiation between platforms, affects
the composition of revenues from advertising and sales. Godes, Ofek,
and Sarvary (2009) investigate a similar question, focusing on competi-
tion between platforms in different media industries. Our paper exam-
ines optimal advertising supported content sampling and content
pricing when the rm can generate revenue from both content sales
and advertising. While papers that examine content sampling from dif-
ferent perspectives include Xiang and Soberman (2011) for preview
provision and Chellappa and Shivendu (2005) for piracy-mitigating
strategies, neither consider the impact of sampling on advertising reve-
nues. To the best of our knowledge, optimal content sampling when
sampling impacts revenues from both content sales and online advertis-
ing has not been addressed by the literature.
This paper is also related to the broad literature on consumer learn-
ing about product attributes. Firms typically enable consumer learning
through disclosing information about their products and services. This
information can be disclosed in various ways, including through infor-
mative advertising (see Anderson & Renault, 2006, and Bagwell, 2007
for a comprehensive survey), or product descriptions or third-party re-
views (Hotz & Xiao, 2013; Sun, 2011). Another way for rms to disclose
2
See Rysman (2009) for a general review of the two-sided markets literature. Anderson
and Gabszewicz (2006) provide a canonical survey of media and advertising.
194 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206

Table 1
Components of the general framework.

General Framework Explicit form

Variables Assumed properties

Publisher Content parameters Expected posterior quality

N content size Vn V n 0 Section 3.3
V maximum content quality Expected content demand
DE p; n Dp; n;
Vn Dp b 0, Dn b 0, D V N 0 Proposition 2
Cost parameters DEn N 0 demand-enhancing sampling Lemma 2
F xed production costs DnE b 0 demand-reducing sampling Lemma 2
cs unit distribution costs
Decision variables
p content price
n sample size

Content market Prior parameters Indirect utility Section 3.1

v0 minimum estimate of V u(p,n)
uncertainty about v0 Conditional indirect utility Section 6.3
Posterior parameters ui(x;pi,ni)
e
v0 n
+n
Preference parameters
valuation of quality
ad attraction/ad repulsion
x preferred product characteristic
sensitivity to mismatch

Advertising market Advertiser parameter Inverse advertising demand

advertising effectiveness a(n) ads in free content a(n) b 0 Section 4
ap(N n) ads in paid content ap(N n) b 0 Section 6.1
Endogenous ad effectiveness e n Section 6.2

information is through sampling. Heiman, McWilliams, Shen, and
Zilberman (2001) and Bawa and Shoemaker (2004) study how sam-
pling affects demand and the evolution of market shares for consumer
goods, while Boom (2009) and Wang and Zhang (2009) investigate
sampling of information goods. However, when rms sample informa-
tion goods, they only offer a portion of the good for free to avoid the
information paradox (Akerlof, 1970). Consumers' inferences about a
product's attributes are most naturally modeled in a Bayesian frame-
work. Bayesian learning processes based on product experience have
been widely employed in the literature, for instance, by Erdem and
Keane (1996), Ackerberg (2003), and Erdem, Keane, and Sun (2008),
and we follow this approach here.
We organize the remainder of the paper as follows. Section 2
presents the general framework and describes how the publisher
operates in two markets: content and advertising. Section 3 describes
the content market and studies the impact of sampling on expected
content demand. Section 4 describes the advertising market. Section 5
analyzes the publisher's optimal content strategy. Section 6 offers ex-
tensions of our analysis. Conclusions and directions for future research
are provided in Section 7. To facilitate exposition, we have relegated
proofs to the Appendix.
2. General framework
We rst introduce the three main components of our modeling
framework: the publisher, the content market, and the advertising
market. Next, we dene the strategies available to the publisher and
characterize the optimal advertising-sales revenue ratio and thus the op-
timal content strategy. Table 1 summarizes the components of the gen-
eral framework (as well as its main assumptions) and indicates where
the reduced-form expressions are replaced by specic functional forms.
2.1. Publisher
We consider a publisher who offers a digital information good with
content of size N N 0 through an online channel. Content size may be
thought of as the number of chapters of a book or movie, the number
of songs on an album, or the number of articles on a news platform.
To mirror the cost properties of information goods, we assume that
the publisher has xed costs F 0 and zero unit costs to produce the
content.3 The cost to provide digital access per subscriber is cs 0 and
the costs of providing free samples are normalized to zero. The qualities


of the content parts are distributed on the quality spectrum 0; V ,
where V is the publisher's private information. We treat quality V as
an outcome of a previous strategic decision and suppose that the pub-
lisher has two decision variables: the sample size n [0, N] and the
price p at which to sell the good.4 Notice that in the context of the
metered model, the sample size n is the meter and p is the price for
the content behind the paywall, which separates free content from
paid content.
2.2. Content market
We consider a market with a unit measure of consumers that
observe the publisher's sampling and pricing decisions. Consumers are
uncertain about content quality. We assume that they update their
prior expectations in a Bayesian fashion through inspection of the free
samples and denote by V e n the consumers' expected posterior quality
given the sample size n. Expected content demand depends on price p,
e n. Specically, we assume that the publisher's ex-
sample size n, and V
pected content demand is given by
E e n:
D p; n Dp; n; V 1
This representation emphasizes that the sample size has both a direct
effect on content demand and an indirect effect that operates through
e n.
the impact of n on expected posterior quality V
3
Throughout the analysis, we assume that the xed cost do not exceed the product
market prot. Hence they do not change the analysis and can therefore be omitted.
4
The choice of (p, n) is not a multidimensional signal for quality as studied, for instance,
by Wilson (1985) and Milgrom and Roberts (1986). In this strand of the literature, n is an
advertising signal for quality. However, in our setting, the publisher's choice of n allows
the consumers to gain information about the actual content quality through their sample
experience before making the purchase decision.
 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206
Expected content demand satises the following basic assumptions.
First, we assume that D
p
b 0, i.e. content demand depends negatively on
price. Second, we impose that D n
b 0, so that a larger sample size has a
direct negative effect on demand for the remaining content accessible
through the paywall. Third, we require that V D
N 0, i.e. content demand
depends positively on expected posterior quality. The overall effect of
the sample size n on expected content demand is given by
E 1
D D D e
V n;
n e
n V
D e
where term V V n captures the indirect effect of the sample size on ex-
pected content demand. It is not clear a priori how the sample size af-
E
fects posterior expectations and hence D n
. If Ve n b 0 , sampling
reduces posterior expectations and is thus demand-reducing. Even if
e n N 0 , that is, if sampling increases posterior expectations, offering
V demand.
an additional sample may be demand-reducing if the direct effect dom-
e n is sufciently large, the in-
inates the indirect effect. However, once V
direct effect is stronger than the direct effect so that DE
N 0 and
n content strategy are determined by characteristics of both the content
sampling has a demand-enhancing effect. In line with Bawa and
Shoemaker (2004), we refer to the direct effect of sampling on content
demand as the cannibalization effect and to the indirect effect as the
expansion effect.
2.3. Advertising market
We consider a market where the advertisements are delivered to the
publisher through a representative advertiser (e.g., an advertising agen-
cy). We assume that the publisher includes one advertisement in each
free content part. The inverse advertising demand is denoted by a(n)
and maps the publisher's choice of n into the market price for ads.
Thus, a(n) can be thought of as the advertiser's willingness to pay for
placing n advertisements. We make the natural assumption that the
price for advertisements decreases in sample size, that is, a(n) b 0.
2.4. Optimal strategy
The publisher receives prots from the content market and from
the advertising market. The expected prot from content sales is pcs 
e n, while the prot (revenue) from including advertisements in
Dp; n; V
the free articles is a(n)n. The publisher makes pricing and sampling deci-
sions so as to maximize its (expected) prot from the two markets:
max e n ann
p; n pcs Dp; n; V
p;n
s:t: p 0
0 n N:
As long as the publisher's prot function (p, n) is concave, standard opti-
mization theory posits that there is a unique constraint global maximizer
(p*, n*) because the constraint set is convex. Depending on the optimal
pricing and sampling decision, the following denition gives the strategies
available to the publisher.
Denition 1. Strategies
Given the optimal pricing and sampling decision (p*, n*), the publisher
adopts either (i) a sampling strategy if p* N 0 and n* (0, N), (ii) a pure
paid content strategy if p* N 0 and n* = 0, or (iii) a pure free content
strategy if p* = 0 and n* = N.
Notice that both the paid content strategy and the free content
strategy are nested within the sampling strategy: The publisher re-
ceives no advertising revenue under a paid content strategy and no
sales revenue under a free content strategy. The following result
195
describes the optimal strategy as the ratio of advertising revenue to
sales revenue.
Proposition 1. Advertising-sales revenue ratio
Under a sampling strategy, the publisher's optimal ratio of advertising
revenue to sales revenue is given by
an n V V
  ; 3
Dp 1
a p
where p (D/p)(p/D) denotes the elasticity of content demand with
respect to price, n (D/n)(n/D) denotes the elasticity of content de-
e V=D
e
mand with respect to sample size, V D=V denotes the elastic-
e nn=V
ity of content demand with respect to quality, V V e denotes
the elasticity of posterior quality expectations with respect to sample
size, and a n(a)(a/n) denotes the price elasticity of advertising
This result has two important managerial insights: First, it shows that
the publisher's advertising-sales revenue ratio and hence its optimal
market and the advertising market. Specically, consumer preferences
determine the characteristics of the content market, captured by the
elasticities of content demand with respect to price, sample size, and
quality. The price elasticity of advertising demand reects advertiser
preferences. This general result thus provides guidance for managers
seeking to better understand the contributions of (expected) sales and
advertising to total revenue.
Second, Proposition 1 shows how changes in the market envi-
ronment, captured by the various elasticities, affect the publisher's
composition of revenues. Unsurprisingly, if the price elasticity p in-
creases, the advertising-sales revenue ratio is lower. Intuitively, for a
given sample size, the optimal price for the content is lower, which
results in a higher sales revenue. In contrast, higher elasticity of con-
tent demand with respect to the sample size n increases the
advertising-sales revenue ratio. Furthermore, the higher the price
elasticity of advertising demand a, the lower is the advertising-
sales revenue ratio.
Proposition 1 also highlights the crucial role which the elasticity of
posterior quality expectations with respect to sample size plays. Be-
cause the elasticity of content with respect to quality V is positive,
the impact of sampling on posterior quality determines the sign of
V V . If V is negative, the ratio of advertising revenue to sales revenue
tends to be high, while it tends to be low if V is positive. Intuitively,
if V b 0, sampling reduces expected content demand as V n b 0, and
hence the advertising-sales revenue ratio is high. In contrast, if V N0,
2 sampling increases expected content demand as consumers revise their
expectations about quality upwards, resulting in a lower advertising-
sales revenue ratio.
Interestingly, the optimal advertising-sales revenue ratio is reminis-
cent of the well-known Dorfman-Steiner condition, which states that a
monopolist's ratio of advertising spending to sales revenue is equal to
the ratio of the elasticities of demand with respect to advertising and
price (Dorfman & Steiner, 1954). Proposition 1 reduces to this result in
the special case when offering additional samples does not affect poste-
rior quality V 0 and if the advertising demand is perfectly elastic
(a ).
Our general framework is agnostic about how consumers form pos-
terior quality expectations. To shed light on effects of sampling on pos-
terior quality expectations, the next section introduces a Bayesian
learning mechanism in which consumers update their prior expecta-
tions about content quality through experience with the sample. Impor-
tantly, we show how the publisher should take the consumers' learning
into account to gauge the effects of offering free samples on content
demand.
196 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206
(A)
Fig. 1. Prior expectations about V (where V 1).
3. Content market
This section analyzes the impact of customer selected sampling on
expected content demand when the qualities of the content parts are


uniformly distributed on the quality spectrum 0; V . We begin by laying
out assumptions regarding consumer behavior and describing the
Bayesian learning mechanism. Taking consumers' learning into account,
we derive the publisher's expected content demand given its pricing
and sampling decisions. Finally, we provide conditions under which
sampling increases or decreases expected content demand.
3.1. Consumer behavior
Consumers know that the

qualities
 of the free samples are uniformly
distributed on the interval 0; V , but they do not know the upper bound
of the publisher's quality spectrum V and are hence uncertain about (av-
erage) content quality.5 Consumers have a common prior belief about V
that may stem, for instance, from reviews, ratings or word of mouth.
The natural conjugate family for a random sample from a uniform distri-
bution with unknown upper bound is the Pareto distribution (DeGroot,
1970). We capture uncertainty about V by a prior belief that consists of a
minimum estimate v0 of the upper bound V and a level of uncertainty
about this value. Specically, we assume that prior beliefs follow a Pare-
to distribution with density function
8 NEVjv1 ; ; vn  np;
< v0
; for v N v0
f vjv0 ; 1
:v
0; otherwise:
We assume N 1 to ensure existence of prior expectations.6 Based
on the consumers' prior knowledge about v0 and , their prior expecta-
tion about V is

 v0
E Vjv0 ; : 4
1 will purchase the information good if and only if the indirect utility from
Obviously, prior expectations increase in v0 and decrease in . Fig. 1
illustrates two prior beliefs along with the corresponding expectations
for different parameter values. Prior expectations are lower than actual
quality in Panel A and higher than actual quality in Panel B. Note that
prior expectations can be higher than actual quality even if v0 b V.
Consumers update their prior belief about the upper bound of the
quality spectrum V by taking the observed qualities of the free samples
into account. Specically, consumers use the n sample qualities Vi = vi
5 8
Note that the upper bound V is monotonically related to the mean, which may be an
alternative way for consumers to think about content quality.
6
Our measure of uncertainty corresponds to the scale parameter of the Pareto distri-
bution. Hence, when uncertainty is higher, the prior distribution is more spread out.
(B)
(i = 1, , n) to form their posterior belief evn about V. Using standard
Bayesian analysis, evn follows a Pareto distribution with minimum
value parameter e v0 n maxfv0 ; v1 ; ; vn g and shape parameter
+ n (DeGroot, 1970).7 Hence, the posterior expectation of V is
given by

 ne
v0 n
E V je
v0 n; :
n1
Consumers infer the expected quality of the information good
EVjv1 ; ; vn  from the average quality of the sampled content parts.
Knowing that qualities are uniformly distributed on the quality spec-
trum offered, the expected quality of the information good is given by


E V je
v0 n;
EVjv1 ; ; vn  : 5
2
Consumers believe that higher quality is better than lower quality
but differ in the way they value quality. To capture this heterogene-
ity, we introduce a preference parameter for quality , which is uni-
formly distributed on the interval [0,1]. We assume that each
consumer either purchases the information good at price p or stays
with the n free samples. A consumer's indirect utility from these
two options is given by

from purchasing at price p
up; n
nEVjv1 ; ; vn n; from staying with the f ree samples:
The parameter captures the intensity of ad-attraction ( N 0) or ad-
repulsion ( b 0) in the population (Gabszewicz, Laussel, & Sonnac,
2005). Thus, when consumers exhibit ad-loving behavior, the utility
of both options is augmented by n, while the utility of both options
is reduced by n in the case of ad-avoiding behavior.
The value of the information good is equal to the number of content
parts multiplied by their expected quality.8 This implies that a consumer
buying exceeds the indirect utility from consuming only the free sam-
ples, that is, if
NnEVjv1 ; ; vn p 0: 6
This condition means that the (quality-weighted) expected value of the
content that has not been sampled must exceed the price. Importantly,
purchase does not depend on consumer behavior towards advertising.
7
The proof of this result appears in the Appendix.
This additivity assumption is justied for independently valued content parts. However,
a concave or convex relationship between the value and the number of content parts might
be more appropriate for interrelated content parts, that is, if the content parts are substitutes
or complements.
 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206 197

3.2. Expected content demand v0 b V, sampling may have a demand-enhancing effect. We next address
this possibility.
Because consumers do not know the upper bound of the quality
spectrum V with certainty, content demand is inuenced by con- 3.3. The role of quality expectations
sumers' posterior quality expectations. Thus, when the publisher
makes decisions about the sample size and the price, it has to base For a given level of prior expectations about content quality, sam-
them on expected content demand as consumers have not yet evalu- pling either increases or decreases expected content demand. Whether
ated sample qualities and updated their expectations about content sampling compensates for cannibalization through consumers' learning
quality. depends on the gap between posterior quality and actual quality.
In order to compute expected content demand, we assume that the Expected posterior quality is
publisher knows the consumers' prior parameters v0 and , which can
be learned using standard market research techniques such as surveys. 8
>
> nvn1
n1
nV
>
< 0
The calculation involves a three-step procedure: In the rst step, the n ; if v0 b V
e n 2 n1n 1V
V 10
publisher computes the expected posterior quality by averaging poste- >
> nv0
>
rior expectations about V as given by (5) across all possible realizations : ; if v0 V
2 n1
of sample qualities:

nE e
v0 n and the quality gap is dened as Ve n V . Consumers overestimate (un-
2
EEVjV 1 ; ; V n  :
2 n1 derestimate) quality if the expected posterior quality is higher (lower)
than actual quality. This leads to the following result.
In the second step, the publisher substitutes the expected posterior
quality into the purchase condition given by (6) to obtain expected con- Lemma 1. Quality gap
When v0 b V , consumers overestimate quality after their sample
tent demand:
experience if
 
E p 2n1
D p; n max 0; 1 : 7  1
Nn nEe
v0 n v0 1 n1
N ; 11
V n
In the third step, the publisher calculates E e
v0 n to obtain the expected
content demand as a function of the underlying model parameters. This and underestimate it if the inequality is reversed. If v0 V , consumers
calculation leads to the following result. overestimate quality irrespective of the sample size and their level of uncer-
Proposition 2. Expected content demand tainty N 1.
When the publisher sells the information good at price p and offers
n {1, , N 1} samples, then Prior expectations can be higher than actual quality even if v0 < V (a
high level of uncertainty about v0 is captured by a low ). The condition
(a) if v0 b V, expected content demand is given by
in Eq. (11) applies when consumers overestimate quality based on pos-
terior expectations: This is likely to be the case for a low and when the
( ) publisher offers a small number of free articles n. On the other hand,
n
E p 2 n1n 1V consumers underestimate quality if their uncertainty is low and the
Dfv0 b V g p; n max 0; 1 : 8 sample size is large.
Nn nvn1 nV n1
0
When v0 b V, the shaded area in Fig. 2 illustrates the set of param-
eters for which consumers overestimate and underestimate quality,
respectively. By construction, where N 1, condition (11) holds and
(b) if v0 V, expected content demand is given by consumers overestimate quality. The parameter region for which con-
sumers overestimate quality shrinks as n gets larger since V e n V as
2
n , meaning that consumers learn actual quality once the sample
  size gets large enough.
E p 2 n1
Dfv0 Vg p; n max 0; 1 : 9
Nn nv0

Expected content demand has the intuitive properties that we as-

sumed in our general framework: it decreases in price and increases
in expected posterior quality. In addition, sampling has both a direct
demand-reducing effect and an indirect effect that operates through
its impact on posterior expectations. The direct effect kicks in
through the factor Nn1
and mirrors the cannibalization effect D
n
b 0.
This follows because a larger sample size reduces the utility of the
remaining content consumers have to pay for. Proposition 2 nests the
expected content demand under a paid content strategy (n = 0), in
which case D(p, 0) is a function of prior quality expectations as there
is no learning. In contrast, when the publisher employs a free content
strategy (n = N), consumers do not purchase the information good
as they can download it for free and hence D(p, N) 0.
Which of the two demand functions reported in Proposition 2
emerges depends on the value of the minimum estimate v0 about V
vis--vis the actual upper bound of the quality spectrum V : If v0 V , Fig. 2. The quality gap for the case v0 b V (where V 10). The shaded area indicates where
sampling necessarily reduces expected content demand. In contrast, if consumers underestimate quality.
198 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206

e n allows us to rewrite the expected content de-

The denition of V N 1.9 For expositional purposes, we normalize the costs of providing
mand derived in Lemma 2 more compactly as digital access cs to zero and present analysis ignoring the integer con-
( ) straint on the number of free samples.
E p
D p; n max 0; 1 ; 12
e n
NnV 5.1. Strategy with known quality

which is a specic version of the reduced-form demand function in We rst derive content demand under a sampling strategy and sub-
Eq. (1). Hence the number of free samples n has both a direct effect on sequently derive the demands for the two boundary strategies when
expected content demand and an indirect effect that operates through consumers know content quality. Next, we characterize the optimal
posterior quality expectations V e n. The next result uses this demand pricing and sampling decisions for each of the three strategies. Finally,
function to identify conditions under which sampling has a demand- we determine the optimal content strategy.
E
enhancing effect (that is, D
n
N 0).
5.1.1. Content demand
Lemma 2. Effects of sampling When consumers know the upper bound V and hence the quality
Offering free samples has a demand-enhancing effect if V N Nn
n
, that is, spectrum, the minimum estimate v0 of the upper bound V is equal to V
if the elasticity of consumers' posterior quality expectations exceeds the with certainty and thus . Proposition 2 implies that content de-
ratio of sampled to paid content. mand can be expressed as
8 9
>
> >
>
Lemma 2 shows that offering free samples may increase expected < p =
content demand through consumers' learning, even though it produces Dp; n max 0; 1 14
>
> V> >
a cannibalization effect. Intuitively, the indirect effect dominates the : Nn ;
2
direct cannibalization effect if sampling results in a sufciently large up-
wards revision of consumers' prior expectations. for n {0, , N 1}. Recall that D(p, N) 0 under a free content
strategy.
4. Advertising market
5.1.2. Optimal pricing and sampling
This section derives inverse advertising demand of a representative In the benchmark case with known quality, the publisher makes its
advertiser that places advertisements in the free samples offered by pricing and sampling decisions so as to
  
the publisher. The representative advertiser can be thought of as an ad- p n
max p; n p 1 n
vertising agency that delivers independent product advertisements to p;n Nn 2
V N
the publisher. Following Godes, Ofek, and Sarvary (2009), we let the s:t: p 0
advertiser's utility from placing n advertisements be given by 0 n N;

2 where content demand is given by Eq. (14) and inverse advertising de-
n
un n an;
2N
where N 0 is a parameter capturing the marginal benet of an adver-
tisement and a denotes the unit price set by the publisher to run an ad-
vertisement. This specication captures decreasing marginal utility
from placing advertisements caused by decreasing consumer recall
and retention rates for ads (Burke & Srull, 1988).
The advertiser maximizes its utility to determine how many adver-
tisements to run on the publisher's platform. The implied inverse adver-
tising demand has the linear form
n
an ; 13
N
under a sampling strategy, p NV 82 V =64
where is referred to as advertising effectiveness. Inverse demand
slopes downward, implying that the publisher's revenue per ad impres-
sion is decreasing in sample size. Intuitively, a(n) can be thought of as
the advertiser's willingness to pay for placing n advertisements.
5. Optimal content strategy
This section analyzes the publisher's optimal content strategy with
customer selected sampling. We rst analyze the benchmark case
in which the consumers know V and hence the publisher's quality
spectrum. In this case, sampling does not affect the consumers' ex-
pectations about quality and simply serves to generating advertising
revenues. Next, we analyze the case where V is not known to con-
sumers. In contrast to the benchmark case, sampling not only
generates advertising revenues but also inuences consumers' ex-
pectations about quality.
We consider three strategies: the paid content strategy, the sam-
pling strategy, and the free content strategy. In order to study the
tradeoffs between the three strategies, we focus on the case where
mand by Eq. (13). From the rst-order conditions, the optimal price for
a given sample size is
NnV
pn : 15
4
This implies that the more free samples the publisher chooses to offer,
the less it will be able to charge the consumer for the remaining content.
The next result summarizes the optimal pricing and sampling decisions
for each of the three strategies.
Lemma 3. Pricing and sampling
Suppose that consumers know the upper bound of content quality
V . Then, (i)
and n N 8V =16, (ii) under a paid content strategy, p NV=4
and n = 0, and (iii) under free content strategy, p = 0 and n = N.
The parameters V and have opposite effects on the optimal price
and on the optimal sample size under a sampling strategy: As expected,
p increases in V while n decreases in quality. In contrast, p decreases in
, and n increases in advertising effectiveness. Both the optimal price
and the optimal sample size increase in content size N. The following
proposition characterizes the publisher's optimal strategy.
Proposition 3. Optimal strategy
If consumers know the quality spectrum, then: (i) if V8, the publisher

should follow a paid content strategy, (ii) if V8 ; V8 2 , the publisher
should employ a sampling strategy, and (iii) if V
8 2, the publisher's
optimal strategy is a free content strategy.
9
When 1, the free content strategy is never optimal because the advertising reve-
nues and hence the publisher's prot is zero. Thus, the publisher's choice is only between
the paid content strategy and the sampling strategy. The analysis where N 1 encom-
passes this special case.
 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206 199

Proposition 3 shows that the choice of the optimal strategy is driven In contrast to our benchmark case, it is not possible to character-
by the relationship between content quality V and advertising effective- ize the publisher's optimal pricing and sampling decisions (and
ness . Thus, for a given content quality, a paid content strategy is hence prots) analytically. Nevertheless, we have the following
optimal if the effectiveness of advertising is sufciently low. For inter- result.
mediate levels of advertising effectiveness, a sampling strategy that
generates revenues from both sales and advertising on the free samples Proposition 4. Optimal strategy
is optimal. If advertising is sufciently effective, the publisher should Suppose that consumers are uncertain about the upper bound of
switch to a free content strategy. content quality V and that the prot function E(n) is strictly concave.
The effects of and V on the optimal strategy can also be understood e 0NV
Then, there are cut-off values 1 V e 0 and 2 V N
4 4
by inspection of the advertising-sales revenue ratio. The ratio follows such that a paid content strategy is optimal for , a sampling
from (3) and is strategy is optimal for ; , and a free content strategy is optimal
! ! for .
V V This result is consistent with the insights from the benchmark case

an 8 8 when quality is known: a paid content strategy is optimal only if the ad-
! :
Dp V V vertising effectiveness is sufciently low, a sampling strategy is optimal
2 for intermediate levels of the advertising effectiveness, and the publish-
4 8
er should switch to a free content strategy once advertising is sufcient-
The ratio of advertising revenue to sales revenue tends to zero as ly effective. Fig. 3 illustrates the optimal strategy for varying advertising
V effectiveness and the expected prot for each strategy (SC for the
approaches the lower bound , implying that the publisher should em-
8 sampling strategy, PC for the paid content strategy, and FC for the
ploy a paid content strategy. A sampling strategy is optimal only if ad-
free content strategy).
vertising is not too effective, that is, as long as V8 2 . Once
Hence prior expectations determine the lower of the two cut-off
exceeds this level, the publisher should switch to a free content
values for a sampling strategy to be optimal whereas posterior expecta-
strategy.
tions for sample size n = N determine the upper cut-off value. In effect,
is determined by the impact of the rst free content part on poste-
5.1.3. Summary
rior expectations, while is determined by posterior expectations
When content quality is known to consumers, the publisher's opti-
after inspection of the last free content part. The next lemma shows
mal strategy is determined by the relation between advertising effec-
that the model where quality V is not known to consumers nests the
tiveness and content quality. The more effective advertising is, the
full information benchmark case (see Proposition 3).
more free samples the publisher should offereven if it solely cannibal-
izes content demand. Lemma 4. Cut-off values
Suppose that consumers are uncertain about content quality V and that
5.2. Strategy with unknown quality the prot function E(n) is strictly concave. Then, when consumers have
correct quality expectations, that is, if v0 V and , the lower
We rst determine the optimal content strategy when consumers bound converges to V and the upper bound converges to V8 2.
8
do not know content quality and compare our ndings to the results
from the benchmark case. Next, we study how the interplay of prior ex-
pectations and advertising effectiveness governs the optimal choice of 5.2.2. The impact of prior expectations
content strategy. Proposition 4 shows that the optimal strategy depends not only on
advertising effectiveness and quality V as in the benchmark case,
5.2.1. Optimal pricing and sampling but also on the specic values of the prior parameters v0 and (as
When content quality is not known to consumers, the publisher well as content size N). Fig. 4 illustrates the optimal strategy for
makes its pricing and sampling decisions so as to given advertising effectiveness and prior expectations. Panel A
! depicts the cut-off thresholds between the different strategies in
 n
E p the ; v0 -space (given = 2). Similarly, Panel B illustrates the
max p; n p 1 n
p;n e n
NnV N publisher's optimal strategy in the (, )-space (given v0 5 ).
s:t: p0
0 n N;

where expected content demand is given by Eq. (12) and inverse adver-
tising demand by Eq. (13). The only difference between this expected
prot and the prot when content quality is known to consumers is
the dependence on expected quality rather than actual (average) qual-
e n is the posterior estimate of av-
ity. Based on (15) and recalling that V
V
erage quality 2 , we thus obtain that

e n
NnV
pn :
2

Substituting p(n) back into the prot function allows us to rewrite the
prot maximization problem as

E Ve n  n
max n Nn n 16
n 4 N
s:t: 0 n N: Fig. 3. Optimal strategy with unknown quality (for v0 5, = 2, V 10, and N = 10).
200 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206

(A) (B)
Fig. 4. Optimal strategy (for V 10 and N = 10).

Here prior expectations are correct and coincide with actual quality effectiveness to depend on content quality. Third, we introduce compe-
when v0 5 and = 2. The following observation summarizes our tition into the model.
insights.
6.1. Including advertisements in the paid content
Observation 1. Prior expectations
If advertising effectiveness is high and actual content quality is relatively In this section, we extend the model by allowing it to include adver-
low, it can be optimal for the publisher to adopt a sampling strategy to gen- tisements in both the free articles and the paid content. To this end, we
erate advertising revenue even when it reduces prior quality expectations assume that the market price for advertisements included in the paid
and content demand. In contrast, if advertising effectiveness is low and ac- content is given by
tual content quality is relatively high, it can be optimal to adopt a paid con-
tent strategy and not reveal high quality, even though sampling would ^
n
increase content demand. ^ p
ap n ; 17
N

The shaded areas in Fig. 4 illustrate these important managerial in- where n ^ Nn and p N 1 denotes the advertising effectiveness for
sights. Notice that the larger areas correspond to the case where prior ads in the paid content. This inverse demand is a natural counterpart
expectations are high relative to actual content quality. In such market to the advertising demand a(n) given by Eq. (13) and indicates
environments, the publisher should sacrice content demand to boost that the ad price depends on the number of articles n ^ that have
advertising revenues. The gure also shows that when prior expecta- not been offered as free samples. Differences in the levels of
tions are sufciently low relative to actual content quality (that is, if v0 advertising effectiveness p and capture differences in reach or
is small or is large), the choice is between a sampling strategy and a the degree of targeting in the advertising markets for paid and free
free content strategy only. Intuitively, the publisher has an incentive to content.
reveal its higher than expected quality through free samples, possibly When allowing for advertisements in the paid content, a consumer's
offering its content for free. In contrast, when prior expectations are suf- indirect utility from the two options is
ciently high, the optimal strategy is either a paid content strategy or a

free content strategy. In such market environment, the publisher has no NEVjv1 ; ; vn  Np; from purchasing at price p
^ p; n
u
incentive to reveal its lower than expected quality when is low. nEVjv1 ; ; vn  n; from staying with the free samples:

Importantly, the utility of the purchase option now depends on the

5.2.3. Summary
overall number of advertisements shown to the consumer rather than
When actual content quality is not known to consumers, the optimal
the number of ads contained in the free samples only. This augmented
strategy is determined by the relation between advertising effectiveness,
specication implies that a consumer will purchase the information
prior quality expectations, and posterior quality expectations. As in the
good if and only if
benchmark case, employing a paid content strategy is optimal only if
advertising effectiveness is sufciently low compared to prior quality
NnEVjv1 ; ; vn  pNn; 18
expectations. For intermediate levels of advertising effectiveness, the
publisher should use a sampling strategy. The publisher should switch
that is, if the expected value of the content that has not been sampled
to a free content strategy once advertising is sufciently effective com-
exceeds its full price p(N n). Intuitively, the full price of the con-
pared to posterior quality expectations. Counter to intuition, it can be
tent is lower if consumers are ad-lovers ( N 0) and higher if they are
optimal for the publisher to generate advertising revenue by adopting
ad-avoiders (in which case exposure to the ads in the paid content re-
a sampling strategy even when sampling reduces both prior quality ex-
sults in a nuisance cost (N n) that has to be added to the price p).
pectations and content demand. In addition, it can be optimal for the
From the purchase condition in Eq. (18), expected content demand
publisher to adopt a paid content strategy and to refrain from revealing
can be derived as
high quality through free samples.
  p 
^ E p; n max 0; 1 1
D : 19
6. Extensions EVjv1 ; ; vn  Nn

This section relaxes three key assumptions and studies the effects on Compared to the case where consumers are ad-neutral, content
our ndings. First, we allow the publisher to generate advertising reve- demand is higher when consumers are ad-lovers and lower when con-
nues from paid content as well. Second, we allow advertising sumers are ad-avoiders.
 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206
The publisher makes its pricing and sampling decisions so as to
 
E
max p; n p Rp n ^ D ^ E p; n n n
p;n N which in turn increases the advertiser's willingness to pay for the adver-
s:t: p 0
0 n N;
where expected content demand is given by Eq. (19) and inverse adver-
^ ap n
tising demand by Eq. (17). Notice that Rp n ^ n
^ are the additional
revenues from including ads in the paid content. By denition, Rp = 0
under a free content strategy, while Rp = (p 1)N under a paid con-
tent strategy. We summarize our insights as follows.
Observation 2. Ads included in paid content
When the publisher includes advertisements in the paid content and
consumers are neutral about advertisements or ad-lovers, the range of ad-
vertising effectiveness for which the sampling strategy is best expands. In
contrast, when consumers are ad-avoiders, the range of for which the
sampling strategy is optimal can shrink.
Fig. 5 illustrates the prot effects of including advertisements in the
paid content when consumers are neutral about advertisements. Intui-
tively, exploiting revenues from advertisements in the paid content in-
creases the unit margin from selling content, which translates into a
higher prot under a sampling strategy. These prot effects are more
pronounced when advertising effectiveness for ads in paid content p
increases. Further, the prot under a sampling strategy is higher when
the consumers are ad-lovers (for given p) and lower when they are
ad-avoiders.
6.2. Endogenizing advertising effectiveness
Up to now, we considered advertising effectiveness to be exogenous
and independent of content quality. In this section, we relax this as-
sumption and allow the advertiser's willingness to pay for advertise-
ments to be positively related to expected posterior quality, in effect
having product quality have a halo effect on ad effectiveness. Specical-
ly, we let inverse advertising demand be given by
n by the dashed lines reproduced from Panel A in Fig. 4). This occurs be-
e n
an
N cause when prior expectations are low, sampling not only reveals high
and capture the effect of expected posterior quality V e n on quality-
e n in the following way:
adjusted advertising effectiveness
e 6.3. The impact of competition
e n V n ;
e 0
V
e 0
where is advertising effectiveness (as introduced in Section 4) and V
denotes prior quality expectations. This specication reects the idea
Fig. 5. Optimal strategy with (dashed lines) and without advertisements in paid content
(solid lines) for = 0, v0 5, = 2, V 10, N = 10, and p = 1.1.
201
that free samples become more attractive as an advertising platform
when consumers' posterior expectations exceed their prior expectations,
tisements (and vice versa).
When advertising effectiveness is positively related to posterior con-
tent quality, offering free samples affects the willingness to pay for ad-
vertisements in two ways: Through changes in the quality-adjusted
advertising effectiveness e n and through changes in the number of
ads shown to consumers. Therefore, advertising demand can be increas-
ing in the number of free samples when prior expectations are suf-
ciently low relative to consumers' updated expectations (see Panel A
in Fig. 6 for a graphical illustration). Over the range of where a(n) in-
creases, the marginal benet of higher advertising effectiveness out-
weighs the disutility of showing an additional ad to consumers.
The publisher makes its pricing and sampling decisions so as to
!
p 
max
E
p; n p 1 e n n n
p;n e n
NnV N
s:t: p 0
0 n N;
where expected content demand is given by Eq. (12) and inverse adver-
tising demand by Eq. (20). The next observation summarizes our
insights.
Observation 3. Endogenous ad effectiveness
When consumers' prior expectations are low and the advertiser's will-
ingness to pay for advertisements is positively related to the expected pos-
terior content quality, it can be optimal for the publisher to adopt a free
content strategy even when baseline advertising effectiveness is low.
Panel B of Fig. 6 illustrates this observation. In the gure, the lines
depict the cut-off thresholds between the different strategies in
the (, v0)-space. The solid lines correspond to the case with quality-
adjusted advertising effectiveness and show that a free content strategy
can also be optimal for low , whereas a sampling strategy yields a
higher prot when advertising effectiveness is exogenous (indicated
20
quality, but also increases the advertiser's willingness to pay for ads
and thus to boost advertising revenues.
Thus far, we have examined a publisher operating in a monopoly
setting. Here, we allow for competition between two publishers that
offer differentiated information goods. We add horizontal product
differentiation to capture the intensity of competition between the
two publishers.10 In the newspaper industry for instance, horizontal
product differentiation may arise due to different political opinions
among consumers (Gabszewicz, Laussel, & Sonnac, 2005).
The analysis of the competitive case indicates that, as in the monop-
oly case, advertising effectiveness is a key driver of the publishers' strat-
egy choice. Specically, if consumers' prior expectations about content
qualities are similar and advertising effectiveness is high, it is optimal
for both publishers to adopt a free content strategy in equilibrium. How-
ever, two other drivers that affect strategy choice in the competitive
case: the gaps in prior expectations about the quality of the competing
products and the degree of horizontal product differentiation. The
following observation summarizes our insights.
Observation 4. Competition
If the gap in prior expectations regarding the product qualities is large
and advertising effectiveness is high, both publishers should adopt a sam-
pling strategy. As the degree of horizontal product differentiation increases
10
All details of the model and the analysis of the competitive case are in the Appendix.
202 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206
(A)
Fig. 6. Impact of endogenizing advertising effectiveness (for = 2, V 10, N = 10; in addition, in Panel A v0 2 and = 2).
and the products become less substitutable, the parameter region in which
the sampling strategy is optimal for both publishers shrinks.
Counter to intuition, our results show that it is not per se optimal to
choose a free content strategy when advertising effectiveness is high.
What matters in addition is the gap in prior expectations between the
two publishers: if it is small, then both rms should adopt a free content
strategy. Instead, if the gap in prior expectations is large, then both rms
should adopt a sampling strategy. Intuitively, the publisher that faces
lower prior expectations from consumers has a stronger incentive to re-
veal its higher than expected qualityand it is a best-response of the
rival publisher to also adopt a sampling strategy.
The second insight is that the parameter region in which the sam-
pling strategy is optimal for both publishers shrinks when products are
perceived as less substitutable (e.g., when the readers with left-wing
preference consider a right-wing newspaper a less good alternative).
For the publishers, this results in less ability to acquire consumers who
purchase from the competitor. Consequently, the publishers try to gener-
ate as much revenue as possible from the advertising market by
employing a free content strategy. This result mirrors the ndings from
the monopoly case: as the degree of horizontal product differentiation
increases and publishers tend to become monopolists in their respective
market segment, they should employ a free content strategy once adver-
tising effectiveness exceeds a certain threshold level.
7. Conclusion
This paper analyzed digital content strategies when content sam-
pling serves the dual purpose of disclosing content quality and generat-
ing advertising revenue. One of the key features of the model is that
consumers evaluate free samples of their choice within the limit set
by the publisher. Consumers then use the information gathered from
the free samples to update their prior expectations about content qual-
ity in a Bayesian fashion. Taking consumers' quality updating into ac-
count, the publisher can adopt a sampling strategy, a paid content
strategy, or a free content strategy.
We derived several important results. First, we show in our general
framework how the publisher's advertising-sales revenue ratio and
hence its optimal content strategy is determined by characteristics of
both the content market and the advertising market. We capture the
characteristics of the content market by the elasticity of consumers' up-
dated quality expectations and the elasticities of content demand with
respect to price, sample size, and posterior quality. The corresponding
characteristic in the advertising market is the price elasticity of advertis-
ing demand. We show that, all else equal, the advertising-sales revenue
(B)
ratio is higher the lower the price elasticity of content demand, the
higher the elasticity of content demand with respect to sample size,
and the lower the price elasticity of advertising demand. In addition,
managers can expect the ratio of advertising revenue to sales revenue
to be low when sampling increases consumers' quality expectations
(resulting from the expansion effect) and high when sampling reduces
quality expectations (resulting from the cannibalization effect).
Second, when consumers make purchase decisions based on the
price of the content behind the paywall, content demand depends on
consumers' posterior quality expectations, which can be inuenced by
the publisher through its sampling decision. Expected content demand
has the natural properties that it is decreasing in price and increasing in
expected posterior quality. Further, sampling has a demand-enhancing
effect through consumers' learning when prior expectations are suf-
ciently low (even though sampling produces a cannibalization effect).
We uncovered the rule of thumb that sampling increases content de-
mand if the elasticity of consumers' updated expectations exceeds the
ratio of sampled to paid content.
Third, we characterize the publisher's optimal content strategy
when consumers are uncertain about actual content quality and learn
about it through inspection of free samples. We identify two cut-off
values that determine the publisher's optimal content strategy: a
lower bound that depends on prior quality expectations (separating
paid from sampling strategies) and an upper bound that depends on
posterior quality expectations (separating sampling from free content
strategies). From a managerial perspective, it can be optimal to reduce
both prior quality expectations and content demand in order to gener-
ate advertising revenue. In addition, it can be optimal for managers to
adopt a paid content strategy and to refrain from revealing high quality
through free samples.
We also explore several model extensions that are relevant for mana-
gerial decision making. First, when the publisher also includes advertise-
ments in the paid content, the analysis shows that the cut-off values
between the content strategies depend not only on the relation between
advertising effectiveness and updated quality expectations, but also on
consumers' attitudes towards advertisements. Second, when the willing-
ness to pay for advertisements is related to content quality, we show that
a free content strategy can be optimal even when advertising effective-
ness is low. Third, we show that under competition the advertising effec-
tiveness is a key driver of the publishers' equilibrium strategy choices. This
analysis also sheds light on recent developments in the newspaper indus-
try and explains why publishers have moved away from pure advertising-
nanced business models to metered models (Abramson, 2010).
Our general framework offers several avenues for future research.
Regarding consumers, we assume they correctly update quality expec-
tations based on their sample experience. One alternative is to assume
 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206 203

a consistent bias in the consumers' judgments. In addition, in circum- A.2. Proofs

stances where the rm selects the samples, consumers are likely to ad-
Proof of Proposition 1. By strict concavity of the prot function, the
just (discount) observed quality, assuming that the publisher has
solution to the problem in Eq. (2) must satisfy the necessary and suf-
provided a non-representative set of samples to choose from in order
cient rst-order conditions
to persuade them to buy the paid content. Further, one could assume
that consumers do not evaluate the quality of all free samples because e n
e n pc Dp; n; V
of sampling costs, for example due to the opportunity cost of time or Dp; n; V s 1 0 A:1
p
mental costs. One could also enrich the model by allowing for internal
competition, where the publisher offers two websites to serve different !
e n Dp; n; V
Dp; n; V e n
categories of consumers, which relates to the versioning literature.11 pcs e n
V
n e
V A:2
Thus, there are several further directions which research in these
areas could take. We view this paper a step in this process and hope ann an 2 3 0
the paper encourages work in these and related directions.
and the constraints 1p = 0, 2n = 0, and 3(n N) = 0, where the
Appendix A i's are non-negative real numbers (whose existence is assured by the
KuhnTucker theorem). Suppressing the arguments of content demand,
A.1. Sampling from a uniform distribution (A.1) can be rewritten as
 
pcs 1
A.1.1. The Pareto distribution 1 1 : A:3
p p D
A random variable X has a Pareto distribution with parameters w0
and (w0 N 0) and ( N 0) if X has a density Dividing Eq. (A.2) through p and substituting from Eq. (A.3) produces

8
< w0   
for x N w0 1 D D e ann an 2 3
f xjw0 ; 1 1 1 V n 0:
:x p D e
n V p p p
0 otherwise:

w0 1
For N 1 the expectation of X exists and it is given by EX . Re- Recalling that n a (from the inverse function theorem) and
1 a n
garding sampling from a uniform distribution, we use the following using the denitions of the respective elasticities, the preceding equa-
result. tion can be rearranged to obtain
Theorem. (DeGroot, 1970)  
  
pD 1 e 1 1 2 3 :
12
Suppose that X1, , Xn is a random sample from a uniform distribu- 1 1 n V V A:4
n
tion of the interval (0, W), where the value of W is unknown. Suppose an p D a a
also that the prior distribution of W is a Pareto distribution with parameters
w0 and such that w0 N 0 and N 0. Then the posterior distribution of W Under a sampling strategy there is an interior solution and hence the
when Xi = xi (i = 1, , n) is a Pareto distribution with parameters w0 and k's are zero. Thus, Eq. (A.4) can be rewritten as
+ n, where w0 = max{w0, x1, , xn}.
an n V V
Proof. For w N w0, the prior density function of W has the following  :
Dp 1 1
form: a p

1 Proof of Proposition 2. (a) In order to calculate Ee

v0 n when v0 b V ,
w :
w1 we rst derive the distribution of e
v0 n maxfv0 ; V 1 ; ; V n g. Before
doing so, we state a preliminary fact: The distribution function of
Furthermore, (w) = 0 for w w0. The likelihood function fn(x1, , xn|w) M = max{V1, , Vn} is given by
of Xi = xi (i = 1, , n), when W = w (w N 0) is given by13:
( F M t PrfmaxfV 1 ; ; V n g t g
1
f n x1 ; ; xn jw f x1 jw f xn jw for maxfx1 ; ; xn g b w PrffV 1 t g fV n t gg
wn  n
0 otherwise: n t
Pr fV i t g : A:5
i1 V
It follows from these relations that the posterior p.d.f. (w|x1, , xn) will
be positive only for values w such that w N w0 and w N max{x1, , xn}. As an immediate implication, the density function of M is given by
Therefore, (w|) N 0 only if w N w0. For w N w0, it follows from Bayes'
theorem that nt n1
f M t n : A:6
V
1
wjx1 ; ; xn f n x1 ; ; xn jww
wn1 Next, we derive the density function of ev0 n. By denition, e
v0 n
cannot be smaller than v0 . Therefore, e v0 n v0 if and only if
(the marginal joint probability density function fn(x1, , xn) of X1, , Xn is maxfV 1 ; ; V n g v0 . The probability of this event follows from
a normalizing constant). Eq. (A.5) and it is given by
 n
v0
11 F M v0 :
For instance, The Boston Globe operates the ad-supported site boston.com and the V
subscriber-only site BostonGlobe.com.
12
Theorem 1, p. 172.
13
Given W = w, the random variables X1, , Xn are independent and identically distrib- v0 n N v0 , let e
For e F  denote the truncated distribution function
uted and the common probability density function of each of the random variables is f(xi|w). of e
v0 n. After removing the lower part of the distribution, we
204 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206

 e
e content strategy, 1 = 3 = 0, leading to p NV=4 and n = 0. Under

F t F M t F M v0 for t v0 ; V . This implies f t f M t for
have
t v0 ; V , and hence a free content strategy, we have that p = 0 and n = N.

Proof of Proposition 3. Using Lemma 3, it is straightforward to derive

n1
ef t nt the prots under a free content strategy (FC) and a paid content strategy
n ; if v0 t V
V (PC). The prots are given by, respectively, FC = ( 1)N and
PC NV=8 . Comparing the two prots shows that FC PC if
by Eq. (A.6). The distribution of e
v0 n has a mixed structure with and only if N V8 1. The prot under a sampling strategy (SC) follows
2
from Lemma 3 and is given by SC NV 16V 2 642 =256.
 n Employing a sampling strategy is optimal if SC N PC and SC N FC. It
v0

Prfe
v0 n v0 g A:7 is immediate that these conditions hold if V8 ; V8 2 . A paid content
V
strategy is optimal if PC SC and PC FC, that is, if V8 . A free con-
and density tent strategy is optimal if FC SC and FC PC, that is, if V8 2.

Proof of Proposition 4. At an interior solution, the optimal sample size

n1 n satises the rst-order condition
ef t nt
n ; if v0 t V: A:8
V e n e n

 V V 2n
Nn 0:
4 4 N
The expectation of this mixed distribution is given by
For a corner solution involving n = 0, the KuhnTucker conditions imply
 n Z
e 0 Ve 0
v0 nt n
V NV
E e
v0 n v0 V dt 0 :
v0 V 4 4
n1 n1
v0 nV
n : At the other extreme, when n = N, the KuhnTucker conditions require
n 1V
that
e N
V
Substituting this expression into Eq. (7) produces Eq. (8). (b) If v0 V, 2 0 :
4
then ev0 n is equal to v0 , which in turn implies that E ev0 n v0 .
Substituting this expression into Eq. (7) yields Eq. (9).
Proof of Lemma 4. Using the denition of V e n in Eq. (10), the lower
Proof of Lemma 1. If v0 b V, the quality gap can be expressed as bound can be expressed in terms of the underlying model parameters as

2 1 Nv0
n1 n1 :
e n V v0 nV 1 :
V A:9 1612
n
2 2 n1n 1V
Settingv0 V and letting yields that V8. Likewise, we have that
Clearly, the sign of the quality gap depends only on the sign of numera-
tor (A.9). The latter can easily be rearranged to obtain Eq. (11). If v0 V, NV
2:
the quality gap can be written as 8 N1

 Letting , we obtain V8 2.
e V n v0 V V
V n ;
2 2 n1 A.3. Analysis of the competitive case

which is strictly positive by our assumptions. We study competition between two publishers indexed by i = 1,2
and suppose that they offer differentiated information goods to a popu-
Proof of Lemma 2. Differentiating Eq. (12) with respect to n yields
lation of consumers through an online platform. We frame the analysis

e nV
NnV e n p in terms of the newspaper market and assume that the readers (con-
DE p; n sumers) can be politically ranked from left to right on the political spec-
 :
n NnV e n 2 trum captured by the unit interval [0,1] (Gabszewicz, Laussel, & Sonnac,
2005). Horizontal differentiation captures different editorial opinions,
e nV
e n N 0, which
Clearly, sampling is demand-enhancing if NnV and we assume that the publishers are located at the extremes of the
e nn
V n political spectrum at x1 = 0 and x2 = 1, respectively. Vertical differen-
can be rewritten as e n
N .
V Nn tiation captures the rms' different content qualities such as the level of
Proof of Lemma 3. The optimal decisions on the size of the sample and investigative reporting (which are unrelated to the publishers' political
on the price follow from solving the KuhnTucker conditions in orientation).
Proposition 1.14 Under a sampling strategy, the k's are zero and it follows We again assume that the perceived quality qi(ni) of information



that p NV 82 V =64 and n N 8V =16 . Under a paid good i is equal to its number of content parts Ni multiplied by its expect-
ed posterior quality, that is, qi ni Ni EV i jv1i ; ; vni . Consumers make
a discrete choice and decide which of the two information goods to pur-
chase. A consumer's conditional indirect utility from buying information
good i is given by
14
It is straightforward to show that the objective function is concave for all parameter
values. ui x; pi ; ni qi ni jxxi jpi ;
 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206 205

where x [0,1] is the consumer's political orientation and the parame-

ter N 0 captures the sensitivity to horizontal mismatch |x xi|. Intui-
tively, the mismatch arises because consumers make a discrete choice
and cannot purchase the information good that perfectly matches
their political preferences. Political orientations are drawn indepen-
dently across consumers from a uniform distribution over the interval
[0,1].
The publishers compete for consumers by making their pricing and
sampling decisions. To derive expected content demands, we determine
the location ^x of the consumer who is indifferent between buying from
publisher 1 and from publisher 2 for given prices p = (p1, p2) and sam-
ple sizes n = (n1, n2). Clearly, the location of the indifferent consumer
^xp; n is a solution to the indifference condition u1 ^xp; n u2 ^xp; n,
which ensures that the indirect utilities from the two information goods
Fig. A.1. Strategy choices and corresponding prots.
are the same. With linear mismatch, the consumer located at ^xp; n seg-
ments the market: Consumers located to the left of ^xp; n purchase
from publisher 1, while consumers located to the right of ^xp; n pur- consumers believe that publisher 1 offers a lower content quality than
chase from publisher 2.15 Ignoring cannibalization, publisher 1 thus publisher 2.
faces ^xp; n consumers, while publisher 2 faces 1^xp; n consumers. The key insights from the analysis of the competitive case can be
To capture that sampling not only reveals quality but also cannibalizes summarized as follows: If consumers' prior expectations about content
N n
sales, we let i i denote the conditional purchase probability given qualities are similar and advertising effectiveness is high, it is optimal
Ni
sample size ni. Hence expected content demands can be expressed as for both publishers to adopt a free content strategy in equilibrium. In-
stead, if the gap in prior expectations is large and advertising effective-
N1 n1 N2 n2 ness is high, both publishers should adopt a sampling strategy. Further,
E
D1 p; n ^xp; n and E
D2 p; n 1^xp; n: the parameter region in which the sampling strategy is optimal for both
N1 N2
publishers shrinks when consumers are more sensitive to horizontal
mismatch and products thus are less substitutable.
Consumers thus choose their preferred publisher based on prices and To understand these insights, it is important to notice that each point
posterior quality expectations and subsequently purchase the content in the (,v01)-space in Fig. A.2 corresponds to a (pure-strategy) Nash
N i ni
with probability . Similar to the monopoly case, the sampling deci- equilibrium of the matrix game described above.16 The solid line depicts
Ni
the cut-off threshold between the two types of Nash equilibria: sam-
sion ni therefore has a direct effect on expected content demand
pling (SC, SC) and free content (FC, FC). Counter to intuition, it is not
through the conditional purchase probability (a cannibalization effect)
per se optimal to choose a free content strategy when advertising effec-
and an indirect effect on ^xp; n (an expansion effect). Note that publish-
tiveness is high. What matters in addition is the gap in prior expecta-
er i's expected content demand is zero under a free content strategy due
tions among the two publishers: if it is small, then both rms should
to the cannibalization effect.
adopt a free content strategy. Instead, if the gap in prior expectations
Publisher i makes its pricing and sampling decisions so as to
is large, then both rms should adopt a sampling strategy. Intuitively,
 
E E n publisher 1 has a stronger incentive to reveal its higher than expected
max i p; n pi Di p; n i i ni
pi ;ni Ni qualityand it is a best-response of publisher 2 to also adopt a sampling
s:t: pi 0 strategy (that is, publisher 2 attains a higher prot under a sampling
0 ni Ni ; strategy than under a free content strategy when taking publisher 1's
strategy as given).
where i is the advertising effectiveness of publisher i's advertising. Fig. A.2 also illustrates that the parameter region in which the sam-
Compared to the monopoly case, each publisher now has to take into ac- pling strategy is optimal for both publishers shrinks when consumers
count the rival's choice of strategy to make its optimal decision. Thus, are more sensitive to horizontal mismatch . Intuitively, a higher
there are nine possible outcomes, summarized in Fig. A.1. If both rms means that products are less substitutable from the viewpoint of the
use a paid content strategy, the rms' corresponding (expected) prots consumers (e.g., when readers with left-wing preferences consider a
are denoted by PP PP
1 and 2 , respectively (and likewise for the other out- right-wing newspaper a less good alternative). For the publishers, this
comes). For each outcome, the prot levels can be obtained by solving results in a lower ability to engage in business stealing in the content
the publishers' decision problems. The optimal strategy choices are market. Consequently, the rms try to generate as much revenue as
then obtained as a Nash equilibrium of the matrix game depicted in possible in the advertising market by employing a free content strategy.
Fig. A.1. This comparative statics result mirrors the ndings from the monopoly
To analyze the optimal strategy choice, we focus on a market envi- case (cf. Fig. 3): When increases and publishers become monopolists
ronment where consumers have different prior beliefs about the con- on their respective market segment, the rms should employ a free con-
tent quality of the two publishers. Specically, we suppose that the tent strategy once exceeds a certain threshold levelirrespective of
consumers' minimum estimate v01 differs from v02, while the publishers the gap in prior expectations (observe that this threshold level lies at
are actually symmetric and in particular offer the same quality spectrum around = 2 in Fig. A.2).
(V 1 V 2). Fig. A.2 illustrates the publishers' optimal strategy choices as Up to now, the focus has been on strategy choices in the upper-right
a function of the gap in prior expectations and advertising effectiveness corner of the matrix game in Fig. A.1. The reason is that these strategy
(i ). The gure holds v02 constant (at v02 = 1) and plots different combinations capture the essence of the debate in the newspaper in-
values of v01 on the vertical axes. Notice that where v01 b v02, dustry: whether to offer the content for free or to employ a metered
model. Of course, there is also a symmetric industry conguration

15 16
See Anderson, de Palma, and Thisse (1992) for an in-depth treatment of Hotelling- The code to numerically compute the Nash equilibria is available from the authors up-
type models. on request.
206 D. Halbheer et al. / Intern. J. of Research in Marketing 31 (2014) 192206
Fig. A.2. Equilibrium strategies (for 02 = 1, = 2, V i 3, and Ni = 10).
where both publishers employ a paid content strategy. Unsurprisingly,
this equilibrium arises if advertising effectiveness is low. In addition,
there are asymmetric strategy choices such as (PC, SC), which are optimal
when consumers overestimate the content quality of publisher 1
(and hence the incentive to not reveal low quality) and underestimate
the content quality of publisher 2 (and hence the incentive to reveal
high quality through free samples).
References
Abramson, J. (2010). Sustaining quality journalism. Daedalus, 3944.
Ackerberg, D. A. (2003). Advertising, learning, and consumer choice in experience good
markets: An empirical examination. International Economic Review, 44(3), 10071040.
Akerlof, G. A. (1970). The market for lemons. Quarterly Journal of Economics, 84(3), 488500.
Anderson, S. P., de Palma, A., & Thisse, J. -F. (1992). Discrete choice theory of product differ-
entiation. Cambridge, MA: MIT Press.
Anderson, S. P., & Gabszewicz, J. J. (2006). The media and advertising: A tale of two-sided
markets. In V. Ginsburgh, & D. Throsby (Eds.), Handbook of the economics of art and
culture (pp. 567614). Elsevier.
Anderson, S. P., & Renault, R. (2006). Advertising content. American Economic Review, 96,
93113.
Bagwell, K. (2007). The economic analysis of advertising. In M. Armstrong, & R. Porter
(Eds.), Handbook of industrial organization, Vol. III. (pp. 17011844). Elsevier.
Bawa, K., & Shoemaker, R. (2004). The effects of free sample promotions on incremental
brand sales. Marketing Science, 23(3), 345363.
Bhargava, H. K., & Choudhary, V. (2008). Research note: When is versioning optimal for
information goods? Management Science, 54(5), 10291035.
Boom, A. (2009). Download for freeWhen do providers of digital goods offer free
samples? Working Paper. Copenhagen Business School.
Burke, R. R., & Srull, T. K. (1988). Competitive interference and consumer memory for
advertising. Journal of Consumer Research, 15(1), 5568.
Chellappa, R. K., & Shivendu, S. (2005). Managing piracy: Pricing and sampling strategies
for digital experience goods in vertically segmented markets. Information Systems
Research, 16(4), 400417.
Cheng, H. K., & Tang, Q. C. (2010). Free trial or no free trial: Optimal software product
design with network effects. European Journal of Operational Research, 205,
437447.
DeGroot, M. (1970). Optimal statistical decisions. New York: Mc Graw-Hill.
Dorfman, R., & Steiner, P. O. (1954). Optimal advertising and optimal quality. American
Economic Review, 44(5), 826836.
Erdem, T., & Keane, M. P. (1996). Decision-making under uncertainty: Capturing dynamic
brand choice processes in turbulent consumer goods markets. Marketing Science,
15(1), 120.
Erdem, T., Keane, M. P., & Sun, B. (2008). A dynamic model of brand choice when price
and advertising signal product quality. Marketing Science, 27(6), 11111125.
Faugre, C., & Tayi, G. K. (2007). Designing free software samples: A game theoretic ap-
proach. Information Technology Management, 8, 263278.
Gabszewicz, J. J., Laussel, D., & Sonnac, N. (2005). Attitudes toward advertising and price
competition in the press industry. In V. A. Ginsburgh (Ed.), Economics of art and
culture (pp. 6174). Elsevier.
GlobeNewswire (April 5). Wall Street Journal offers digital open house today; access in-
cludes online, apps for iphone, ipad. http://www.globenewswire.com/-newsroom/
news.html?d=250593
Godes, D., Ofek, E., & Sarvary, M. (2009). Content vs advertising: The impact of competi-
tion on media rm strategy. Marketing Science, 28(1), 2035.
Heiman, A., McWilliams, B., Shen, Z., & Zilberman, D. (2001). Learning and forgetting:
Modeling optimal product sampling over time. Management Science, 47(4),
532546.
Hotz, V. J., & Xiao, M. (2013). Strategic information disclosure: The case of
multiattribute products with heterogeneous consumers. Economic Inquiry,
51(1), 865881.
Kind, H. J., Nilssen, T., & Srgard, L. (2009). Business models for media rms: Does com-
petition matter for how they raise revenue? Marketing Science, 28(6), 11121128.
Milgrom, P., & Roberts, J. (1986). Price and advertising signals of product quality. Journal
of Political Economy, 94(4), 796821.
Newspaper Association of America (2012). Paid digital content benchmarking study.
Rysman, M. (2009). The economics of two-sided markets. Journal of Economic Perspectives,
23(3), 125143.
Shapiro, C., & Varian, H. R. (1998). Information rules. Boston, MA: Harvard Business School
Press.
Sun, M. (2011). Disclosing multiple product attributes. Journal of Economics and Management
Strategy, 20(1), 134145.
Wang, C., & Zhang, X. (2009). Sampling of information goods. Decision Support Systems,
4(1), 1422.
Wilson, R. (1985). Multi-dimensional signalling. Economics Letters, 19(1), 1721.
Xiang, Y., & Soberman, D. A. (2011). Preview provision under competition. Marketing
Science, 30(1), 149169.