From Firm-Controlled to Consumer-Contributed: Consumer Co-Production of Personal Media Marketing Communication

This paper explores the idea of adapting a co-production strategy from service marketing to marketing communication sent to personal media. It provides empirical evidence supporting a co-production approach applied as a communication strategy in the context of a text message mobile coupon marketing campaign.

This is the author’s final, peer-reviewed manuscript as accepted for publication. The
publisher-formatted version may be available through the publisher’s web site or your
institution’s library.
This item was retrieved from the K-State Research Exchange (K-REx), the institutional
repository of Kansas State University. K-REx is available at http://krex.ksu.edu
From firm-controlled to consumer-contributed: consumer co-
production of personal media marketing communication
Todd J. Bacile, Christine Ye, Esther Swilley
How to cite this manuscript
If you make reference to this version of the manuscript, use the following information:
Bacile, T. J., Ye, C., & Swilley, E. (2014). From firm-controlled to consumer-contributed:
Consumer co-production of personal media marketing communication. Retrieved from
http://krex.ksu.edu.
Published Version Information
Citation: Bacile, T. J., Ye, C., & Swilley, E. (2014). From firm-controlled to consumer-
contributed: Consumer co-production of personal media marketing communication.
Journal of Interactive Marketing, 28(2), 117-133.
Copyright: © 2013 Direct Marketing Educational Foundation, Inc., dba Marketing
EDGE. Published by Elsevier.
Digital Object Identifier (DOI): doi:10.1016/j.intmar.2013.12.001
Publisher’s Link: http://www.sciencedirect.com/science/article/pii/S1094996813000637

1
From Firm-Controlled to Consumer-Contributed: Consumer Co-Production of Personal
Media Marketing Communication
Todd J. Bacile
Loyola University New Orleans, Joseph A. Butt, S.J., College of Business, 6363 St. Charles
Avenue, New Orleans, LA 70118, USA
Christine Ye
Westminster College, Bill and Vieve Gore School of Business, 1840 South 1300 East, Salt Lake
City, UT 84105, USA
Esther Swilley
Kansas State University, College of Business Administration, Department of Marketing, 201A
Calvin Hall, Manhattan, KS 66506-0506, USA

2
Abstract
Fueled by the sociocultural shift from firm-controlled to consumer-contributed media, the
researchers explore the idea of adapting a co-production strategy from service marketing to
marketing communication sent to personal media. Eleven field experiments with firms, along
with a structural model tested on survey data, provide empirical evidence supporting a co-
production approach applied as a communication strategy in the context of a text message mobile
coupon marketing campaign. The results demonstrate a co-produced direct marketing
communication strategy increases attitude toward the communication, purchase intent, and
purchase activity, while also acting as a risk-reducing mechanism. Furthermore, perceived
customization of the communication interacts strongly with risk perception and marginally with
coupon proneness as related to attitude toward the communication when marketers enter the
world of consumers’ personal media. A push versus pull framework and a co-produced
communication framework are put forth to suggest various areas marketers can make available
for consumers to co-produce in a marketing communication exchange.
“[C]ustomers in contexts other than those traditionally considered service interfaces are in reality
involved in service-like processes.” – Christian Grönroos (2006, p. 329)
Introduction
The rising importance and prevalence of personal media on mobile phones are creating
new opportunities for marketers, but a lack of strategies in these communication channels is
surfacing (Hennig-Thurau et al. 2010). A key differentiating element tempering firms’ success is
personal media are more important to consumers than mass media, causing many consumers to
elect not to receive intrusive marketing messages and to become less tolerant of irrelevant
messages. The prominence of identifying effective strategies is amplified by the upsurge in direct
marketing communication, hereafter referred to as communication, sent to such media. For
example, spending for communication sent to mobile and social media will reach $8 billion and
$5 billion, respectively, by 2016 (VanBoskirk 2011).

3
For our purposes, personal media refers to highly individualized and important
communication tools primarily for interpersonal communication, such as the telephone, chat, text
messaging, and social media through a mobile device (Lüders 2008). The mobile phone is the
platform extraordinaire for such media as it is personal and drives media use. Both the individual
nature and importance of personal media within consumers' lives are apparent (Hennig-Thurau et
al. 2010). As such, psychological barriers exist which firms must overcome when sending
communication through personal media, such as risk perception of a negative outcome including
privacy concerns, irrelevant messages, message volume, and intrusiveness (Deighton and
Kornfeld 2009; Sultan, Rohm, and Gao 2009). Currently, the primary strategy firms use to
reduce risk perception of a negative outcome is to grant consumers the right to opt-in and opt-out
(i.e. give or remove permission) to receive communication sent to personal media (Barwise and
Strong 2002). However, other than these two all-or-nothing inputs, firms typically do not allow
consumers to participate in further decision making inputs into the communication process. This
lack of consumer participation is reminiscent of traditional mass media where the firm views the
consumer as a passive audience member (Wind and Rangaswamy 2001). Aside from opt-in or
opt-out decisions, in both mass and personal media, the time, frequency, subject, and type of
communication are generally decided by marketers. This consistency between mass and personal
media is surprising, considering labels such as "customized" and "personalized" that often
differentiate personal media from mass media (Shankar and Balasubramanian 2009).
The current paper suggests firms view personal media communication as an interactive
process with consumers in contrast to the classic one-way marketing communication model. This
proposed type of exchange positions personal media communication similar to the modern
viewpoint of service as interactive firm-consumer processes. Various service philosophies such

4
as service logic (Grönroos 2006), service-dominant logic (Vargo and Lusch 2004), and unified
services theory (Sampson and Froehle 2006) each discuss consumers providing inputs to firms
during interactive processes as co-producing activities central to service exchange. An example
of such inputs includes consumers specifying their desired preferences within a service offering.
Thus, the purpose of this research, and one hypothesized to improve firm outcomes and
consumer psychological and behavioral responses, is to empirically assess if a service strategy
such as co-production can be adapted as a marketing communication strategy in personal media.
This study contributes to marketing theory and practice in a number of ways. The
theoretical contribution of this manuscript is to extend the domain of service to marketing
communication. Personal media communication is becoming service-like in nature. Technology
has rendered complex formerly simple dissemination of firm communication. Marketing
communication is evolving from a one-way, firm-controlled message subject with a simple effort
allocation approach, to a two-way, consumer participatory service-like exchange merging firm-
consumer processes. Firms should engage consumers by enabling the co-production of some
aspect of the communication process. This is a unique perspective not present within emerging
models of direct marketing via personal media (e.g. Deighton and Kornfeld 2009; Shankar et al.
2010; Sultan, Rohm, and Gao 2009). As shown in two studies, the extension of marketing
communication theory is possible with the application of this service strategy. The authors offer
a push versus pull framework and a personal media co-production framework discussing many
characteristics of these communication exchanges that are ideally suited to co-produce.
The empirical contribution of the current research demonstrates large improvements in
the effectiveness of promotional communication in the form of a text message mobile coupon via
a series of field experiments in Study 1. The results imply co-producing communication

5
improves consumer response. Using survey data with structural equation modeling then uncovers
mechanisms underlying this effectiveness in Study 2, which hones in on perceived customization
as the key driver behind enhanced response to a firm’s mobile coupon promotion. Risk
perception is present when communication is sent to personal media, and we show a co-
production strategy attenuates this risk. In addition to acting as a risk-reducing mechanism,
results suggest co-production improves attitude toward the communication and purchase intent,
with perceived customization of the communication interacting strongly with risk perception and
marginally with coupon proneness as each relates to attitude toward the communication. The
remainder of this paper discusses communication theory and the theoretical development on the
ways in which personal media communication is evolving toward a service-like offering. The
hypotheses and conceptual model are formulated and assessed in two studies, followed by a
discussion of all findings, theoretical implications, and managerial ramifications.
Theoretical Development
Communication Theory
Marketers using traditional mass media to disseminate communication typically follow a
process consistent with the transmission model of mass communication (Shannon and Weaver
1949). The basis of this model is information processing and communication theory. These steps
within the communication process have been adapted to various communication models, yet the
core steps remain the same. Communication is a process, where a breakdown in any step disrupts
the entire process. A source creates a message by encoding it into a format conducive for a
particular medium. The message is sent via a medium, upon which a receiver must decode the
message. Encoding the proper message containing information of interest to a receiver and using

6
the correct medium should result in optimal communication between two parties. Passive
feedback is generated back to the source after receiving a message (i.e. ignoring a message,
remembering a message for later use, making a purchase, or telling others) with limited ability
for specific direct communication in response to a message (i.e. active feedback). This
overarching control over one-way communication is consistent with the political economy of
communication theory (Graham 2007; Innis 1942), which posits that a select few controlling
entities, such as owners of communication networks and resourceful brands, attempt to persuade
the thoughts and actions of consumers because these entities have the power to do so. Mass
media marketing communication is based upon this format of one-way communication.
In large part, mass media marketing communication is sub-optimal and described as
“wasteful marketing” (Sheth and Sisodia 2006, p. 7). Mass media’s ineffectiveness, coupled with
the interactive nature of emergent technologies such as personal media, are causing new
paradigms to materialize (Deighton and Kornfeld 2009; Shankar et al. 2010; Sultan, Rohm, and
Gao 2009). Yet, such newer frameworks are still derived from mass media’s centrally managed
philosophy. Emerging frameworks are adapting mass media’s need to infer and exploit
consumers’ information to craft communication in a manipulative or intrusive manner. As an
example, Deighton and Kornfeld (2009) offer up multiple consumer roles emerging across a two-
by-two linear matrix: Accessibility ↔ Identity and Information ↔ Meaning. Within this matrix
the researchers posit that the most attractive form of interactivity is a higher degree of knowing
an individual consumer’s identity and a higher degree of brand meaning in specific contexts for
individual consumers. Largely absent from this framework is a consumer role with high degrees
of identity and meaning, while avoiding direct marketing’s need to infer, exploit, or compete
against consumer-provided primary or secondary information.

7
This paper suggests that firms enabling consumers to actively participate in some aspect
of the communication process prior to receiving a marketing message via personal media avoids
the need for firms to infer or exploit consumer information. The difference from mass media’s
communication model is a consumer participating (i.e. co-producing) by providing preferences
to a firm prior to sending marketing messages to a consumer. Firm-consumer participation in this
context is similar to the provision of service.
Co-Production and A Service Orientation
A consistent theme across many service philosophies is co-production, which is defined
as consumers participating with firms in at least one interactive decision-making process prior to
the usage of a good or service (Etgar 2008; Grönroos 2006). In service exchange, the co-creation
of value is always present, but co-production may vary from none at all to a high degree
depending on the level of interactive firm-consumer processes prior to usage (Vargo and Lusch
2008). Firms offer co-production opportunities by not only enabling consumers to participate in
labor, but also using customer-provided information, such as consumers providing preferred
characteristics and attributes within an offering (Grönroos 2006; Sampson and Froehle 2006).
Etgar (2008, p. 97) states “co-production is an explicit result of decision making by consumers
reflecting their own preferences.” Consistent with this rationale, researchers use customization to
represent or establish a direct link with co-production (Auh et al. 2007; Etgar 2008).
Co-production of some aspect of the communication process occurs when a consumer
decides a marketing message’s delivery time, frequency, recipient, subject, format, preferred
media channel, or any other message characteristic prior to receiving a marketing
communication. Consumers providing preferences to firms for any of these characteristics are in

8
effect co-producing the communication process. Whereas mass media limit a consumer as an
inactive participant, co-producing the communication process in personal media enables active
participation of a consumer to decide some aspect of a communication prior to receiving. This
communication-based view of co-production is consistent with Etgar’s (2008) service-based
view of co-production. Positing the domain of co-production as applicable to personal media
marketing communication is a logical evolutionary path when one considers the progression
toward the modern understanding of service.
The emergence of a service orientation is expanding the purview of service from earlier
connotations of an intangible act, such as a gas station attendant pumping fuel, to firm-consumer
interactions, exchanges, and processes. The modern definition of service is the application of
interactive firm-consumer processes to create value. This definition captures the essence across
many service philosophies, including service logic (Grönroos 2006), service-dominant logic
(Vargo and Lusch 2004), and unified services theory (Sampson and Froehle 2006).
The Nordic school’s service logic distinguishes a service from a physical good by
proposing that the former is a firm-consumer process exchange (Grönroos 2006). Services are
processes, rather than objects for economic, transactional exchange; and these firm-consumer
processes take place in interactions. Hence, service logic defines service as processes which aim
to solve consumers’ problems through firm-consumer interactions, be it in-person or technology-
mediated. Consumers co-producing service processes are co-creators of value.
Using similar reasoning, service-dominant logic implies all firms are service providers,
with service being the fundamental basis of exchange (Vargo and Lusch 2004; 2008). Service-
dominant logic suggests a shift is occurring, where intangible resources and co-creation of value
through service provision are replacing the traditional view of marketing based on the economic

9
exchange of goods. Service-dominant logic defines service as the application of specialized
processes by one entity to benefit another entity. Firms use such processes with consumers to co-
create value. A distinct component of value co-creation is co-production (Vargo and Lusch
2008). Service always co-creates value, yet involving consumers as co-producers is optional.
Unified services theory’s view of service is consistent with service logic and service-
dominant logic. Unified services theory differentiates service processes from non-service
processes by involving a consumer to provide specific inputs into the production of service
(Sampson and Froehle 2006). Production processes that do not require specific consumer inputs
are managed differently, such as mass manufacturing. Groups of consumers contribute ideas and
market research within a manufacturing context, yet their direct inputs are absent (i.e. a specific
product is not produced for a specific consumer based on inputs). Co-production is a method for
a consumer to provide specific inputs into the service production process.
Associating co-production with promotional communication processes separate and
distinct from product offerings may seem incongruent. Research illustrates this incongruence by
often focusing on co-production of a good or service product (e.g. Auh et al. 2007; Bendapudi
and Leone 2003). However, co-production is applicable to underlying process activities and
interactions to exchange information that will ultimately co-create value (Grönroos 2006). Such
processes can produce accurate and useful personal media communication, leading to positive
outcomes for firms and consumers (Duncan and Moriarty 2006).
Co-production of some aspect of the communication process by enabling consumers to
participate by providing preferences supports the notion that communication exchange is a firm-
consumer encounter becoming process-like (Payne, Storbacka, and Frow 2008). Direct marketing
communication sent to personal media is now a construction process based on mutual

10
understanding between firms and consumers (Firat and Dholakia 2006). This recognition of
communication as a process-like firm-consumer exchange encounter is similar to the modern
view of a service orientation. Within a service orientation marketers adopt a participatory role in
communication iterations with consumers to promote learning, create accurate offers, build
relationships, and co-create value (Ballantyne and Varey 2006; Duncan and Moriarty 2006).
We believe that merging a co-production strategy to personal media communication is
consistent with the co-production consumer engagement model (Etgar 2008). This model
identifies situations where firms should involve consumers in co-production and customization.
The first two stages (co-production antecedent conditions and consumer motivations to co-
produce) are useful to assess if co-producing personal media communication is a viable strategy.
Ideal antecedent conditions in Stage 1 include assessing technological, economic, and cultural
factors. These conditions are prevalent in personal media, as the increasing number of electronic
touch points creates easy to use and low cost methods for consumers to interact with firms
(Prahalad and Ramaswamy 2004). Higher per capita incomes in mature markets tend to have
cultures that place a high value on customization, increasing the likelihood of consumers
preferring to co-produce (Etgar 2008).
The second stage, consumer motivation to co-produce in personal media, is driven by one
particular psychological factor - importance. When tasks, situations, decisions, or encounters are
perceived to be highly important, the integrative control model (Brehm 1972; Wortman and
Brehm 1975) posits people need control, or more accurately, perceived control. Important
situations amplify the necessity to avoid negative outcomes. The concept of importance is
relevant to personal media given that consumers place a high degree of importance upon these
channels. Perceived control enables consumers to tolerate the risk of potentially negative

11
outcomes (Lee and Allaway 2002). To achieve higher perceived control consumers use co-
production and customization (Auh et al. 2007; Bateson 1985). One caveat of co-production and
customization is that a positive manifestation may not result. Complexity and uncertainty can
occur when a consumer is beset with information overload or unsure of the best options
(Huffman and Kahn 1998). However, customization does not produce these negative occurrences
when a consumer is conversant in choice options that best meet their specific needs (Godek,
Yates, and Yoon 2002).
The present paper conducts two studies to assess some of the benefits of a personal media
co-production strategy for firms. With the presence of the antecedent conditions and motivation
to co-produce, we posit co-producing some aspect of the communication process will influence
purchase behavior in Study 1. We support this position by drawing from extant research in co-
producing goods and services, and applying this logic to co-producing communication. Co-
producing a good or a service creates an offering with specific options meeting individual needs
of a consumer, making it more likely a purchase will occur (Wind and Rangaswamy 2001). It is
also possible the mere act of involvement in creating an offering is a source of value, which
ultimately increases purchase behavior (Auh et al. 2007). In either case, co-production functions
as a powerful strategy to improve consumer purchase behavior which has yet to be assessed in
the domain of marketing communication. We propose purchase redemption, which is a response
to a communication resulting in a purchase, will be higher for consumers who co-produce some
aspect of the communication process versus non-co-producers.
H1: Purchase redemption will be higher for consumers co-producing some aspect of the
communication process within personal media compared to consumers who do not
co-produce.

12
Study 1
Study 1 was designed to test H1 under actual market conditions. A mobile marketing firm
agreed to a field experiment to compare purchase redemption of consumers in two groups: those
who co-produce some aspect of the personal media communication process versus those who do
not co-produce. Among personal media consumers use daily, none are more important than
mobile phones (Hennig-Thurau et al. 2010). The most popular form of direct marketing
communication sent to mobile phones are text message mobile coupons (m-coupons) containing
a discount offer and/or promotional information (Dickinger and Kleijnen 2008; Shankar and
Balasubramanian 2009). The particular feature of the m-coupon to be co-produced in Study 1 is
the delivery time of the communication. Delivery time is an ideal feature to investigate in the
current context since the success of m-coupon campaigns highly depend on messages’ time of
delivery (Bacile and Goldsmith 2011; Barwise and Strong 2002). Mobile devices offer users
value-for-time and afford consumers the opportunity to receive information anytime and
anywhere (Clarke 2001). By enabling consumers to decide their preferred delivery time, in effect
consumers are co-producing the communication process, since this type of digital
communication does not exist until consumers receive the message. This operationalization
aligns with the core idea of co-production, “which may take place within the production process
which precedes the usage stage,” (Etgar 2008, p. 98) .
Field Setting and Sample
Four casual restaurants in two cities with populations between 100,000 and 500,000 in
the southeastern United States agreed to participate. Restaurants were the industry of choice due
to ecological validity based on two reasons: restaurants frequently make use of coupons in
marketing promotions and m-coupons are anticipated to be heavily used by the restaurant

13
industry (eMarketer 2010). A total of 11 experiments were conducted. The restaurants used
promotional communication in print media, signage in their surrounding areas, Web sites, and
retail signage outside and inside the restaurants asking consumers to opt-in to a new text message
promotion. The advertisements directed consumers to each restaurant’s Web site, where a Web
form was used by consumers to enter their mobile phone number to opt-in. The Web page
housing the opt-in form randomly presented one of two possible forms to consumers. This
insured random assignment to each of the two groups. On one form the consumers were only
required to opt-in by entering their mobile phone number, thus representing a control group of
non-co-production subjects. In contrast, the other form contained two additional fields
consumers used to select the day of the week and time of the day to receive an m-coupon, thus
representing the treatment group of co-production subjects.
A field study can limit the control of other factors that may influence results, such as one
group having subjects who are more or less technology ready than the other group. However, the
field experiment randomly assigned subjects to each group. The average control group size (n =
116) and average treatment group size (n = 119) in each of the 11 field experiments were
sufficiently large to cancel the main effects of uncontrolled factors.
Measure
The number of purchase redemptions produced by the m-coupon (i.e. the redemption
rate) was assessed. Each consumer in the treatment and control group was sent an m-coupon with
a code number indicating to the restaurant a consumer was redeeming either a co-produced or
non-co-produced m-coupon. At the time of purchase the text message was shown to the cashier

14
who recorded the code. The use of two unique coupon codes made it easy to compare the number
of purchase redemptions in the treatment group compared to the control group.
Analysis and Results
In each of the 11 experiments the same m-coupon wording and discount amount was
issued to consumers in each of the two groups. For example, in one experiment both groups
received an m-coupon good for 25% off a meal; however, the treatment group’s m-coupons were
sent on the day and time as specified by each consumer. The range of discount amount offered
across each of the 11 experiments was 20%-35% with a mean discount amount of 25%.
Table 1 contains details and results of the experiments. All but two of the experiments
were significantly higher as assessed with a chi-square test. When consumers co-produced the
delivery time aspect of the communication process by informing the restaurants of their
preferences, purchase redemptions were higher than those who did not co-produce. Redemption
rates for co-produced m-coupons ranged from 16.0% to 37.5%, compared to a non-co-produced
range of 5.8% to 11.6%, across the 11 experiments providing ample support for H1.
---------------------------------------------
Insert Table 1 about here
---------------------------------------------
Discussion
Across all of the m-coupon campaigns in Study 1, the purchase redemption for
consumers co-producing the communication process was higher compared to non-co-producers.
In addition, one restaurant recorded the number of new versus current customers who opted-in.
New customers accounted for 40% of the opt-in consumers, suggesting purchase redemptions
may have increased incremental sales. Study 1 empirically illustrates the successful application

15
of a co-production strategy to communication. The results imply marketers wishing to enter
consumers’ personal media should avoid standardized communications used in mass media.
Enabling consumers to co-produce some aspect of the communication process creates personally
relevant messages playing to the strength of personal media. Co-producing communications in
this context adds value to consumers, allowing them to receive a more relevant, usable
communication, which in turn leads to a stronger purchase redemption rate for firms.
Study 2
Hypotheses Development and Conceptual Model
Building off the results from Study 1, Study 2 investigates psychological responses to co-
producing some aspect of the communication process for an m-coupon. The conceptual model in
Figure 1 adapts endogenous constructs from studies assessing mass media marketing
communication to a personal media context. In particular, attitude toward the marketing
communication and purchase intent are important antecedents, which typically correlate strongly
with purchase behavior (Ajzen 1991; Muehling and McCann 1993). The effect of co-producing
some aspect of the communication process, which is modeled with perceived customization, and
the effects of perceived risk and coupon proneness are included to hypothesize their effects on
attitude and purchase intent in a personal media context.
----------------------------------------------
Insert Figure 1 about Here
----------------------------------------------
Consistent with Study 1, Study 2 uses a text message m-coupon and the opportunity to
co-produce the delivery time to represent the co-produced aspect of the communication process.
This type of digital communication does not exist until consumers specify when they prefer to

16
receive it. Therefore, perceived customization represents subjects’ perception that the act of co-
producing this characteristic of the message creates a customized delivery time based on a
preference communicated to a firm.
One interesting possibility related to personal media communication is that perceived risk
may be present (Sultan, Rohm, and Gao 2009). Perceived risk is defined here as a consumer's
perception of potentially adverse consequences in relation to receiving a communication from a
firm (Dowling and Staelin 1994). Perceived risk has numerous conceptualizations within the
marketing literature, but often represents a consumer’s perception of a negative outcome in a
given situation or task, such as an unwanted result, cause to worry, or uncertainty (Dowling and
Staelin 1994). The present study adopts Taylor’s (1974) psychological aspect of perceived risk
when receiving a communication. Consumers experience an increase in risk when opening up
their mobile phone to marketers due to unknown factors such as future messages exceeding
optimal frequency, being overly intrusive, and lacking timeliness or relevance (Barwise and
Strong 2002; Dickinger and Kleijnen 2008). Concern over any of these factors increases risk
perception of a potentially negative outcome (Deighton and Kornfeld 2009). This representation
of concern over a potential negative outcome is consistent with psychological discomfort when
referring to perceived risk (Zaltman and Wallendorf 1983).
One mechanism to reduce risk perception is the act of giving a consumer a choice (Taylor
1974). Granting a choice is a form of co-production and customization (Etgar 2008). In turn,
customization increases the perception of control (Auh et al. 2007; Bateson 1985). The
perception of control reduces perceived risk when using remote technology services offered by
firms (Lee and Allaway 2002). Moreover, co-production and customization are theorized as risk-
reducing techniques because consumers can reduce ambiguity, worry, or concern for a negative

17
outcome by exerting their preferences (Etgar 2008). In the current context, co-producing the
communication process creates a communication customized to some degree based on a
consumer’s preference. Thus, perceived customization reduces the perceived risk of personal
media communication.
H2: Perceived customization has a direct, negative effect on the perceived risk of
receiving a communication sent to personal media.
Attitude toward the communication is defined here as a tendency to respond favorably or
unfavorably to a particular marketing communication (Lutz 1985). Attitude toward a traditional
marketing communication is more favorable when the message contains useful and relevant
information (Muehling and McCann 1993). The essence of co-production and customization is
consumers indicating what preferences are more useful and relevant. This logic is applied in the
current context by hypothesizing customization will improve attitude. In addition, perceived
customization has an indirect effect on purchase intent through attitude, as research suggests
attitude fully mediates this relationship (Ajzen 1991; Muehling and McCann 1993).
H3: Perceived customization has a direct, positive effect on attitude toward the
communication sent to personal media.
H4: The effect of perceived customization on purchase intent is fully mediated through
attitude toward the communication.
Risk perception has a significant influence on consumers to engage in new technologies
and is an antecedent of attitude toward new technologies (Pavlou 2003). Swilley (2010) suggests
violation of privacy and uncertainty associated with personal media are potential negative
outcomes, which result in perceived risk having a negative influence on attitude toward mobile
phones. Such potentially negative outcomes are what consumers are apprehensive about
regarding communication sent to personal media (Hennig-Thurau et al. 2010).
H5: Perceived risk has a direct, negative effect on attitude toward the communication.

18
Low risk situations are of little concern to people, but as risk goes above some baseline
context value, there is more of a need to reduce it (Kahneman and Tversky 1979; Taylor 1974).
As previously stated, co-production and customization are methods for risk reduction (Etgar
2008). If this is the case, then the impact of a high degree of perceived risk on attitude should be
attenuated with a high degree of perceived customization. Moreover, perceived customization
should produce a beneficial impact on the risk  attitude relationship, thus improving attitude.
However, this risk  attitude relationship will reduce attitude when there is a lower degree of
perceived customization. In summary perceived customization moderates the effect of perceived
risk on attitude.
H6: A higher degree of perceived customization will attenuate the effect of perceived risk
on attitude toward the communication.
The inclusion of the coupon proneness construct in the model is relevant due to the
experimental stimuli being a communication in the form of an m-coupon. Coupon proneness is
defined as an incremental propensity to respond to a purchase offer because the coupon form of
the purchase offer positively affects purchase evaluation (Lichtenstein, Netemeyer, and Burton
1990). Prior studies examining traditional paper coupons show support for the relationship
between an increase in coupon proneness and an increase in attitude toward a coupon
(Lichtenstein, Netemeyer, and Burton 1990; Mittal 1994; Shimp and Kavas 1984). Similar
support is put forth in a theoretical framework for non-traditional coupons available for
download from Web sites (Fortin 2000). In addition, extant m-coupon research has included
coupon proneness measures as the basis to form multiple groups to use for moderation
assessments on an endogenous attitude construct (Dickinger and Kleijnen 2008), yet has not
modeled the direct effect of coupon proneness on attitude. H7 fills this gap.
H7: Coupon proneness has a direct, positive effect on attitude toward the communication.

19
An additional interaction effect between perceived customization and coupon proneness,
as related to attitude toward the communication, is hypothesized. Consumers have a favorable
response when coupons match preferences, such as a coupon for a preferred brand (Bawa,
Srinivasan, and Srivastava 1997). Coupons possess several characteristics upon which a
consumer’s preference can be matched, such as preferred brand, location, redemption time,
media distribution channel, face value, and the type of deal offered (Neslin and Clarke 1987).
Characteristics such as these create an overall perception of attractiveness for coupons
(Swaminathan and Bawa 2005). In the context of the present study, time is a key factor which
creates more favorable disposition toward coupon use. With traditional paper coupons it takes
time for consumers to search for, clip, save, and locate a coupon when redeeming (Babakus, Tat,
and Cunningham 1988). The added time requires added effort, which forms a non-monetary cost
that adversely affects one’s perceived value of coupon usage (Bawa and Shoemaker 1987), and
ultimately makes a coupon less attractive (Colombo, Bawa, and Srinivasan 2003).
It’s possible that enabling consumers to customize creates a more attractive coupon
offering by better matching a consumer’s preference. In particular, customizing the delivery time
of an m-coupon comparatively minimizes the time needed to search for, clip, save, and locate a
traditional paper coupon for redemption. When compared to a lower degree of perceived
customization, the impact of a higher degree of perceived customization should somewhat
neutralize the effect that a lower degree of general coupon proneness would have on attitude.
Moreover, a lower level of coupon proneness should lead to a less favorable attitude, yet a higher
degree of perceived customization should increase the strength of this coupon proneness 
attitude relationship.

20
H8: A higher degree of perceived customization will produce a stronger positive effect of
coupon proneness on attitude toward the communication compared to a lower degree
of perceived customization.
Method and Design
An online survey collected data from subjects who were randomly assigned to either a
co-production or non-co-production situational condition. The description of the survey given to
subjects was a restaurant opening a new location near them needed feedback on potential
marketing communications. The restaurant used was fictitious to reduce the occurrence of prior
positive / negative attitudes or experiences with a particular firm influencing the results.
Respondents completed the four coupon proneness items prior to the experimental stimuli
exposure, in an effort to assess each participant’s general level of coupon proneness.
All of the information between the two conditions was identical, except for the ability to
co-produce or not co-produce the delivery time aspect of the communication process. Subjects
were told what an m-coupon was and that consumers were required to opt-in to receive
messages. In the co-production condition the subjects were shown a screenshot of a Web form
used to opt-in to receive m-coupons. Subjects also were shown two form fields they could use to
customize their preferred delivery day and time of the m-coupon. In the non-co-production
condition these two fields were not present. All subjects then viewed an image of a mobile
phone with a text message m-coupon on the screen good for 25%-off-a-meal. Subjects were
asked to imagine this m-coupon was sent to their phone. Subjects in the co-production condition
were asked to imagine receiving the m-coupon on their desired day and time. The survey was
then completed.
Sample

21
Consumers responded to an ad posted on a popular U.S. social media site, indicating
respondents were needed to participate in exchange for financial compensation. Investigations of
factor structures, means, standard deviations, and reliabilities of data from consumers who
respond to such online ads show the data do not vary from in-person paper-and-pencil data or in-
person computer-based lab data (Howell et al. 2010). Furthermore, collecting respondents
through social media ads is used currently by researchers (e.g. Dickinger and Kleijnen 2008).
Subjects within the convenience sample were screened based on: owning a mobile phone
and being non-students. Following established practice using online samples (e.g. Landwehr,
McGill, and Herrmann 2011), participants were removed if the survey was completed in less
than half the expected time or had a monotonous answering pattern. This resulted in a final
sample of N = 332 (57% female, M age = 34 years old, M income = $55,000).
Construct Measures
Modified scales from prior research measured the latent constructs. All scales contain
multiple items using seven-point Likert scales with extreme bi-polar anchors (i.e. strongly
disagree/agree). Four items are used to measure perceived customization in order to capture
respondents’ perceptions of the customized delivery time of the communication. Wording and
items were adapted from previous scales (Coulter and Coulter 2002; Steenkamp and Geyskens
2006) to measure perceived customization, with wording changes to reflect the delivery time
aspect and mobile context. The perceived risk construct scale was adapted using four items from
Cox and Cox (2001) to assess the degree of a negative outcome associated with receiving the
communication. The attitude toward the communication scale was adapted using four items from
Holbrook and Batra (1987). This scale assesses attitude toward ads and attitude toward coupons.
The purchase intent scale was adapted using four items from MacKenzie, Lutz, and Belch (1986)

22
to assess the likelihood of purchase by redeeming the m-coupon. The coupon proneness scale
was adapted using four items from Lichtenstein, Netemeyer, and Burton (1990). Table 2 lists
items, standardized factor loadings, t-values, and reliabilities.
-----------------------------------
Insert Table 2 about Here
-----------------------------------
Data Analysis Procedure
Structural equation modeling (SEM) assessed the psychometric properties of the
constructs, model fit, and H2-H8. The inclusion of co-production and non-co-production
situational conditions created experimental-like data. The authors followed the recommendations
proposed by Bagozzi (1977) when using SEM to assess this type of data. This technique uses a
multiple item measure (perceived customization) to represent the dichotomous groups (co-
production or non-co-production) in the data, which produces a larger variance to help assess
relationships in the model while controlling for measurement error. MacKenzie (2001) states
SEM’s advantages make it ideal to use with experimental-like data.
Results
A marker variable assessed whether common method bias inflated or deflated the results
within the cross-sectional data. A theoretically unrelated single-item measure (I like to day
dream) on a seven-point Likert scale anchored by strongly disagree / strongly agree served as the
marker. The two lowest correlations with this marker (r = .002 and r = -.004) are well below the
suggested .20 threshold for problematic method variance (Malhotra, Kim, and Patil 2006).
Nonetheless, a discounted correlation matrix using the more conservative bias estimate (r = -
.004) was compared to the unadjusted matrix per Lindell and Whitney (2001). All correlations

23
remained significant with signs unchanged. In sum, method bias is not a significant risk to the
data.
Two manipulation checks of the co-produced versus non-co-produced subject groups
assessed if the experimental manipulation was successful. First, the four manifest items assessing
perceived customization were summed and averaged to compare the co-producers and the non-
co-producers. The co-producing subjects had a significantly higher level of perceived
customization (M=5.82, SD=1.15, p < .001) compared to the non-co-producing subjects
(M=4.04, SD=1.84). Second, a single item to assess the attractiveness of the coupon (This
coupon is attractive) measured on a seven-point Likert scale anchored by strongly agree /
strongly disagree was included in regard to H8. The co-producing subjects believed the coupon
was significantly more attractive (M=4.35, SD=1.70, p < .05) than the non-co-producers
(M=3.94, SD=1.82). Taken together, the experimental manipulation was deemed to be a success.
The psychometric properties of the constructs were evaluated through confirmatory factor
analysis (CFA). Each item was allowed to load on one factor and could not cross-load on other
factors. In addition, all constructs were tested simultaneously in one model. The results show the
measurement model fit the data well (χ2 = 340.04, df= 160, χ2/df = 2.13; CFI = .97; TLI = .97;
RMSEA = .058; RMSEA 90% CI: .050-.067). The chi-square statistic was significant (p < .001);
however, chi-square is sensitive to larger samples (n > 200; Hu and Bentler 1999). All measures
in the analysis were assessed to be reliable, with construct reliability estimates ranging from .84
to .95. All items loaded strongly and significantly on their respective factors, ranging from .70
to .95. Convergent validity was established with each latent variable’s AVE exceeding .50
(Fornell and Larcker 1981). Discriminant validity was established with the square root of the
AVE for each construct exceeding the correlation between all other constructs (Fornell and

24
Larcker 1981). Table 3 contains each construct’s mean, standard deviation, AVE, reliability
estimate, and the results for the CFA.
-----------------------------------
Insert Table 3 about Here
-----------------------------------
Next, the structural model (see Figure 1) was tested and provided an excellent fit to the
data (χ² =542.51, df= 226, χ²/df= 2.40; CFI = .95; TLI = .95; RMSEA = .065; RMSEA 90% CI:
.058-.072). Perceived customization had a strong, positive relationship with co-production; and
all hypothesized standardized path coefficients were significant and in the expected direction.
The results fully support the direct paths in H2, H3, H5, and H7. Perceived customization had a
negative path coefficient with perceived risk and a positive path coefficient with attitude toward
the communication. Perceived risk had a negative path coefficient with attitude toward the
communication. Coupon proneness had a positive path coefficient with attitude toward the
communication. Table 4 provides results of the structural model testing including standardized
path estimates and R2 estimates.
-------------------------------------
Insert Table 4 about Here
--------------------------------------
The bootstrap procedure recommended by Zhao, Lynch, and Chen (2010) assessed the
indirect effect in H4. This method was chosen over the Sobel test because it is a more powerful
assessment of mediation (Preacher and Hayes 2008). A 2,000-iteration analysis shows the
standardized indirect effect of perceived customization on purchase intent was positive (.42) and
significant (p< .001). Furthermore, the results of the confidence interval show a 95% confidence
level of the true value for the standardized indirect effect lies between .32 and .51. Baron and
Kenny’s (1986) four step mediation test was then used in SEM to determine partial or full
mediation. The direct effect of perceived customization  attitude toward the communication,

25
and attitude toward the communication  purchase intent were each significant. Then the direct
effect of perceived customization  purchase intent was found to be significant in isolation.
Finally, the effect of perceived customization  purchase intent was no longer significant when
all constructs and paths were entered into the model simultaneously, thus illustrating full
mediation. Moreover, H4 is supported as attitude toward the communication fully mediates the
relationship between perceived customization and purchase intent.
The interaction effects in H6 and H8 were assessed using the method suggested by
Mathieu, Tannenbaum, and Salas (1992) and endorsed by Cortina, Chen, and Dunlap (2001) for
an interaction of two multi-item constructs in SEM. The reliability for each of the interaction
terms was estimated using the formula put forth by Bohrnstedt and Marwell (1978). This
technique creates each interaction term and adds each to the model as a separate construct. The
reliabilities and squared correlations from the two independent linear terms are used to calculate
the lambda and variance for each of the added interaction constructs. Results of the model with
and without the included path for each of the interaction constructs were compared separately.
Table 5 presents the fit statistics and the change in χ2 values which assess H6 and H8. The
change in χ2 indicates a significant interaction term for H6 (∆χ2 = 8.2, ∆df = 1, p < .01) and in the
hypothesized direction. Perceived customization interacted with perceived risk (β=.14; p < .01),
as related to attitude. A higher degree of perceived customization in situations with a higher
degree of perceived risk increases attitude significantly, compared to a lesser degree of perceived
customization. Moreover, H6 is supported as perceived customization significantly attenuates the
perceived risk  attitude effect, leading to a higher attitude.
The change in χ2 indicates a marginally significant interaction term for H8 (∆χ2 = 2.7, ∆df
= 1, p < .10) in the predicted direction. Perceived customization and coupon proneness have a

26
marginally significant interaction effect (β=.08; p <.10), as related to attitude. It’s noteworthy to
point out the selected method of assessing this interaction in a structural model is a strict test. A
follow up regression analysis of this hypothesized interaction effect provides more acceptable
support (β=.09, t(3, 328)=2.00, p < .05). However, we choose to label H8 as marginally
significant based on the more rigorous assessment. This interaction suggests a lesser degree of
coupon proneness and a higher degree of perceived customization leads to higher attitude toward
the communication, compared to a lesser degree of coupon proneness and a lesser degree of
perceived customization. Moreover, enabling consumers to co-produce improves attitude toward
the communication for those consumers possessing a lesser degree of coupon proneness.
-------------------------------------
Insert Table 5 about here
-------------------------------------
Discussion
The results of the hypotheses testing from Study 2 illustrate the effect of co-production
when used with direct marketing communication disseminated to personal media. The
conceptual model centers around perceived customization’s effect on the other constructs in the
model. Perceived customization functions as a risk-reducing construct, both directly and
interacting with perceived risk in the risk  attitude relationship. While customization has been
mentioned as a risk reducing technique in the co-production of products, Study 2 expanded this
effect to risk perception regarding marketing communication.
An examination of the results reveal perceived risk is above the mid-point of the scale (M
= 4.27, SD = 1.85) for non-co-producers of the communication process. In contrast, perceived
risk is significantly lower for co-producers (p < .001, M = 2.99, SD = 1.79). Thus, risk
perception is present in personal media communication, yet reduced with co-production.

27
Not all consumers are predisposed to react favorably to certain marketing
communications, such as a coupon discount offer. Coupon proneness assesses this propensity.
The marginally significant interaction in H8 suggests that a co-production strategy may be one
mechanism to transform a less favorable initial reaction to m-coupons to a more favorable
reaction which can improve one’s attitudinal response. A co-produced communication strategy in
this context may engender certain consumers to respond to this type of promotion, who
otherwise would have a limited amount of interest. This finding is financially interesting due to
the assertion that consumers who respond to coupon offers have a proclivity for higher dollar
amount purchases per order and higher purchase frequencies (Hale 2010).
General Discussion and Conclusions
Summary and Implications
The overarching theoretical contribution of this paper is the conceptual shift concerning
marketing communication sent to personal media as an attention-getting promotion disconnected
from direct consumer participation to a service-like, participatory offering. The agent of change
driving this shift is evolving technology, which requires marketers to enable consumers to
participate in interactive decision making in the communication process. This conceptual shift
aligns well with various service orientations, such as service logic, service-dominant logic, and
unified services theory, each of which identify co-production as an element within the provision
of service. Some marketers will quickly dismiss this alternative notion by arguing marketing
communication and service have never been closely associated. However, improvements to a
long standing system require a new or alternative paradigm (Kuhn 1996). An important

28
theoretical implication is that co-production is not limited to service outcomes. Consumers can
effectively co-produce marketing communication to produce mutually beneficial exchange.
Related to this overarching contribution is an additional theoretical implication which
improves communication theory by expanding the understanding of push versus pull marketing
communication. Acknowledging co-production as a relevant personal media communication
strategy improves upon the mass media communication process. Mass media derives its model
from mass communication theory, which is the process currently applied by marketers in
personal media, albeit with limited success. An improvement to marketing communication
theory acknowledges consumer involvement in co-producing the communication process, as this
will usher in an improved practice appropriate for personal media. In this sense, co-production
centers around push versus pull marketing communication and who has input into a message:
firms (push) or consumers (pull). Heavier reliance on pull is one implication of a service
orientation applied to promotion. Panel A in the top portion in Figure 2 shows that the current
opt-in strategy for personal media enables consumers to pull communications by agreeing to be
an audience. However, after this initial pull action, firms reassume a traditional push mentality.
Consumers are enabled to pull for a moment, followed by firm-controlled push messages.
---------------------------------------------
Insert Figure 2 about here
---------------------------------------------
The current paper suggests the firm-consumer push-pull encounter must involve an
additional, more participatory pull-like process given to consumers. Panel B in Figure 2 accounts
for this by expanding the push-pull mindset. As depicted, the push-pull continuum widens when
introducing pull with co-production. The expanded continuum grants more participation to
consumers aside from the initial opt-in decision. To truly enable communication to be more
effective and valuable to consumers and firms, a deeper level of participation must be offered to

29
consumers. The degree of participation may vary, but it must be more than opting to become a
passive target to communications. After the decision to opt-in, consumers are actively involved
in pull-like decisions. Consistent with this logic, the previously labeled pull is more accurately
characterized as pull with firm production, as consumers initially decide to opt-in with a pull-like
action, but then the firm controls and produces standardized communication. After enabling
consumers to co-produce, the stages within the communication process then play out. The right
side of Figure 2 is a new concept to marketing communication. Consumers are now enabled to
provide not only passive, but also active feedback with co-production. Providing active feedback
in this manner gives firms the opportunity to avoid duplicating several decades of ineffective
marketing communication (Rust and Oliver 1994).
This implication of co-produced communication is a major shift away from strategies and
concepts within traditional marketing communication (Duncan and Moriarty 1998; Godfrey,
Seiders, and Voss 2011) and personal media marketing communication (Deighton and Kornfeld
2009; Shankar et al. 2010; Sultan, Rohm, and Gao 2009). Whether a firm uses traditional or
personal media, in either case attempts are made by a company to infer consumer preferences
and/or exploit existing information about particular consumers. The inference and/or exploitation
arises from information available to a firm through primary (i.e. individual customer records) or
secondary (i.e. mining online databases and sources) information. However, the information is
not specifically provided from a consumer to a firm for a particular communication exchange.
The information may be accurate for a consumer, but it may not be context specific for a
particular communication exchange. This lack of context is a gap that is now filled with a co-
production communication strategy.

30
Co-production is a dramatic move to escape the dependence of marketers on the need to
infer or exploit consumer information; and a major addition to personal media frameworks.
Whereas Deighton and Kornfeld’s (2009) interactivity paradigms include thought tracing and
activity tracing in an attempt to gather information second-hand, a co-production strategy offers
customer-specific information to a firm first-hand. In addition, the interactivity paradigms of
property, social, and cultural exchanges have roots in marketers wanting to have a pedestal to
persuade or manipulate consumers either apparently or covertly. Co-production creates a
communication exchange based on transparency, relevance, and usefulness while providing high
degrees of identity and meaning. The identity of the consumer is known along with preferences;
and the meaning of the brand to each consumer becomes clearer with a higher degree of
contextual meaning. Personal media enable consumers to talk back to firms with active feedback,
meaning the era of the communication exchange in addition to property exchange has begun.
Co-production is a form of consumer empowerment, which is in line with the shifting
balance of media power toward consumers and away from firms. Co-production of marketing
communication disrupts this power balance by allowing for more assertiveness on the part of the
consumer. This control of the interaction by the consumer goes beyond the five digital paradigms
discussed by Deighton and Kornfield (2009) to one that is a communication exchange. An
understanding of the conditions under which the communication exchange takes place can give a
better understanding of when and how consumers are more willing to exert their power. Firms
need to understand the barriers they need to overcome in order to invite the communication
exchange. Both firms and consumers should have an understanding of the outcomes each party
is willing to entertain for the communication exchange to be a success.

31
The evolution toward co-produced communication brings us to an interesting question:
which characteristics of personal media communication are viable alternatives to enable
consumers to co-produce? Under the premise of Lancaster’s (1966) characteristics theory, the
value of personal media marketing communication is not the marketing message itself, but rather
the function of the attributes of the communication. This rationale is consistent with
attractiveness theory used in traditional promotions research (Bawa, Srinivasan, and Srivastava
1997; Swaminathan and Bawa 2005). Similar to how products contain many valuable
characteristics that create utility for a consumer, these newer forms of marketing communication
also contain valuable characteristics. One characteristic of marketing communication (delivery
time) is shown to produce favorable outcomes for firms and consumers across both studies, yet
additional characteristics are viable options for firms to enable consumers to co-produce.
Figure 3 presents a suggested personal media co-production framework divided into three
stages: pre-communication, communication, and post-communication. The top portion of Figure
3 presents a number of characteristics present within personal media marketing communication
ideally suited to co-produce. It is important to note that Figure 3 does not make any normative
predictions of which characteristics should be applied in certain marketing contexts. Providing
this type of non-normative framework supplies a broad picture of co-production possibilities and
will aid in the future work of assessing co-produced personal media communication.
In the pre-communication stage marketers make available certain communication
characteristics ideally suited for co-production in a given context. Among these characteristics
there is delivery time (i.e. day of the week or time of day) when a consumer prefers to receive a
communication, frequency of messages sent to a consumer, a consumer’s preferred location,
message format (such as plain text, an image, or a video embedded within a communication),

32
specific type of product a consumer prefers to receive messages about (i.e. subject of a
communication), enabling a consumer to choose who is a recipient of a message (i.e. a consumer
recommending a brand’s message to a friend), type of communication preferred (such as a
discount message or an informational message), and the economic incentive1. Certain consumers
and brands may find varying levels of value among the different characteristics, but this list
provides a basis for a beginning to consider what consumers can co-produce in communication.
---------------------------------------------
Insert Figure 3 about here
---------------------------------------------
The middle portion of Figure 3 is the communication stage, which represents a consumer
receiving a marketing communication that he or she co-produced. The bottom portion of Figure
3 is the post-communication stage. This stage represents post-communication outcomes for
consumers and firms, several of which were assessed across both studies in the current paper.
Enabling consumers to co-produce communication characteristics may lead to an increase in
consumer purchase intent, an increase in consumer attitude toward the communication, an
increase in consumer response rate (i.e. purchases), an increase in consumer word-of-mouth
(WOM) recommendations, a decrease in consumer perception of risk, a decrease in consumer
opt-out rate, and an increase in consumer satisfaction. Future research may be able to build upon
this foundation to identify additional co-production options and positive outcomes for firms and
consumers. One particularly interesting area of future research would be to assess the weighted
importance to consumers of the various characteristics open to co-production and how these
characteristics affect different outcomes.
1 Internet sites such as priceline.com enable consumers to specify a price for a product. While this is a viable
business model for core product offerings, the idea may also be useful for discount amounts made available to
consumers via marketing communication.

33
Another theoretical implication emerges by juxtaposing the goods-services contrast to the
mass-personal media contrast. The goods-dominant logic is a derivative of mass manufacturing
(Vargo and Lusch 2004), upon which a guiding philosophy for mass media communication
emerged. This philosophy was that firms would operate as a sole decision maker within mass
media, similar to the efficiencies achieved by excluding consumer participation in standardized,
mass manufactured goods (Vargo and Akaka 2009). The position of this research is that service
logic, service-dominant logic, and unified services theory are the prevailing ideologies for
marketing communication sent through personal media. Moreover, the production process for
goods versus services is similar to the production process for mass media versus personal media
communication, respectively. A key differentiating factor is co-produced communication.
Typically in mass media marketing communication a firm investigates many of the
elements contained in Figure 3. A firm infers the best options to drive a specific outcome for
itself in an effort to produce the largest marginal impact on consumer purchasing or persuasion.
Test marketing may occur, often referred to as A/B testing, in small groups of consumers, yet the
marketer is still the chief decision maker largely basing judgments on inference without direct
consumer participation in the decision making process. Passive consumers provide general
information upon which marketers make decisions. The introduction of co-production into
communication enables a marketer to forgo A/B testing by enabling consumers to actively
choose A or B. This is a world of change that marketing, in a communication context, has not
experienced before. Figure 2’s two types of feedback align with this change. Traditionally in
mass media firms push out messages and then assess the results with passive feedback. Now in
personal media firms can involve consumers by granting co-production opportunities. This is
active feedback, in that, consumers are no longer a passive audience, but now become active

34
participants in the communication process. The emergence of co-production and active feedback
introduces a potentially rich area for behavioral and modeling researchers, such as including co-
production in the comparison of A/B testing (A or B or co-production). Future research can
examine if the benefits of co-producing some of Figure 3’s elements will trade off against a
firm’s persuasion or behavioral goals.
Interestingly, mass media marketing communication may also benefit from a co-
production strategy. Years ago this was impossible, but now the evolution of technology and
media makes this a reality. The anticipated growth of interactive television requires new thinking
into how mass promotional communication through this medium will include consumer
participation. In addition, mobile television and multiscreen marketing are estimated to have
exponential growth in the near future, which will morph a mass medium with personal media
(television, social media, and mobile devices). Marketers must avoid old ways of thinking by
limiting consumer participation with newer media to opting-in to become a passive audience.
Three noteworthy managerial implications demand attention. First, although not formally
hypothesized, as a point of comparison in Study 1 the purchase redemption of a paper coupon
with the exact same verbiage and face value was compared to the m-coupon campaign for one
restaurant. Paper coupons were manually distributed to people in the surrounding geographic
area. Consumers who wanted the coupon were given a form requiring them to check a box and
answer two brief questions (to account for minimal effort required by consumers who opt-in to
receive an m-coupon). Purchase redemption was 3% for the paper coupon campaign, which is
typical for paper coupons in the U.S. (historically 1%-to-3%). These results suggest an
opportunity for marketers to widen the purchase redemption gap between paper and m-coupons.

35
Second, a co-production strategy in this context provides guidance to firms that are
struggling to use sales promotions with personal media. An example is the economic losses by
firms using social couponing through Groupon and similar services (Kumar and Rajan 2012).
Firms recognize promotions are needed through new media to spur customer activity; however
social coupons appear to be a losing proposition due to the massive discounts (50%+) and fees.
In contrast, the current paper’s Study 1 illustrates how co-production is successfully used with a
personal media sales promotion at a more acceptable discount rate.
Third, co-production may lead to an increase in consumer opt-in rates. Opt-in must occur
before exposure to messages, and identifying ways to increase opt-in rates is important. If firms
publicize consumers have more decisional input into the creation of marketing communication,
this may cause some consumers to opt-in who otherwise would not participate. Risk perception
is likely to be present prior to opting-in. Communicating a co-production strategy with opt-in
signage and verbiage may attenuate these risks.
Limitations and Future Research
One limitation is delivery time was the only co-produced attribute of the communication
process examined in both studies. It’s possible to co-produce other characteristics as depicted in
Figure 3’s framework. Attractiveness of a promotional communication varies on a number of
different characteristics of a communication, suggesting that future research can examine
additional characteristics and perhaps compare the weighted importance. In particular, it would
be interesting to compare co-produced discount offers with different face value amounts.
Study 1’s restaurants had a high response rate, in part, due to subjects’ co-producing a
characteristic aligning well with the product offering. Despite the results, some firms or products

36
would not benefit from customizing the timing of a communication. For example, consumers
may know when they prefer to eat at a restaurant more so than when they prefer to visit a
hardware store. However, the results from Study 1 point to the idea of what such a strategy offers
to firms if co-producing marketing communication aligns well with a product offering.
Only a single consumer-firm co-production opportunity takes place in each study, where
consumers co-produce only before receiving a single communication. Research should examine
a lengthier dialogical orientation (Ballantyne and Varey 2006) where multiple communication
exchanges with co-production occur. Godfrey, Seiders, and Voss (2011) suggest that consumers
react negatively when the amount of communication is not at an ideal level. Future research
should examine possible outcomes of co-production with varying levels of message volume and
co-production opportunities. What is the ideal level of the number of co-production
opportunities, and is there a balance between co-production opportunities, and specific
opportunities that cannot be co-produced by the consumer?
The present investigation did not seek to identify if the increase in attitude was not only
due to customization, but also consumer empowerment or the need for uniqueness. Not much is
known about the psychological antecedents and outcomes of perceived customization. It is most
likely a multilevel variable and this suggested area of future research would be beneficial.
Both Study 1 and 2 are limited to a convenience sample in a single product category,
meaning findings may not generalize to all consumers across various products. Examination of
multiple product categories, such as high versus low involvement and hedonic versus utilitarian
can also be addressed by future research. Also, demographic information was not captured from
consumers in Study 1. Future research should investigate how actual purchase behavior varies
across demographics when using co-produced communication and personal media.

37
Mobile technologies are changing the retail landscape. The interactivity between
consumers, as well as consumers and retailers, may alter strategies that retailers have always
considered. As Shankar et al. (2010) discuss how retailers need to anticipate the changes brought
on by mobile technologies, retailers need to also understand consumer perceptions of the use of
different handheld devices. The only form of personal media under investigation is mobile
phone text messaging. Co-producing social media direct marketing communication, as well as
the use of other personal devices, such as tablet computers, are ripe areas for future research.
Lastly, mobile marketing acceptance has been studied cross-culturally, with acceptance in
both mature and emerging markets (Sultan, Rohm, and Gao 2009). However, co-production
models may not be as impactful in other markets, as some cultures are irritated by mobile
communications (Liu et al. 2012). An examination of cultural effects of co-production may lead
to a better understanding of the new customer and retailer dynamics.
References
Ajzen, Icek (1991), “The Theory of Planned Behavior,” Organizational Behavior and Human
Decision Processes, 50, 2, 179-211.
Auh, Seigyoung, Simon J. Bell, Colin S. McLeod, and Eric Shih (2007), “Co-production and
Customer Loyalty in Financial Services,” Journal of Retailing, 83, 3, 359-370.
Babakus, Emin, Peter Tat, and William Cunningham (1988), “Coupon Redemption: A
Motivational Perspective,” Journal of Consumer Marketing, 5, 2, 37-43.
Bacile, Todd J. and Ronald E. Goldsmith (2011), “A Services Perspective for Text Message
Coupon Customization,” Journal of Research in Interactive Marketing, 5, 4, 244-257.
Bagozzi, Richard P. (1977), “Structural Equation Models in Experimental Research,” Journal of
Marketing Research, 14, 2, 209-226.
Ballantyne, David and Richard J. Varey (2006), “Introducing a Dialogical Orientation to the
Service-Dominant Logic of Marketing” in The Service-Dominant Logic of Marketing:
Dialog, Debate, and Directions, Robert F. Lusch and Stephen L. Vargo, eds. New York:
M. E. Sharpe, 224-235.

38
Baron, Reuben M. and David A. Kenny (1986), “The Moderator-Mediator Variable Distinction
in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations,”
Journal of Personality and Social Psychology, 51, 6, 1173-1182.
Barwise, Patrick and Colin Strong (2002), “Permission-Based Mobile Advertising,” Journal of
Interactive Marketing, 16 (1), 14-24.
Bateson, J. E. G. (1985), “Self-Service Consumer: An Exploratory Study, “ Journal of Retailing,
61, 3, 49-76.
Bawa, Kapil and Robert W. Shoemaker (1987), "The Effects of a Direct Mail Coupon on Brand
Choice Behavior," Journal of Marketing Research, 24, 4, 370-376.
---, Srini S. Srinivasan and Rajendra K. Srivastava (1997), "Coupon Attractiveness and Coupon
Proneness: A Framework for Modeling Coupon Redemption," Journal of Marketing
Research, 34, 4, 517-525.
Bendapudi, Neeli and Robert P. Leone (2003), "Psychological Implications of Customer
Participation in Co-Production," Journal of Marketing, 67, 1, 14-28.
Bohrnstedt, George W. and Gerald Marwell (1978), “The Reliability of Products of Two
Random Variables,” in Sociological Methodology, K. F. Schuessler, ed. San Francisco,
CA: Jossey-Bass, 254-273.
Brehm, Jack W. (1972), Responses to Loss of Freedom: A Theory of Psychological Reactance.
Morristown, NJ: General Learning Press.
Clarke, Irvine (2001), “Emerging Value Propositions for M-Commerce,” Journal of Business
Strategies, 18, 2, 133-148.
Colombo, Richard, Kapil Bawa, and Srini S. Srinivasan (2003), "Examining the Dimensionality
of Coupon Proneness: A Random Coefficients Approach," Journal of Retailing and
Consumer Services, 10, 1, 27-33.
Cortina, Jose, Gilad Chen, and William P. Dunlap (2001), “Testing Interaction Effects in
LISREL: Examination and Illustration of Available Procedures,” Organizational
Research Methods, 4, 4, 324-360.
Coulter, Keith S. and Robin A. Coulter (2002), “Determinants of Trust in a Service Provider:
The Moderating Role of Length of Relationship,” Journal of Services Marketing, 16, 1,
35-50.
Cox, Dena and Anthony Cox (2001), “Communicating the Consequences of Early Detection:
The Role of Evidence and Framing.” Journal of Marketing, 65, 3, 91-103.
Deighton, John and Leora Kornfeld (2009), “Interactivity’s Unanticipated Consequences from
Marketers and Marketing,” Journal of Interactive Marketing, 23, 1, 4-10.
Dickinger, Astrid and Mirella Kleijnen (2008), “Coupons Going Wireless: Determinants of
Consumer Intentions to Redeem Mobile Coupons,” Journal of Interactive Marketing, 22,
3, 23-39.

39
Dowling, Grahame R. and Richard Staelin (1994), "A Model of Perceived Risk and Intended
Risk-Handling Activity," Journal of Consumer Research, 21, 1, 119-134.
Duncan, Tom, and Sandra E. Moriarty (1998), "A Communication-Based Marketing Model for
Managing Relationships," Journal of Marketing, 62 (2), 1-13.
--- and --- (2006), “How Integrated Marketing Communication’s “Touchpoints” Can
Operationalize the Service-Dominant Logic” in The Service-Dominant Logic of
Marketing: Dialog, Debate, and Directions, Robert F. Lusch and Stephen L. Vargo, eds.
New York: M. E. Sharpe, 236-244.
eMarketer (2010), “Mobile Users Ready for Location-Based Text Marketing,” (accessed
January 20, 2011), [available at http://www.emarketer.com/Article.aspx?R=1007782].
Etgar, Michael (2008), “A Descriptive Model of the Consumer Co-Production Process,” Journal
of the Academy of Marketing Science, 36, 1, 97-108.
Firat, A. Fuat and Nikhilesh Dholakia (2006), “Theoretical and Philosophical Implications of
Postmodern Debates: Some Challenges to Modern Marketing,” Marketing Theory, 6, 2,
123-162.
Fornell, Claes and David F. Larcker (1981), "Evaluating Structural Equation Models with
Unobservable Variables and Measurement Error,” Journal of Marketing Research, 18, 1,
39-50.
Fortin, David R. (2000), "Clipping Coupons in Cyberspace: A Proposed Model of Behavior for
Deal-Prone Consumers," Psychology & Marketing, 17, 6, 515-534.
Godek, John, J. Frank Yates, and Yeosun Yoon (2002), “Customization and Personalization:
The Influence of Perceived Control and Perceived Capability on Product Evaluations,”
Advances in Consumer Research, 29, 1, 157.
Godfrey, Andrea, Kathleen Seiders, and Glenn B. Voss (2011), "Enough Is Enough! The Fine
Line in Executing Multichannel Relational Communication," Journal of Marketing, 75
(4), 94-109
Graham, Phil (2007), "Political Economy of Communication: A Critique," Critical Perspectives
on International Business, 3, 3, 226-245.
Grönroos, Christian (2006), "Adopting a Service Logic for Marketing," Marketing Theory, 6, 3,
317-333.
Hale, Todd (2010), “The Coupon Comeback,” (accessed June 10, 2012), [available at
http://blog.nielsen.com/nielsenwire/consumer/the-coupon-comeback/].
Hennig-Thurau, Thorsten, Edward C. Malthouse, Christian Friege, Sonja Gensler, Lara
Lobschat, Arvind Rangaswamy, and Bernd Skiera (2010), “The Impact of New Media on
Consumer Relationships,” Journal of Service Research, 13, 3, 311-330.
Holbrook, Morris and Rajeev Batra (1987), “Assessing the Role of Emotions as Mediators of
Consumer Responses to Advertising,” Journal of Consumer Research, 14, 3, 404-420.
Howell, Ryan T., Katrina S. Rodzon, Mark Kurai, and Amy H. Sanchez (2010), "A Validation of
Well-Being and Happiness Surveys for Administration via the Internet," Behavior
Research Methods, 42, 3, 775-784.

40
Hu, Li-tze and Peter M. Bentler (1999), “Cutoff Criteria for Fit Indexes in Covariance Structure
Analysis: Conventional Criteria Versus New Alternatives,” Structural Equation
Modeling, 6, 1, 1-55.
Huffman, Cynthia and Barbara E. Kahn (1998), “Variety for Sale: Mass Customization or Mass
Confusion?,” Journal of Retailing, 74, 4, 491-513.
Innis, Harold A. (1942), "The Newspaper in Economic Development," Journal of Economic
History, 2, 1-33.
Kahneman, Daniel and Amos Tversky (1979), "Prospect Theory: An Analysis of Decision under
Risk,” Econometrica, 47, 2, 263-292.
Kuhn, Thomas S. (1996), The Structure of Scientific Revolutions (Third Edition). Chicago, IL:
University of Chicago Press.
Kumar, V. and Bharath Rajan (2012), “Social Coupons as a Marketing Strategy: A Multifaceted
Perspective,” Journal of the Academy of Marketing Science, 40, 1, 120-138.
Lancaster, Kelvin J. (1966), "A New Approach to Consumer Theory," Journal of Political
Economy, 74, 2, 132-157.
Landwehr, Jan R., Ann L. McGill, and Andreas Herrmann (2011), "It’s Got the Look: The Effect
of Friendly and Aggressive 'Facial' Expressions on Product Liking and Sales,” Journal of
Marketing, 75, 3, 132-146.
Lee, Jungki and Arthur Allaway (2002), “Effects of Personal Control on Adoption of Self-
Service Technology Innovations,” Journal of Services Marketing, 16, 6, 553-572.
Lichtenstein, Donald R., Richard G. Netemeyer, and Scot Burton (1990), "Distinguishing
Coupon Proneness from Value Consciousness – An Acquisition-Transaction Utility-
Theory Perspective," Journal of Marketing, 54, 3, 54-67.
Lindell, Michael K. and David J. Whitney (2001), "Accounting for Common Method Variance in
Cross-Sectional Research Designs," Journal of Applied Psychology, 86, 1, 114-121.
Liu, Chia-Ling 'Eunice', Rudolf R. Sinkovics, Noemi Pezderka, and Parissa Haghirian (2012),
"Determinants of Consumer Perceptions toward Mobile Advertising - a Comparison
between Japan and Austria," Journal of Interactive Marketing, 26 (1), 21-32.
Lüders, Marika (2008), "Conceptualizing Personal Media," New Media and Society, 10, 5, 683-
702.
Lutz, Richard J. (1985), "Affective and Cognitive Antecedents of Attitude Toward the Ad: A
Conceptual Framework" in Psychological Processes and Advertising Effects: Theory,
Research and Application, L. F. Alwitt and A. A. Mitchell, eds. Hillsdale, NJ: Lawrence
Erlbaum Associates, 45-63.
MacKenzie, Scott B., Richard J. Lutz, and George E. Belch (1986), “The Role of Attitude
Toward the Ad as a Mediator of Advertising Effectiveness: A Test of Competing
Explanations,” Journal of Marketing Research, 23, 2, 130-143.
--- (2001) , "Opportunities for Improving Consumer Research through Latent Variable Structural
Equation Modeling," Journal of Consumer Research, 28, 1, 159-166.

41
Malhotra, Naresh K., Sung S. Kim, and Ashutosh Patil (2006), “Common Method Variance in IS
Research: A Comparison of Alternative Approaches and a Reanalysis of Past Research,”
Management Science, 52, 12, 1865-1883.
Mathieu, John E., Scott I. Tannenbaum, and Eduardo Salas (1992), “Influences of Individual and
Situational Characteristics on Measures of Training Effectiveness,” Academy of
Management Journal, 35, 4, 828-847.
Mittal, Banwari (1994), "An Integrated Framework for Relating Diverse Consumer
Characteristics to Supermarket Coupon Redemption," Journal of Marketing Research,
31, 4, 533-544.
Muehling, Darrel D. and Michelle McCann (1993), “Attitude Toward the Ad: A Review,”
Journal of Current Issues and Research in Advertising, 15, 2, 25–58.
Neslin, Scott A. and Darral G. Clarke (1987), "Relating the Brand Use Profile of Coupon
Redeemers to Brand and Coupon Characteristics," Journal of Advertising Research, 27,
1, 23-32.
Pavlou, Paul A. (2003), “Consumer Acceptance of Electronic Commerce: Integrating Trust and
Risk with the Technology Acceptance Model,” International Journal of Electronic
Commerce, 7, 3, 69-103.
Payne, Adrian, Kaj Storbacka, and Pennie Frow (2008), “Managing the Co-creation of Value,”
Journal of the Academy of Marketing Science, 36, 1, 83-96.
Prahalad, C. K. and Venkat Ramaswamy (2004), The Future of Competition: Co-creating
Unique Value with Customers. Boston, Mass: Harvard Business School Press.
Preacher, Kristopher J. and Andrew F. Hayes (2008), "Asymptotic and Resampling Strategies for
Assessing and Comparing Indirect Effects in Multiple Mediator Models," Behavior
Research Methods, 40, 3, 879-891.
Rust, Roland T., and Richard W. Oliver (1994), “The Death of Advertising,” Journal of
Advertising, 23, 4, 71-77.
Sampson, Scott E. and Craig M. Froehle (2006), “Foundations and Implications of a Proposed
Unified Services Theory,” Production and Operations Management, 15, 2, 329-343.
Shankar, Venkatesh and Sridhar Balasubramanian (2009), “Mobile Marketing: A Synthesis and
Prognosis,” Journal of Interactive Marketing, 23, 2, 118-129.
---, Alladi Venkatesh, Charles F. Hofacker, and Prasad Naik (2010), "Mobile
Marketing in the Retailing Environment: Current Insights and Future Research Avenues,"
Journal of Interactive Marketing, 24 (2), 111-120.
Shannon, Claude E. and Warren Weaver (1949), The Mathematical Theory of Communication,
Urbana, IL: University of Illinois Press.
Sheth, Jagdish N. and Rajendra S. Sisodia (2006), Does Marketing Need Reform? Fresh
Perspectives on the Future. Armonk, NY: M.E. Sharp.
Shimp, Terence A. and Alican Kavas (1984), "The Theory of Reasoned Action Applied to
Coupon Usage," Journal of Consumer Research, 11, 3, 795-809.

42
Steenkamp, Jan-Benedict E.M. and Inge Geyskens (2006), "How Country Characteristics Affect
the Perceived Value of Web Sites," Journal of Marketing, 70, 3, 136-150.
Sultan, Fareena, Andrew J. Rohm, and Tao Gao (2009), “Factors Influencing Consumer
Acceptance of Mobile Marketing: A Two-Country Study of Youth Markets,” Journal of
Interactive Marketing, 23, 4, 308-320.
Swaminathan, Srinivasan and Kapil Bawa (2005), "Category-specific Coupon Proneness: The
Impact of Individual Characteristics and Category-specific Variables," Journal of
Retailing, 81, 3, 205–214.
Swilley, Esther (2010), “Technology Rejection: The Case of the Wallet Phone,” Journal of
Consumer Marketing, 27, 4, 304-312.
Taylor, James W. (1974), "The Role of Risk in Consumer Behavior," Journal of Marketing, 38,
2, 54-60.
VanBoskirk, Shar (2011), “US Interactive Marketing Forecast, 2011-2016,” (accessed February
21, 2012), [available at http://www.forrester.com/rb/Research/
us_interactive_marketing_forecast%2C_2011_to_2016/q/id/59379/t/2].
Vargo, Stephen L. and Robert F. Lusch (2004), “Evolving to a New Dominant Logic for
Marketing,” Journal of Marketing, 68, 1, 1-17.
--- and --- (2008), “Service-Dominant Logic: Continuing the Evolution,” Journal of the Academy
of Marketing Science, 36, 1, 1-10.
--- and Melissa Archpru Akaka (2009), "Service-Dominant Logic as a Foundation for Service
Science: Clarifications," Service Science, 1, 1, 32-41.
Wind, Jerry and Arvind Rangaswamy (2001), "Customerization: The Next Revolution in Mass
Customization," Journal of Interactive Marketing, 15, 1, 13-32.
Wortman, Camille B. and Jack W. Brehm (1975), "Responses to Uncontrollable Outcomes: An
Integration of Reactance Theory and the Learned Helplessness Model" in Advances in
Experimental Social Psychology (Vol. 8), L. Berkowitz ed. New York: Academic Press,
277-336.
Zaltman, Gerald and Melanie Wallendorf (1983), Consumer Behavior, New York: Wiley.
Zhao, Xinshu, John G. Lynch Jr., and Qimei Chen (2010), "Reconsidering Baron and Kenny:
Myths and Truths about Mediation Analysis," Journal of Consumer Research, 37, 2, 197-
206.

Table 1
Results from Field Experiments in Study 1
-----------------------------------------------------------------------------------------------------------------------------------------
#Co-Produced #Co-Produced Co-Produced #Non-Co-Produced #Non-Co-Produced Non-Co-Produce
M-Coupons M-Coupons Redemption M-Coupons M-Coupons Redemption
City Restaurant Sent Redeemed Rate Sent Redeemed Rate
-----------------------------------------------------------------------------------------------------------------------------------------
A 1 68 17 25.0%a 57 6 10.5%
A 2 112 42 37.5%b 132 11 8.3%
A 2 121 37 30.6%a 140 13 9.3%
A 2 140 31 22.1%b 151 17 11.3%
A 2 145 38 26.2%b 153 14 9.2%
A 3 81 22 27.2%b 86 5 5.8%
A 3 112 24 21.4%b 106 8 7.5%
A 3 125 23 18.4%a 118 11 9.3%
A 3 125 20 16.0% 128 12 9.4%
B 4 120 22 18.3% 121 14 11.6%
B 4 126 28 22.2%b 120 9 7.5%
-----------------------------------------------------------------------------------------------------------------------------------------
TOTALS: 1,275 304 23.8%c 1,312 120 9.1%
-----------------------------------------------------------------------------------------------------------------------------------------
Co-produced m-coupon redemption rate is significantly higher than non-co-produced m-coupon: a p < .05 b p <
d free drink with purchase of a meal is approximately an overall discount of 20% off
e BOGO 50%: buy one meal, get one meal 50% off is approximately an overall discount of 25% off
f the overall average discount offered across all campaigns was approximately 25% off

Table 2
Scales and Items Used in Study 2
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
Standardized Factor
Construct Items Loading
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
Perceived customization - The coupon’s delivery time features are customized. .88
(anchored with - The delivery time features of this coupon are adaptable to my needs. .89
strongly disagree/strongly agree) - Some features of this coupon can be changed by me. .86
- The day and time when the coupon is sent is customizable to meet my schedule. .91
Perceived risk - Getting this coupon is risky. .90
(anchored with - This coupon can lead to bad results. .95
strongly disagree/strongly agree) - This coupon can lead to uncertain outcomes. .92
- Getting this coupon would cause me to worry. .82
Attitude toward the - I definitely dislike / I definitely like the coupon. .90
communication - I definitely react unfavorably / I definitely react favorably to the coupon. .94
- I definitely feel negative / I definitely feel positive toward the coupon. .92
- The coupon is definitely bad / The coupon is definitely good. .88
Purchase intent Indicate the probability that you will try this coupon for the restaurant when it
becomes available in your area:
- Definitely Unlikely / Definitely likely .93
- Definitely Improbable / Definitely probable .93
- Definitely Impossible / Definitely possible .86
- No, definitely not / Yes, definitely .92
Coupon proneness - Redeeming coupons makes me feel good. .71
(anchored with - I am more likely to buy brands for which I have a coupon. .83
strongly disagree/strongly agree) - When I use coupons, I feel that I am getting a good deal. .78
- I enjoy using coupons, regardless of the amount of money I save by doing so. .70
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
NOTE: N = 332
* Denotes a path constrained to 1 for model identification

Table 3
Means, Standard Deviations, Correlations, and Reliability Estimates for Study 2
-----------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------------
CFA results: χ2 = 340.04, df = 160, χ2/df = 2.13; CFI = .97; TLI = .97; RMSEA = .058; RMSEA 90% CI: .050-.0
N = 332; all constructs measured on seven-point scales; SD = Standard Deviation; AVE = Average Variance Ext
Construct Reliability.
The square roots of the average variance extracted for each construct are presented in bold on the diagonal of the
* p < .01
Constructs Range Mean SD AVE CR 1 2 3 4
------------------------------------------------------------------------------------------------------------------------------------------
1. Perceived customization 1 - 7 4.90 1.78 .78 .94 .88
2. Perceived risk 1 - 7 3.66 1.93 .81 .94 -.37* .90
3. Attitude toward comm. 1 - 7 5.10 1.76 .83 .95 .53* -.42* .91
4. Purchase intent 1 - 7 5.34 1.63 .83 .95 .49* -.26* .79* .91
5. Coupon proneness 1 - 7 5.28 1.34 .58 .84 .27* -.03 .31* .41*

Table 4
Structural Model Standardized Paths and Variance Explained Results for Study 2
-----------------------------------------------------------------------------------------------------------------------------------------
Standardized
Hypothesized and Non-Hypothesized Paths Path Coefficients R2 (construc
-----------------------------------------------------------------------------------------------------------------------------------------
---- Categorical co-production conditions → Perceived customization .50* .25 (Perceive
H2: Perceived customization → Perceived risk -.38* .14 (Perceive
H3: Perceived customization → Attitude communication .39* .35 (Attitude)
H4: Perceived customization fully mediated through Attitude → Purchase intent .42* .62 (Purchase
H5: Perceived risk → Attitude communication -.27*
H6: Perceived customization X Perceived risk → Attitude communication .14**
H7: Coupon proneness → Attitude communication .22*
H8: Perceived customization X Coupon proneness → Attitude communication .08***
-----------------------------------------------------------------------------------------------------------------------------------------
Structural Model Fit: χ² = 542.51, df= 226, χ²/df = 2.40; CFI = .95; TLI = .95; RMSEA = .065; RMSEA 90% CI:
* p < .001
** p < .01
*** p < .10

Table 5
Structural Model Fit and Interaction Assessments
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Fit / Path Research model without
interaction constructs
Model with Perceived
customization X Perceived
risk interaction construct
Model with Perceived
customization X coupon
proneness interaction construct
χ² | df | χ²/df 542.51 | 226 | 2.40 534.30 | 225 | 2.38 539.86 | 225 | 2.40
CFI .95 .95 .95
TLI .95 .95 .95
RMSEA .065 .064 .065
RMSEA 90% CI .058 to .072 .057 to .072 .058 to .072

Figure 1
Conceptual Model
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
Note: - Dashed path in Study 2’s structural model is supported by previous research. No formal hypothesis is test
- H4: the effect of perceived customization on purchase intent is fully mediated through attitude toward the
H7 +
H5 ‐
H6 +
H4 +H3 +
H2 ‐
Attitude toward the
communication
Purchase
intent
Co-production
situational
condition
Perceived risk
Perceived
customization
Coupon proneness
H8 +

49
Figure 2
Conceptual Push-Pull Framework Featuring Co-Produced Marketing Communication
Firm assumes control Consumers Merely Opting-in Co-Producing Communication
(Mass media) (Personal media currently) (Personal media in the future)
A: Current conceptualization of consumers participating in personal media marketing communications:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
B: New conceptualization of consumers co-producing the communication process within personal media:
Lesser degree of Limited Higher degree of
consumer participation consumer participation consumer participation
Push Pull with firm production Pull with co-production
Push Pull
Decoding
Source
(Marketer)
Encoding
Medium
(Personal Media)
Receiver
(Consumer)
Message
(Marketing Communication)Passive
Feedback
Active
Feedback
Feedback

Figure 3
Personal media co-production framework
Pre-communication
stage
Personal media marketing
communication
characteristics ideally
suited for co-production
Communication
stage
Consumer receives a co-
produced communication
Post-communication
stage
Possible outcomes
produced by co-produced
communications for firms
and consumers
Note: dotted boxes were featured as part of the empirical assessment in the current paper’s Study 1 and
Delivery Time
Frequency Location
Format (text / image / video)
Product Recipient
Type (
Marketing message is sent to a consume
Increase consumer attitude
toward communication
Increase consumer
purchase intent
Increase consumer
response rate
Increase consumer WOM
recommendations
Decrease
consumer risk