Online Information Review

Informing brand messaging strategies via social media analytics

Constantinos K. Coursaris Wietske van Osch Brigitte A. Balogh
Article information:
To cite this document:
Constantinos K. Coursaris Wietske van Osch Brigitte A. Balogh , (2016),"Informing brand messaging
strategies via social media analytics", Online Information Review, Vol. 40 Iss 1 pp. 6 - 24
Permanent link to this document:
http://dx.doi.org/10.1108/OIR-02-2015-0062
Downloaded on: 13 February 2016, At: 07:14 (PT)
References: this document contains references to 33 other documents.
Downloaded by New York University At 07:14 13 February 2016 (PT)

To copy this document: [email protected]

The fulltext of this document has been downloaded 77 times since 2016*
Users who downloaded this article also downloaded:
Wu He, Guandong Xu, (2016),"Social media analytics: unveiling the value, impact and implications of
social media analytics for the management and use of online information", Online Information Review,
Vol. 40 Iss 1 pp. - http://dx.doi.org/10.1108/OIR-12-2015-0393
Yoosin Kim, Rahul Dwivedi, Jie Zhang, Seung Ryul Jeong, (2016),"Competitive intelligence in social
media Twitter: iPhone 6 vs. Galaxy S5", Online Information Review, Vol. 40 Iss 1 pp. 42-61 http://
dx.doi.org/10.1108/OIR-03-2015-0068
Youquan Wang, Zhiang Wu, Zhan Bu, Jie Cao, Dun Yang, (2016),"Discovering shilling groups
in a real e-commerce platform", Online Information Review, Vol. 40 Iss 1 pp. 62-78 http://
dx.doi.org/10.1108/OIR-03-2015-0073

Access to this document was granted through an Emerald subscription provided by emerald-
srm:198285 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.

*Related content and download information correct at time of download.

 The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1468-4527.htm

OIR
40,1
Informing brand
messaging strategies via
social media analytics
6 Constantinos K. Coursaris and Wietske van Osch
Received 26 February 2015
Department of Media and Information,
Revised 23 April 2015 Michigan State University, East Lansing, Michigan, USA, and
31 May 2015
Accepted 22 July 2015 Brigitte A. Balogh
Department of Advertising and Public Relations,
Michigan State University, East Lansing, Michigan, USA
Downloaded by New York University At 07:14 13 February 2016 (PT)

Abstract
Purpose – The purpose of this paper is to offer a theory-driven, evidence-based approach to
developing a brand’s messaging strategies on social media encompassing three messaging dimensions,
namely, appeal, content, and richness.
Design/methodology/approach – Using longitudinal data from three Fortune 200 companies – Delta
Airlines, Wal-Mart, and McDonald’s – the authors empirically investigate comprehensive strategic
messaging framework. Using ANOVAs and regression analyses, the authors test a set of hypotheses
regarding the relations between a brand’s purchase involvement, its message appeal, message content,
and message richness, and engagement.
Findings – Findings reveal significant relations between purchase involvement and appeal. Furthermore,
the authors find that abstract content categories are best combined with richer media. Finally, both
transformation appeal and richer media have a highly significant and positive effect on engagement.
Research limitations/implications – The authors offer a theoretical ground and empirical
validation of both a comprehensive typology of content categories and a holistic strategic messaging
framework that can fill a significant void in the social media marketing literature that lacks
integrative models for assessing, classifying, analyzing, and in turn, informing future social media
marketing strategies.
Practical implications – The validated framework can help managers better understand the diversity
of messaging components as well as offer an analytical tool for assessing the nature of engagement
associated with each appeal and category.
Originality/value – To the best of the author’s knowledge, this paper offers the first comprehensive
typology of content categories and validates it in the context of a strategic messages framework using
real-world data finding strong support for all hypotheses.
Keywords Social media, Strategy, Engagement, Marketing, Empirical, Messaging
Paper type Research paper

Introduction
Recent studies have revealed that companies have grown their social media marketing
spending and are expected to continue to increase social media budgets in the next five
years (Moorman, 2015), making social media one of the fastest growing marketing
platforms in the world. The opportunity for novel forms of consumer interaction and
the unlimited scope of consumer markets, makes social media an attractive alternative
or complement to traditional marketing channels. Combined with its free, personal,
Online Information Review
Vol. 40 No. 1, 2016
pp. 6-24 This research was supported by Leo Burnett Detroit (www.lbdetroit.com). The content is
© Emerald Group Publishing Limited
1468-4527
solely the responsibility of the authors and does not necessarily represent the official views of
DOI 10.1108/OIR-02-2015-0062 Leo Burnet Detroit.
 as well as social nature, social media does not merely offer significant cost advantages Informing
compared to traditional media channels (Hoffman and Fodor, 2010), but further opens brand
up possibilities for targeting smaller niche markets – effectively reaching market
segments of one – and accruing unparalleled media exposure and referral marketing
messaging
(Carpenter and Shankar, 2012). strategies
Similar to the social media hype among marketing practitioners, an explosion of social
media marketing scholarship can be witnessed (Fan and Gordon, 2014; Laroche et al., 2013; 7
McCarthy et al., 2014; He and Zha, 2014; He et al., 2013, 2014). Yet, despite the burgeoning
social media marketing literature, few studies have provided theoretically grounded,
empirically validated, comprehensive typologies for assessing, classifying, analyzing, and
in turn, informing future social media marketing strategies. Our analysis of the social
media marketing literature revealed only four such typologies, each with substantial blind
spots, inadequate theoretical grounding, and limited empirical validation.
In order to fill this void in the literature, this study involves the theoretical
Downloaded by New York University At 07:14 13 February 2016 (PT)

amalgamation of three perspectives for analyzing brand messages – purchase

involvement, message appeal, and message richness – as well as the bottom-up
development of an empirical typology for classifying messaging content. While the
inclusion of existing marketing and communication theories offer adequate insights for
hypothesizing and understanding the relation between purchase involvement,
messaging appeal, and the media richness in the composition of a brand message,
no applicable typologies for categorizing message content were identified.
Therefore, using a Multi-Grounded Theory (MGT) approach (Goldkuhl and
Cronholm, 2010), we develop a typology of Facebook page marketing messages.
Despite our focus on a single channel, Facebook pages, generalizability of our findings
is enhanced by means of a longitudinal multiple-case design – involving the analysis of
three top Fortune companies – Delta Airlines, Wal-Mart, and McDonalds – that
represent the full range of product-service involvement categories – namely, airlines,
big box retailers, and fast-moving consumer goods (FMCGs)[1]. A total of 369 Facebook
posts were collected for analysis over two distinct six-week periods in the Spring and
Fall of 2012. Our final typology for classifying brand Facebook page posts
encompasses seven higher level, overarching categories, and 29 sub-categories, which
provide a comprehensive, yet, generalizable typology that allowed for the classification
of every single post coded across the two data periods.
Findings from our ANOVA and regression analyses reveal significant relations
between the four message components – purchase involvement, appeal, content, and
richness. Specifically, our findings show that brand purchase involvement – whether it
is high or low – significantly predicts the messaging appeal employed by brands in
their Facebook page posts. Furthermore, our analysis of content categories reveals
significant relationships between the context of a brand Facebook post and the media
type – text only, photo, video, or URL – included in the message. Additionally, findings
show that specific message appeal and high media richness of brand posts result in
significantly higher consumer engagement (i.e. likes, comments, and shares), whereas
no such relationship can be found for content.
The proposed typology of content categories as well as the results of our ANOVA
and regression analyses provide important insights for future research on social media
marketing strategies. Furthermore, the findings about effective message appeal and
message richness as well as the classification of message content are important tools
for informing the design and development of successful social media marketing
messages. Hence, this study offers a comprehensive, logical, and analytical framework
 OIR for effectively predicting engagement and consumer-brand interaction that can guide
40,1 both researchers and practitioners in their analyses and development of effective social
media marketing strategies.
This paper is organized as follows. First, we review literature on purchase
involvement, messaging appeal, messaging content typologies, and media richness as
well as propose a set of hypotheses that guide our empirical analysis. Then, we present
8 our research design, including case selection, data collection, typology development,
as well as the coding and data analysis process. Subsequently, we present our findings
including the typology and hypotheses testing. Finally, we discuss our findings and
implications for future research and practice.

Theoretical underpinnings
In order to provide a holistic assessment of the messaging behavior of brands on social
Downloaded by New York University At 07:14 13 February 2016 (PT)

media, we draw on existing theories regarding four characteristics of the brands and
the message, namely, purchase involvement, messaging appeal, messaging type, and
media type, as follows.

Purchase involvement
Purchase involvement or purchase-decision involvement refers to the extent of interest
and concern that a consumer brings to bear upon a purchase-decision task (Mittal,
1989). Purchase involvement thus encompasses the time, effort, and costs invested in
making a purchase, such as the research that may precede the transaction. Higher
levels of economic and time concerns are associated with higher levels of purchase
involvement. Similarly, in purchases that involve high uncertainty, consumers
experience high involvement because of the perceived risk of negative consequences
(Houston and Rothschild, 1978).
Purchase involvement thus relates to price comparison and risk reduction.
While involvement is a characteristic that resides within the consumer, certain
product categories can be conceived of as high or low involvement. High involvement
products are generally expensive and are associated with high potential risk.
Consumers do not always find these products inherently enjoyable, but instead they
invest time and attention because the product is important, expensive, and/or risky.
Examples include buying a home, financial investments, and/or airline tickets
(primarily business travel). When shopping for high purchase involvement goods,
consumers are seeking extensive information to support decision making. Hence,
messaging by high involvement brands focusses on informing rather than
entertaining (Lally, 2007).
Low involvement products, on the other hand, are commodities that do not
personally engage the consumer, hence, are neither perceived as fascinating nor risky
(Lally, 2007). Examples include food, beverages, and office supplies. In the case of low
involvement products, such as a cola drink, brands employ emotional appeal to get
consumers to select their brand over any other. Since consumers are unlikely to search
and evaluate product information extensively, messaging by low involvement brands
focus on rapid hedonic or otherwise affective appeals (Hawkins et al., 1983).

Messaging appeal: informational vs transformational

The literature on traditional advertising includes a variety of classification approaches
for messaging strategies, but applying the schemes consistently has proven
 challenging for both practitioners and researchers (see, e.g. Simon, 1971; Frazer, 1983; Informing
Aaker and Norris, 1982) as many approaches are either too complex or too simplistic. brand
In response to this apparent contradiction, Laskey et al. (1989) developed a dichotomous
framework that has proven to be both practically and theoretically valuable.
messaging
The primary classification scheme proposed by Laskey et al. (1989) is based on the strategies
typology proposed by Puto and Wells (1984), informational vs transformational
messages. Informational advertising emphasizes the transmission of factual, verifiable 9
data about the brand (i.e. rational appeal), while transformational advertising focusses
on the consumer’s experience with the brand (i.e. emotional appeal).
Rossiter and Bellman (2005) further define these major categories based on the
underlying purchase motivation with respect to the brand or category. Informational
messages address negatively originated motivations, such as problem removal or
avoidance, normal depletion, or incomplete satisfaction. Transformational messages,
containing a dominant psychological element, address positive-ending motives such as
Downloaded by New York University At 07:14 13 February 2016 (PT)

sensory gratification, intellectual stimulation, or social approval.

Laskey et al. (1989) used the informational/transformational dichotomy as a first step
in categorizing television advertising messages. Following the initial dichotomization of
messages, the authors proceeded to further divide messages within each category into
exhaustive and mutually exclusive sub-types. Specifically, they identified five types of
informational message strategies (comparative, unique selling proposition, preemptive,
hyperbole, and generic-informational) and four transformational approaches (user image,
brand image, use occasion, and generic-transformational).
Given the limited applicability of these sub-types to social media marketing messages,
we only adopt the dichotomous framework from Laskey et al. (1989) and thus analyze
social media marketing messages from the informational-transformational dichotomy.
In addition, we develop a typology that is relevant to social media and reflective of the
observed data, which is discussed in the methods section below.
Amalgamating the theory on purchase decision involvement – as outlined above – and
the informational-transformational dichotomy, we anticipate that high involvement brands
adopt rational appeals, i.e., informational appeal, whereas low involvement brands adopt
emotional appeals, i.e., transformational appeal. Therefore:
H1a. The higher the brand involvement, the more informational the messaging
appeal.
H1b. The lower the brand involvement, the more transformational the messaging
appeal.
Furthermore, because of the hedonic nature of transformational messages, these
messages display improved fit with the inherently hedonic interests and motivation of
consumers using social media. Hence, we anticipate that transformational messages
will be related to higher consumer engagement:
H2. The more transformational the messaging appeal, the higher the consumer
engagement.
Messaging typologies
An assessment of existing typologies for assessing Facebook or other social media
messages reveals that only a few such typologies exist to provide a theoretical foundation
for this study. Furthermore, from the identified typologies, as summarized in Table I,
none adopt a strategic lens or address the use of social media as a critical element
 OIR Jensen and Jepsen
40,1 Source (2006) Waters et al. (2009) Jenkins (2011) Kwok and Yu (2012)

Industry Online marketing Non-profit For-profit brands Hospitality

communications (varied) (restaurants)
Platform Online Facebook YouTube Facebook
Total n na 275 profiles 30 social video 982 messages
10 (275 non-profit) advertisements (12 restaurants)
Data analysis Literature review Content analysis Exploratory Algorithmic text
method mining via SVM
Metrics Four categories Disclosure, media Emotional content, Likes, comments,
(typology of OMCs and types, contact points message and message keywords
dimensions) related tools creative strategy,
virality
Domain Communication Communication Shareability Engagement
Downloaded by New York University At 07:14 13 February 2016 (PT)

Limitations Theoretical Provides basic Provides only Post type popularity

(abstract) rather descriptives of emotional contest; does not
than empirical non-profits’ qualification of consider brand
(concrete); does not Facebook profiles messaging content, strategy or
Table I. consider brand without considering no consideration for messaging content;
Overview of existing strategy, or qualify strategy, content, product category, no cross-industry
social media messaging content or product category and purchase insight
message typologies involvement

of the business and marketing strategy. Additionally, only one study focusses on
engagement metrics.
As Table I shows, the first study ( Jensen and Jepsen, 2006) focusses on online
communication tools as highly diverse and distinct from offline tools. Although the
study provides relevant insights into the diversity of online marketing communication
(OMC) tools, the study is solely conceptual with no concrete empirical insights
regarding messaging content, engagement metrics, or business strategies.
The second study by Waters et al. (2009) presents the results of a content analysis of
275 non-profit company Facebook profiles. The study only provides basic profile
descriptives that led to the authors’ conclusion that most non-profit organizations have
not integrated Facebook applications into their social network presence. Hence, there is
no discussion of Facebook as a strategic marketing tool.
The third study by Jenkins (2011) involves a virality analysis of 30 YouTube videos
by for-profit brands. Although the study assesses emotional content, it ignores the
cognitive content embedded in these video messages as well as subsequent consumer
engagement characterizing virality.
The fourth and final study (Kwok and Yu, 2012) provides the most detailed typology
through an analysis of 982 Facebook messages by restaurants. However, the authors
merely classify messages into two categories: communication and marketing messages;
and compare their respective popularity. However, the mutual exclusivity
(i.e. discriminant validity) of these two categories can be challenged, and they
provide little insight into underlying appeal, content, and engagement.
In short, then, the existing typologies provide few theoretical insights that could
inform a typology for classifying and analyzing Facebook-based marketing messages.
Hence, in the next section, we will present how the message typology of brand
Facebook page communications was developed.
 Media richness theory (MRT) Informing
MRT (Daft and Lengel, 1986) posits that each communication medium can be described brand
by its ability to reproduce any associated contextual cues, e.g. visual ones such as
gestures, during a message’s transmission. This ability is referred to as the medium’s
messaging
“richness” and, by extension, increases its effectiveness. Media richness is a function of strategies
the medium’s capacity for immediate feedback, the number of cues available, and
language variety (Daft and Lengel, 1986). 11
Facebook pages – as used by brands – offer different media types, each of which
represents a different richness level. More specifically, Facebook page posts can be
solely text-based or support photo, video, or URL. Hence, a continuum of message
richness emerges for Facebook posts, ranging from leaner media-types – i.e., text-only
posts – to richer media types – i.e., posts containing embedded photos, videos, or URLs.
Here, we treat URLs as rich media since their inclusion in a Facebook page post is
associated with a thumbnail image and these URLs frequently link to external, rather
Downloaded by New York University At 07:14 13 February 2016 (PT)

than embedded, videos.

Ultimately, richer, more personal communication is more effective than leaner, less
rich media (Newberry, 2001), hence, we anticipate that the use of richer media – such as
photo, video, or URL – is associated with higher consumer engagement. Therefore, we
propose that (Table II and Figure 1):
H3. Greater message richness – through the inclusion of richer media types such as
photos, videos, or URLs – is associated with higher consumer engagement.

Hypotheses Specification

H1a Higher purchase involvement Winformational appeal

H1b Lower purchase involvement Wtransformational appeal
H2 Transformational appeal Whigher engagement
H3 Richer media Whigher engagement Table II.
H4a Abstract content categories Wricher media Summary of
H4b Product awareness messages Wphoto/URL hypotheses

Message
Characteristics
Brand
H1 Message H2
Purchase Engagement
Appeal
Involvement

Message
Content
H3
H4
Figure 1.
Message
Proposed research
Richness
model
 OIR Research design
40,1 As brand-consumer interactions are highly contextual, a single case-study design – by
analyzing a single Facebook page to explore messaging strategies and categories –
would pose a challenge to generalizability. Hence, using a multiple-case study
approach, three Facebook pages, Delta Airlines, Wal-Mart, and McDonald’s, were
selected to represent like-brands and their respective social media marketing
12 communications. A multiple-case study research design involves the collection and
comparison of data on two or more cases (Yin, 1994). The advantage of having multiple
cases is the potential to augment a within-case analysis with cross-case comparisons
(Eisenhardt, 1989) and to strengthen results through pattern-matching, thereby,
increasing confidence in the robustness of theoretical results (Yin, 1994). The unit of
analysis was the brand Facebook page post. Results were obtained based on an
extensive content analysis of two six-week periods of messaging by the three brands,
as will be further explained below. In what follows, we describe the case selection
Downloaded by New York University At 07:14 13 February 2016 (PT)

process as well as the data collection and analysis process.

Multiple-case study design and case selection
With respect to social media marketing, Delta Airlines, Wal-Mart, and McDonald’s
represent leaders in their respective domains (see Table III). All three brands are among
the top Fortune companies ranked by gross revenue, enjoy strong brand equity, and
maintain a considerable social media presence. Second, these brands represent different
levels of purchase-decision involvement, that is, the extent of interest and concern that
a consumer brings to bear upon a purchase-decision task (Mittal, 1989) in relation to the

Delta Airlines Wal-Mart McDonald’s

Purchase-decision High (expensive durables/ Mixed (products evoking Low (inexpensive

involvement consumables such as airline either low or high equipment or low cost
tickets) involvement) consumables)
Product/service Airlines; consumables Big box retailers; Fast-moving consumer
industry category consumables, and goods (FMCGs);
durables consumables
Facebook page www.facebook.com/delta www.facebook.com/ www.facebook.com/
Wal-Mart McDonalds
No. of fans
(3/12/12)a 98,550 13,419,825 1,800,469
Talking about
b
(3/12/12) 5,373 357,878 608,486
Were here
(3/12/12)c 111 512,215 2,195,043
Data collection 24 posts 64 posts 23 posts
Period 1 (03/12-
04/30/12)
Data collection 56 posts 169 posts 33 posts
Period 2 (08/19-
09/30/12)
Notes: aThese represent the engagement metrics at time of first data collection phase (Spring 2012);
b
number of unique people who have created a story about the page (e.g. like the page, post on wall, like/
Table III. comment/share page posts, answer question, respond to event, mention page, tag page, check in at
Juxtaposing the page’s location, or recommend location); ccount of total check-ins and mobile device location shares the
three case companies page accrued over the past seven days
 products and services provided by these brands. Representative case studies (brands) Informing
were selected from each of three purchase-decision involvement levels, namely: low brand
involvement as found for McDonald’s and other FMCGs; mixed involvement as is the
case with Wal-Mart; and other big box retailers; and, high purchase-decision
messaging
involvement found in Delta Airlines and other expensive and greater risk-bearing strategies
goods and services. Third, the purchase cycle was one additional consideration in our
case selection process resulting in a selection of brands representing both consumables 13
and durables (Sullivan and Sheffrin, 2003). Purchase cycle – consumables vs durables –
is a key factor in brand messaging when considering purchase-decision frequency and
the need for tailored messaging.

Data collection
The data collection process was divided over two six-week periods, one collected in
Spring 2012 and one in Fall 2012 (see Table III). Our motivation for the temporal
Downloaded by New York University At 07:14 13 February 2016 (PT)

division of our data collection process was twofold. First, it further increased the
longitudinal nature of our data set, which helps to further mitigate generalizability
concerns. Second, it allowed us to assess if any temporal bias in messaging strategies,
categories, and consumer engagement exists as a result of seasonal events. The last
two rows in Table III above summarize the number of brand posts analyzed for each
brand for each of the two data collection periods.

Typology development
In the absence of an existing typology of Facebook messages, part of our coding
scheme had to be developed from the bottom-up, which will be explained here. The
remainder of the coding scheme – encompassing categories for messaging appeal,
media type, messaging timing, and engagement are discussed in the next paragraph in
which the coding process is explained.
In order to inductively develop the Facebook messaging typology, we used an
alternative and extended coding approach referred to as MGT (Goldkuhl and
Cronholm, 2010). Whereas the Original Grounded Theory approach (Glaser
and Strauss, 1967) embraces complete inductivism, the MGT approach combines and
integrates aspects of inductivism and deductivism in a dialectical, iterative,
and enriching fashion. Hereto, MGT combines two forms of grounding, namely:
empirical grounding, which is the original form of grounding as advocated in grounded
theory; and internal grounding, which is a conceptual form of grounding as advocated
by the MGT approach and establishes an explicit congruence within the theory itself.
The combining of these two grounding approaches resulted in a three-stage typology
development process that unfolded through the oscillation of theory generation and
theory grounding (i.e. validation), as follows:
(1) Inductive coding (or open coding): the initial theory development process was
performed inductively with an open mind and as free as possible from pre-
categorizations, implying that this stage allowed the data to speak for itself. In our
study, we explored and inductively grouped a pilot data set of two weeks of brand
Facebook page messages (March 1-15)[2] from the three case companies without
theoretical grounding with the aim of developing initial categories and labels.
(2) Conceptual refinement: after the inductive establishment of initial categories,
the second step in the theory development process encompassed the critical
reflection on the data itself. Although various approaches to conceptual
 OIR refinement exist, we used functional determination (i.e. assessing the function a
40,1 phenomenon has and how this is related to other functions). Hence, this stage
thus involved the formulation of definitions of emergent second-order
messaging categories.
(3) Theory condensation (selective coding): following the establishment of second-
order (i.e. lower level) messaging categories, the final step in the typology
14 development process involved condensation or selective coding; the densifying
of the typology, by identifying a few core categories under which we could
include all the data, and proposing the final coding scheme.
The MGT process was conducted by two graduate research assistants under the
guidance of the primary investigators. An initial interrater reliability of 93.7 percentage
agreement and 0.73 Cohen’s κ (i.e. substantial agreement; cf. Landis and Koch, 1977)
provided a strong assessment of the coding process reliability and the emergent coding
Downloaded by New York University At 07:14 13 February 2016 (PT)

scheme validity.
Following the completion of the coding scheme (i.e. the developed typology), we
empirically validated the coding scheme by testing it with the two six-week periods of
data from the three brands, as will be explained in the coding process below. The final
coding scheme is presented in the results section.

Coding process
All brand posts were analyzed by the same pair of independent coders that completed
the typology development. The final coding scheme encompassed the inductively
developed typology of brand messaging types as well as categories for messaging
appeal (transformational vs informational), message richness vis-à-vis media types (text
only, includes URL, photo, and/or video), timing (date/time of post; date/time of first
and last comment), and consumer engagement (likes, comments, and shares; or votes in
the case of polls).
The interrater agreement between the two coders for Period 1 (Spring 2012)
was 95.6 percent (0.73 Cohen’s κ) and for Period 2 (Fall 2012) was 97.5 percent
(0.74 Cohen’s κ). Face-to-face meetings were organized to discuss and reconcile
disagreements. After the entire data set had been coded, summary tables and graphs
were constructed for each category of the coding scheme.
The empirical validation of the inductively developed typology resulted in minor
changes, primarily in terms of integrating two lower level categories that were not
mutually exclusive. Furthermore, a few posts in the expanded data set presented novel
messaging types that were re-evaluated and necessitated the mere adjustment of
existing rather than the construction of additional sub-categories.

Data analysis
In order to test our hypotheses, we used SPSS ANOVA – for pairwise comparisons
between brands – and SPSS regression – for analyzing the effects of different
messaging content types, message appeal, and message richness on engagement.
Due to the extreme differences in the number of posts for each brand and the high
sensitivity of ANOVA to unequal sample sizes (Howell, 2009), we used SPSS select
cases to randomly select an equal amount of posts across all three brands as
determined by the brand with the least posts, namely McDonalds (n ¼ 56). For SPSS
regression, we used the full data set, i.e., without randomly selecting an equal subset of
posts per brand. Since the aim of the regression analysis was not to compare across
 brands but to find relations between the different messaging components – appeal, Informing
content category, and media type – and consumer engagement, the unequal sample brand
sizes for each brand were unimportant.
For the regression of the different messaging components – appeal, content
messaging
category, and media type – we used the raw scores for likes, comments, and shares, as strategies
well as computed an additional dependent variable, weighted engagement, which we
defined and calculated as follows: 15
Weighted engagement ¼ 0:5  SðLÞþ 1  SðC Þþ 1:5  SðS Þ

where L refers to likes, C refers to comments, and S refers to shares.

The weighing of likes, comments, and shares, was based on the increasing
level of cognitive involvement and exposure (in terms of visibility in one’s own
Downloaded by New York University At 07:14 13 February 2016 (PT)

personal network) associated with each of these engagement activities, respectively

for the consumer.
For the analysis of message richness, we additionally computed a compound
message richness variable that encompasses the combined values for photos, videos,
and/or URLs. That is a text-only post will have a message richness value of 0, a post
that includes only one of these media types will have a value of 1, and a post that
includes two or all three of these media types would have a value of 2 or 3, respectively.

Results
Typology
The typology that emerged from the MGT approach described above, revealed seven
overarching messaging categories of brand awareness, corporate social responsibility,
customer service, engagement, product awareness, promotional, and seasonal were
developed encompassing a total of 23 sub-categories across these seven umbrella
categories. Although a single Facebook message could be categorized under more than
one umbrella category – e.g., a post that fosters both brand awareness and product
awareness simultaneously – the underlying sub-categories are designed to be mutually
exclusive. Table IV offers definitions and written examples of the seven overarching
messaging categories as well as each of the underlying sub-categories. Additionally,
Table IV provides information on the occurrence of each of these overarching and
sub-categories per brand and per time period (i.e. Spring vs Fall 2013) for all of the
posts (rather than the randomly selected subset).
One obvious observation about the content categories in the inductive typology is
their distinctiveness with respect to complexity and abstraction. Whereas some of the
content categories are abstract – namely, brand awareness, corporate social
responsibility (CSR), and customer service – other categories are concrete – namely,
promotional and seasonal messages. Using MRT, we anticipate that for Facebook page
posts to be effective, abstract categories are associated with richer media, in order to
make the message more informative and appealing to consumers. Hence, the following
hypothesis emerges:
H4a. More abstract content categories – i.e. brand awareness, CSR, and customer
service – are associated with richer media (i.e. video, photo, or URL).
Furthermore, for concrete messages as well as for the engagement category – because
of the highly varied nature of the sub-categories for engagement – we anticipate no
 Downloaded by New York University At 07:14 13 February 2016 (PT)

16
OIR
40,1

Table IV.
Summary of
inductive Facebook
messaging typology
Typology Occurrence (% of total)
Sub- Delta Delta Wal- Wal- McDonald’s McDonalds
Categories categories Definitions/examples T1 T2 Mart T1 Mart T2 T1 T2

Brand Posts that build company presence and attentiveness in digital consumer
awareness market 10 (1.68) 10 (1.68) 3 (0.50) 19 (3.19) 6 (1.01) 5 (0.84)
Promotions Posts that contain the use of celebrity and/or event sponsorship which
mentions brand’s name, e.g., Jewel’s performance of “The Wal-Mart Song” – 2 (0.34) 2 (0.34) 1 (0.17) 6 (1.01) –
Heritage Posts that seek to bring consumer into company’s history; trivia, and
employee spotlights, e.g., Delta album shows flight uniforms over time 9 (1.51) 3 (0.50) – – – 2 (0.34)
Located Posts that contain elements promoting events including in-store citings,
store events, etc. – – – 16 (2.69) – 1 (0.17)
News Post that contains news elements found that inform the consumer of current
events, company business – 3 (0.50) – 1 (0.17) – –
Operations Posts that inform consumers about production processes and behind-the-
scenes operations, e.g., Delta video showing JFK airport renovations 1 (0.16) 1 (0.17) 1 (0.17) – – 1 (0.17)
Personality Posts that contain build brand personality, use a spokesperson/
personification, use first person, etc. – 1 (0.17) – – – –
Pop-culture Posts that build brand pop-culture through product use/display promotions,
cascading trends, etc. – – – 1 (0.17) – 1 (0.17)
Corporate Posts that build brand image of supporting, strengthening community 1 (0.17) 3 (0.50) 8 (1.34) 17 (2.86) 2 (0.34) –
social Awareness Posts contain elements that support non-profit organizations and/or raise
responsibility awareness of causes, e.g., Wal-Mart post advertises Project 7’s Save the Earth
(CSR) Fresh Mint Gum 1 (0.17) 2 (0.34) 8 (1.34) 11 (1.85) 2 (0.34) –
Fundraisers Posts highlighting charity fundraiser support, e.g., Wal-Mart’s March of
Dimes of St. Jude’s Children’s Hospital – 1 (0.17) – 6 (1.01) – –
Customer Posts that aim to build consumer knowledge about product, industry, and
service brand changes 5 (0.84) 5 (0.84) 2 (0.34) 3 (0.50) – 1 (0.17)
Apologies Posts that contain product recall information – – – – – –
Openings Posts that contain store opening notifications, e.g., Wal-Mart posts about a
new location opening 4 (0.67) 2 (0.34) 1 (0.17) 2 (0.34) – –
Outages Posts that contain service outage notifications – – 1 (0.17) – – –

(continued )
 Downloaded by New York University At 07:14 13 February 2016 (PT)

Typology Occurrence (% of total)

Sub- Delta Delta Wal- Wal- McDonald’s McDonalds
Categories categories Definitions/examples T1 T2 Mart T1 Mart T2 T1 T2

PSA Posts that contain a consumer notice even if it was originally posted by
another source, e.g., Delta informs travellers to pack 3-ounce approved
liquid containers 1 (0.17) 3 (0.50) – 1 (0.17) – 1 (0.17)
Engagement Posts that build consumer connections/communities through brand
interaction 16 (2.69) 38 (6.39) 43 (7.23) 60 (10.08) 11 (1.85) 29 (4.87)
Assistance Posts that include advice, home improvement, cooking, life tips, recipes for
the consumer, e.g., Wal-Mart posts a dinner recipe made with products sold
in-store 1 (0.17) 10 (1.68) 3 (0.50) 10 (1.68) 1 (0.17) 1 (0.17)
Community Posts encouraging consumers to follow another brand platform (e.g. Twitter,
YouTube), e.g., Wal-Mart posts about following @Wal-MartSpecials
on Twitter – 1 (0.17) 2 (0.34) 6 (1.01) – –
Likes Posts that ask consumers to “Like” a message, e.g., Wal-Mart post requests to
“Like” post if consumers are a fresh berry fan – – 6 (1.01) 14 (2.35) – –
Photos/ Posts which direct consumers to look at new photo albums and/or videos
videos posted by the brand, e.g., Delta promotes new photo album on Sky magazine
over the years 1 (0.17) 3 (0.50) – 3 (0.50) – 23 (3.87)
Polls Posts with questions prompting answers from consumers through multiple
choice questions – – – 15 (2.52) 2 (0.34) 1 (0.17)
Questions Posts that prompt consumer for answers through fill in the blank/open-ended
questions, e.g., Wal-Mart posts the question “What are you planting?” with
ad for gardening supplies 6 (1.01) 8 (1.34) 29 (4.87) 5 (0.84) 3 (0.50) –
Appreciation Posts that recognize, show gratitude to consumers, e.g., Delta offers 20% of
Sky Club membership for reaching 300,000 likes 2 (0.34) – – 6 (1.01) 1 (0.17) –
Directional Post that direct a consumer to click/do something (except for Liking),
e.g., McDonald’s instructs to watch a video of Kyle, a rising basketball player;
voice by LeBron James 6 (1.01) 16 (2.69) 3 (0.50) 1 (0.17) 4 (0.67) 4 (0.67)

(continued )
strategies
Informing

messaging

17
brand

Table IV.
 Downloaded by New York University At 07:14 13 February 2016 (PT)

18
OIR
40,1

Table IV.
Typology Occurrence (% of total)
Sub- Delta Delta Wal- Wal- McDonald’s McDonalds
Categories categories Definitions/examples T1 T2 Mart T1 Mart T2 T1 T2

Product All posts which build product knowledge/understanding, and existence 6 (1.01) 23 (3.87) 43 (7.23) 78 (13.11) 14 (2.35) 19 (3.19)
Awareness Name Posts that mention products sold at the store but are not specifically
brands produced by the posting brand, e.g., Wal-Mart advertises that they sell new
Philadelphia Indulgence in a post 1 (0.17) 23 (3.86) 33 (5.55) 59 (9.92) – –
House Posts mentioning products sold, produced by brand, e.g., McDonald’s posts
brands about their seasonal minty-green Shamrock Shake 5 (0.84) – 10 (1.68) 19 (3.19) 14 (2.35) 19 (3.19)
Promotional Posts that are designed to stimulate immediate or near future purchases
through monetary incentives 6 (1.01) 5 (0.84) 4 (0.67) 17 (2.86) 2 (0.34) 14 (2.35)
Deal Posts include instant reward for consumer to make a purchase, e.g.,
discounts, coupons, deals, specials e.g., Delta offers 20% off selected tickets
to Latin America and the Caribbean 2 (0.34) 5 (0.84) 3 (0.50) 14 (2.35) – 8 (1.34)
Chance Post offers incentives to make a future purchase for a potential reward, such
as a contest, giveaway 4 (0.67) 0 (0) 1 (0.17) 2 (0.34) 2 (0.34) 6 (1.01)
Rebate Posts that contain immediate purchase incentives but with a future reward
such as a rebate – – – 1 (0.17) – –
Seasonal Posts that remind, inform consumers of seasonal, annual events, and related
products by the brand 4 (0.67) 3 (0.50) 27 (4.54) 20 (3.36) 11 (1.85) 2 (0.34)
Holiday Posts that mention or advertise specific holidays such as Valentine’s,
Christmas, or New Years 1 (0.17) – 8 (1.34) 7 (1.18) 3 (0.50) 1 (0.17)
Season Posts that reference a climatic or sports season e.g., McDonald’s posts a video
to advertise their McCafe Shamrock shake-down 0 (0) 2 (0.34) 7 (1.18) 9 (1.51) 4 (0.67) 0 (0)
Event Posts mentions timely event that is not a holiday such as graduation, tax day,
or spring break e.g., Wal-Mart posts a question to ask how consumers are
enjoying the warm weather 3 (0.50) 1 (0.17) 12 (2.02) 4 (0.67) 4 (0.67) 1 (0.17)
 significant relations with message richness. Hence, engagement, promotional, and Informing
seasonal posts can be enacted with any media type. brand
Finally, for product awareness, we anticipate posts in this content category to be
significantly related to photo messages, since product awareness – i.e. enhanced
messaging
product recognition and recall – is most easily established by including a photo of the strategies
product or a direct link (i.e. URL) to the product. Thus:
H4b. Product awareness messages are associated with posts containing a photo 19
or URL.
The next section will discuss the results from the hypotheses testing.

Hypothesis testing
In what follows, we will discuss the findings of our hypotheses testing as summarized
Downloaded by New York University At 07:14 13 February 2016 (PT)

in Table V.
With respect to messaging appeal – i.e., informational vs transformational
messages – we analyzed the link between brand purchase-decision involvement and
appeal and between appeal and engagement.
First, a one-way ANOVA for brand purchase involvement on appeal reveals that
there are significant differences between Delta, on the one hand, and McDonalds and
Wal-Mart, on the other hand, with Delta employing significantly more informational
messages (Welch test; p ¼ 0.019). Purchase decision involvement for Delta is much
higher than for Wal-Mart and McDonalds, hence, rational appeal is an important
element of their marketing communication strategy. For Wal-Mart and McDonalds, on
the other hand, transformational messages represent their dominant marketing
communication strategy.
Second, a regression of appeal on engagement shows that transformational
messages indeed have a positive effect on weighted engagement ( p ¼ 0.024). Thus,
emotional appeal in Facebook page posts results in enhanced consumer engagement.
With respect to the effect of message richness on engagement – H3 – we found that
richer media indeed significantly predict weighted engagement ( p ¼ 0.005).
With respect to the content categories from our messaging typology, we analyzed the
link between the level of abstraction of the messaging category and the richness of the
media type used for each posts. The results of the regression analysis showed that abstract
messages are indeed significantly correlated with richer media ( p ¼ 0.05). More specifically,
brand awareness posts – which involve corporate story-telling – are associated with videos
( p ¼ 0.012). CSR and customer service messages are accompanied by URLs ( p ¼ 0.034 and
p ¼ 0.023, respectively), which is likely related to the need for additional background
information pertaining to the post content. Hence, H4a was supported.

Hypotheses Specification p-value Result

H1a Higher purchase involvement Winformational appeal 0.019 Supported

H1b Lower purchase involvement W transformational appeal 0.019 Supported
H2 Transformational appeal Whigher engagement 0.024 Supported
H3 Richer media Whigher engagement 0.005 Supported
H4a Abstract content categories Wricher media 0.05 Supported Table V.
H4b Product awareness messages Wphoto/URL 0.013/0.039 Supported Hypotheses testing
 OIR In line with our expectations, significant associations between neither concrete
40,1 messages – i.e., promotional and seasonal messages – nor engagement posts and
message richness were found. Thus it appears that promotional, seasonal,
and engagement posts are accompanied by any media-type regardless of their
richness. Furthermore, in support of H4b, we found that indeed product awareness
posts are more likely accompanied by posts including photos ( p ¼ 0.013) or URL
20 ( p ¼ 0.039) in order to support enhanced product recognition and recall.

Exploratory analysis of remaining relations

In addition to assessing the hypothesized relations tested above, there were many more
remaining, non-hypothesized relations due to the lack of appropriate theories. Hence,
we also conducted exploratory regression analyses for testing the significance of the
relation between the various content categories and messaging appeal as well as
Downloaded by New York University At 07:14 13 February 2016 (PT)

between content categories and engagement using SPSS regression. Furthermore,

we used additional ANOVA analyses to test for any significant differences in message
content categories and media types employed by the three different brands.
Our exploratory analysis of relations between content categories and strategies
revealed that brand awareness and product awareness messages are most likely
accompanied by informational strategies, whereas customer service posts are most
likely accompanied by transformational strategies. With respect to weighted (i.e.
consumer) engagement, we found none of the content categories to be significant
predictors. This implies that whether engagement with a post is high or low depends to
a large extent on the appeal and the message richness of the post rather than the
content per se.
Furthermore, we found that Delta and Wal-Mart use significantly more engagement
and promotional content categories (see Table III) than McDonalds, however, the usage
of the remaining content categories from our typology was consistent across the three
brands, despite their substantial differences in purchase-involvement and appeal. With
respect to media types, our exploratory analysis reveals that Delta and McDonalds use
significantly more photos than Wal-Mart, yet, for the other media types – i.e., text only,
videos, and/or URL – there were no significant differences in the proportion of their
usage by the three brands.

Post hoc test for temporal bias

Post hoc analyses were conducted for all variables to assess temporal biases between
the data from the Spring and the Fall. None of the variables showed significance, except
for weighted engagement (F ¼ 4.682; p ¼ 0.031). This implies that whereas the brands’
usage of specific messaging appeal, content categories, and media types were
consistent between the two periods, consumer engagement with the brand grew until
the Fall, likely reflecting the overall increase in Facebook fans and messaging
frequency since the Spring period. The lack of temporal bias for all remaining variables
indicates that the typology can be consistently applied across time periods.

Discussion
As social media marketing spending continues to grow and social media marketing
scholarship proliferates, developing theoretical grounded, empirically validated,
comprehensive typologies for classifying and analyzing social media marketing
messages becomes imperative to the effective implementation of social media
 marketing strategies with the aim of positively impacting consumer engagement and Informing
brand image. Nevertheless, our analysis of the existing social media marketing brand
literature reveals only four such typologies, each with substantial blind spots,
inadequate theoretical grounding, and limited empirical validation.
messaging
To fill this void in the literature, this paper presented the results of a theoretical strategies
amalgamation of three perspectives for analyzing brand messages – purchase
involvement, message appeal, and message richness – as well as the bottom-up 21
development of an inductively developed typology for classifying messaging content
using a MGT approach. The typology was developed through using data of three top
Fortune companies – Delta Airlines, Wal-Mart, and McDonalds – and subsequently
tested with data from two distinct six-week periods.
Using the full set of data (n ¼ 369 posts), hypotheses testing – using SPSS ANOVA
and regression – reveals that purchase involvement of the brand significantly predicts
the messaging appeal employed by brands in their Facebook page posts. Furthermore,
Downloaded by New York University At 07:14 13 February 2016 (PT)

our analysis of content categories reveals significant relationships between the context
of a brand Facebook post and the media type – text only, photo, video, or URL –
included in the message. Additionally, findings show that specific message appeal and
high message richness of brand posts result in significantly higher engagement,
whereas content categories have no effect on engagement.

Research and practical implications

Therefore, this study offers several contributions. First by proposing and testing a
holistic framework including a comprehensive typology of content categories as well as
an assessment of purchase involvement, message appeal, and message richness, we
aim to bring value to both researchers and practitioners. In doing so, we hope this
classification scheme can inform future research that aims to compute additional
correlations among the four components of social media marketing messages – namely
purchase involvement, appeal, content, and richness – as well as consumer
engagement.
Second, we hope that the proposed framework can help managers better understand
the diversity of messaging components and the potential for infinite variations between
these components as well as offer an analytical tool for assessing the nature of
engagement associated with each appeal and category.
Third, our findings regarding the effect of message appeal, richness, and content
categories on engagement – in particular the insight that engagement is largely
predicted by appeal and richness rather than content – can help inform best practices
specific to the level of purchase involvement of the brand.

Limitations and future research

Although this paper represents an initial attempt to offer a comprehensive typology of
Facebook page marketing messages, more work remains to be done. Particularly, we
identify three limitations of the current study and associated directions for extending
the current work. First, the applicability of the proposed typology could be validated in
the context of other social media platforms, to see if the same messaging categories
hold across these media platforms and if different platforms have different emphases.
Anecdotally, a quick scan of Delta Airlines’ social media usage reveals that Delta
Airlines uses Facebook largely for creating brand awareness, engagement, and
communicating news and promotions, while Twitter is predominantly adopted for
 OIR customer service – in particular its DeltaAssist channel. Hence, although the same
40,1 categories hold across media vehicles, various media will place different emphasis on
the multiple messaging categories. Consequently, through exploring the typology
in the context of different platforms, we can develop an understanding of the diverse
messaging strategies and patterns elicited by the specific affordances of each medium
(e.g. immediacy and message length).
22 Second, although the selection of the three brands for this study – Delta Airlines,
Wal-Mart, McDonald’s – was aimed at providing a comprehensive sample in terms of
industry categories and consumption frequencies – airlines, big box retailers, and
FMCGs/durables-consumables – as well as purchase involvement (Day, 1970) – high,
medium, low involvement, respectively – future research should aim to broaden the
scope of industry categories and the number of brands analyzed within each category.
One way to do this is by analyzing if the same messaging categories can be applied or
need to be modified for analyzing non-profit organizations.
Downloaded by New York University At 07:14 13 February 2016 (PT)

Third, although we directly tested the relations between message appeal and
message richness, on the one hand, and engagement, on the other hand, no such direct
relation was tested between message content and engagement. Rather, our existing
study proposed a mediated relationship where message content needs to be coupled
with a specific level of message richness in order to generate engagement. Our reason
for proposing the mediated relationship is partially related to the fact that our proposed
content typology may not be generalizable to all organization types, specifically in the
popularity of each content category, and hence, was outside the scope of this project.
However, we do encourage future research to also explore direct effects of message
content on engagement as the former relates to the domain of application (e.g. for-profit
vs non-profit; education vs retail).
To conclude, the typology provided in this paper provides a comprehensive lens for
Facebook pages marketing messages, in particular, and for social media messaging, in
general. While several additional directions for future research have been identified, the
multiple-case and longitudinal nature of our study, as well as the careful selection of
critical (i.e. successful), but highly diverse case companies enhance confidence in the
robustness and generalizability of our findings.

Notes
1. Sometimes also referred to as consumer packaged goods.
2. Please note that the pilot data set was separate from our final Spring data set.

References
Aaker, D. and Norris, D. (1982), “Characteristics of TV commercials perceived as informative”,
Journal of Advertising Research, Vol. 22 No. 2, pp. 22-34.
Carpenter, G.S. and Shankar, V. (2012), Handbook of Marketing Strategy, Edward Elgar
Publishing, Cheltenham and Northampton, MA.
Daft, R.L. and Lengel, R.H. (1986), “Organizational information requirements, media richness and
structural design”, Management Science, Vol. 32 No. 5, pp. 554-571.
Day, G.S. (1970), Buyer Attitudes and Brand Choice, Free Press, New York, NY.
Eisenhardt, K. (1989), “Building theories from case study research”, Academy of Management
Review, Vol. 14 No. 4, pp. 532-550.
 Fan, W. and Gordon, M.D. (2014), “The power of social media analytics”, Communications of the Informing
ACM, Vol. 57 No. 6, pp. 74-81.
brand
Frazer, C. (1983), “Creative strategy: a management perspective”, Journal of Advertising, Vol. 12 messaging
No. 4, pp. 36-41.
Glaser, B.G. and Strauss, A.L. (1967), The Discovery of Grounded Theory: Strategies for
strategies
Qualitative Research, Aldine Publishing Company, Chicago, IL.
Goldkuhl, G. and Cronholm, S. (2010), “Adding theoretical grounding to grounded theory – 23
towards multi-grounded theory”, International Journal of Qualitative Methods, Vol. 9 No. 2,
pp. 187-205.
Hawkins, D.I., Best, R.J. and Coney, K.A. (1983), Consumer Behavior: Implications for Marketing
Strategy, Business Publications Inc., Plano, TX.
He, W. and Zha, S. (2014), “Insights into the adoption of social media mashups”, Internet Research,
Vol. 24 No. 2, pp. 160-180.
Downloaded by New York University At 07:14 13 February 2016 (PT)

He, W., Wang, F.K. and Zha, S.H. (2014), “Enhancing social media competitiveness of small
businesses: insights from small pizzerias”, New Review of Hypermedia and Multimedia,
Vol. 20 No. 3, pp. 225-250.
He, W., Zha, S.H. and Li, L. (2013), “Social media competitive analysis and text mining: a case
study in the pizza industry”, International Journal of Information Management, Vol. 33
No. 3, pp. 464-472.
Hoffman, D.L. and Fodor, M. (2010), “Can you measure the ROI of your social media marketing?”,
MIT Sloan Management Review, Vol. 52 No. 1, pp. 41-49.
Houston, M.J. and Rothschild, M.L. (1978), “Conceptual and methodological perspectives on
involvement”, in Jain, S.C. (Ed.), 1978 Educators’ Proceedings, American Marketing
Association, Chicago, IL, pp. 184-187.
Howell, D.C. (2009), “Unequal cell sizes do matter”, available at: www.uvm.edu/~dhowell/
StatPages/More_Stuff/Unequal-ns/unequal-ns.html (accessed April 19, 2013).
Jenkins, B. (2011), “Consumer sharing of viral video advertisements: a look into message and
creative strategy typologies and emotional content”, Capstone Project, available at: www.
american.edu/soc/communication/upload/blaise-jenkins.pdf (accessed December 3, 2012).
Jensen, M.B. and Jepsen, A.L. (2006), “Online marketing communications: need for a new typology
for IMC?”, Journal of Website Promotion, Vol. 2 Nos 1/2, pp. 19-35.
Kwok, L. and Yu, B. (2012), “Spreading social media messages on facebook: an analysis of
restaurant business-to-consumer communications”, Cornell Hospitality Quarterly, Vol. 20
No. 10, pp. 1-11.
Lally, L. (2007), “Degrees of delight: a model of consumer value generated by e-commerce”, IRMA
International Conference, pp. 1006-1007.
Landis, J.R. and Koch, G.G. (1977), “The measurement of observer agreement for categorical
data”, Biometrics, Vol. 33 No. 1, pp. 159-174.
Laroche, M., Habibi, M.R. and Richard, M.O. (2013), “To be or not to be in social media: how brand
loyalty is affected by social media?”, International Journal of Information Management,
Vol. 33 No. 1, pp. 76-82.
Laskey, H., Day, E. and Crask, M.R. (1989), “Typology of main message strategies for television
commercials”, Journal of Advertising, Vol. 18 No. 1, pp. 36-41.
McCarthy, J., Rowley, J., Jane Ashworth, C. and Pioch, E. (2014), “Managing brand presence through
social media: the case of UK football clubs”, Internet Research, Vol. 24 No. 2, pp. 181-204.
Mittal, B. (1989), “Measuring purchase-decision involvement”, Psychology & Marketing, Vol. 6
No. 2, pp. 147-162.
 OIR Moorman, C. (2015), “CMO survey report: highlights and insights”, The CMO Survey, available at:
http://cmosurvey.org/results/survey-results-february-2015/ (accessed January 23, 2015).
40,1
Newberry, B. (2001), “Media richness, social presence and technology supported communication
activities in education”, available at: http://learngen.org/resources/module/lgend101_
norm1/200/210/211_3.html (accessed April 20, 2013).
Puto, C.P. and Wells, W.D. (1984), “Informational and transformational advertising: the
24 differential effects of time”, in Kinnear, T.C. (Ed.), Advances in Consumer Research XI,
Association for Consumer Research, Provo, UT, pp. 638-643.
Rossiter, J.R. and Bellman, S. (2005), Marketing Communications: Theory and Applications,
Pearson, Prentice Hall, New York, NY.
Simon, D. (1971), The Management of Advertising, Prentice Hall, Englewood Cliffs, NJ.
Sullivan, A. and Sheffrin, S.M. (2003), Economics: Principles in Action, Pearson Prentice Hall,
Upper Saddle River, NJ.
Downloaded by New York University At 07:14 13 February 2016 (PT)

Waters, R.D., Burnett, E., Lamm, A. and Lucas, J. (2009), “Engaging stakeholders through social
networking: how nonprofit organizations are using Facebook”, Public Relations Review,
Vol. 35 No. 2, pp. 102-106.
Yin, R.K. (1994), Case Study Research: Design and Methods, Applied Social Research Methods
Series, Vol. 5, 2nd ed., Sage, Thousand Oaks, CA.

About the authors

Dr Constantinos K. Coursaris is an Associate Chairperson and the Director of Graduate Studies
for the Department of Media and Information, and an Associate Professor in the Department of
Advertising and Public Relations at the Michigan State University. He is also an Adjunct
Professor with Usability/Accessibility Research and Consulting. Constantinos’ background
consists of extensive scholarly and applied experience in media and information spaces. Coursaris
studies user motivations, expectations, and experiences with new media and the consequent design
implications with a focus on social systems. He supplements his research expertise with consulting on
social media for governance and/or marketing, and has trained diverse organizations in North-
America, Europe, the Middle East and North Africa (MENA). Dr Constantinos K. Coursaris is the
corresponding author and can be contacted at: [email protected]
Dr Wietske van Osch is an Assistant Professor in Media and Information at the Michigan
State University. She studies social media and online communities for collaboration, knowledge
sharing, boundary-spanning, and innovation with a focus on the design of generative systems
and communities, both inside and outside organizations. Her research has been acknowledged
with many of the domain’s most prestigious dissertation and paper awards, including awards
from the ACM SIGMIS, Academy of Management, and Association for Information Systems.
Her work on Enterprise Social Media has recently received a three-year grant from the National
Science Foundation. Current research projects involve extensive industry collaborations for
research on enterprise social software, big data, and social media marketing with companies
including Steelcase and Leo Burnett.
Brigitte A. Balogh (MA, The University of Texas at Austin) is a PhD candidate at the
Michigan State University in Media and Information Studies specializing in Advertising and
Public Relations. Her primary areas of research include positive deviance, native advertising,
brand communities, and enterprise social software. Her dissertation focusses on brand equity,
purchase intent, and consumer engagement intention with social media, and the community and
cultural factors impacting these areas.

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]