Vous êtes sur la page 1sur 8

Journal of Business Research 68 (2015) 978985

Contents lists available at ScienceDirect

Journal of Business Research

Online brand community engagement: Scale development and validation

Brian J. Baldus a,, Clay Voorhees b, Roger Calantone b
California State University, Sacramento, Tahoe Hall 2015, 6000 J Street, Sacramento, CA 95819-6088, United States
The Eli Broad College of Business, North Business Complex, 632 Bogue Street, Michigan State University, East Lansing, MI 48824-1122, United States

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

Article history: In a quest for connecting with customers, the world's largest brands have gone online to develop communities to
Received 30 October 2013 interact with consumers. Despite widespread adoption less is known about what motivates consumers to contin-
Received in revised form 23 September 2014 ually interact in these communities. Across six studies, we develop and test a typology of online brand commu-
Accepted 24 September 2014
nity engagement (i.e., the compelling intrinsic motivations to continue interacting with an online brand
Available online 11 October 2014
community). We identify 11 independent motivations and test the scale's predictive power for participation in
an online brand community. This study provides a much needed renement to the disparate conceptualizations
Online brand communities and operationalizations of engagement in the literature. As a result, academic researchers can now rely on a di-
Scale development verse set of motivational measures that best t the context of their research, adding to the versatility of future
Branding research studies. The results provide managers with new insight in the motivations for and impact of interacting
Consumer motivation in online brand communities.
Brand community 2014 Elsevier Inc. All rights reserved.

1. Introduction measurement of these motivations can also assist in the development

of operational standards of excellence for this maturing medium of
Brands as disparate as the Boston Red Sox, Salesforce.com, Starbuck's brand communication. Despite this practical need, academic research
Coffee, Dell, General Motors, and Procter & Gamble are making signi- on the consumer motivations to participate in online brand communi-
cant investments in online brand communities in an effort to cultivate ties has struggled to keep pace with the changing landscape of the
stronger relationships with their consumers. Often, these communities industry (e.g., Brodie, Ilic, Juric, & Hollebeek, 2013). While early investi-
began as simple text forums where consumers share thoughts and gations in brand communities provide us with operational denitions of
questions about a brand, but over the past 15 years these communities these investments: Online brand communities represent a network of
have evolved and some have even blossomed into strategic market- relationships between consumers and the brand, product, fellow con-
ing investments designed to offer unique brand experiences in rich sumers, and the marketer (McAlexander, Schouten, & Koenig, 2002,
interactive multimedia environments. This increased sophistication p. 39) and insight into early motivations for community engagement
not only offers consumers a new array of opportunities within (e.g., Dholakia, Bagozzi, & Pearo, 2004) they fail to capture the complex-
these communities, but also carries a substantial cost for the brands. ity of motivations driving consumer engagement in communities for
For example, General Motors recently announced that they invest three reasons (e.g., Cova & Pace, 2006). First, these initial investigations
$30 million annually to simply generate content for their community are now a decade old and these initial conceptualizations don't account
on Facebook and are planning to continue this investment, despite for the new possibilities of interaction due to recent technological inno-
cutting their $10 million Facebook advertising budget (Barkholz & vations and substantial investments in these communities by their
Rechtin, 2012). brands.
While each brand community has a unique purpose, the one univer- Second, early investigations were necessarily limited to extreme
sal is that they represent an explicit marketing investment on behalf of lead users. Brand communities now have moved into the mainstream
the rm to develop long term connections with their current and poten- and it is common to nd as many early and late majority consumers
tial consumers (Zaglia, 2013). In order to increase returns on these sub- interacting in these communities as lead users. The increased diversity
stantial investments, marketing managers require better consumer in online brand communities challenges community managers to in-
insights into the motivations to participate in brand communities and crease participation rates and necessitate a broader set of marketing
the resulting attitudinal and nancial benets to the brand. Improved tools to reach the diverse types of community members.
Third, no prior study has undertaken a dedicated effort to understand
the unique dimensions of engagement for online brand communities.
Corresponding author. Tel.: +1 916 278 6978.
Several studies have examined channels, in general Calder, Malthouse,
E-mail addresses: baldus@csus.edu (B.J. Baldus), voorhees@broad.msu.edu and Schaedel (2009), brand channels (Hollebeek, Glynn, & Brodie,
(C. Voorhees), rogercal@broad.msu.edu (R. Calantone). 2014) and C2C communication (Hennig-Thurau, Gwinner, Walsh, &

0148-2963/ 2014 Elsevier Inc. All rights reserved.
B.J. Baldus et al. / Journal of Business Research 68 (2015) 978985 979

Table 1
Operationalizations of engagement in the marketing literature.

Online context

Citation Construct denition Dimensions/ Data General Dedicated
outlets MROCs
items collections social media brand
(forums and
outlets communities
media sites)

Current research Online brand community engagement is the 11 dimensions/ 2 qualitative/

compelling, intrinsic motivations to continue 42 items 4 quantitative
interacting with an online brand community

Hollebeek et al. (2014) Consumer brand engagement is a consumer's 3 dimensions/10 1 qualitative/

positively valenced cognitive, emotional and items 3 quantitative
behavioral brand-related activity during, or
related to, specic consumer/brand
interactions. p. 6

Sprott et al. (2009) Brand engagement in self-concept is the 1 dimension/8 5 quantitative

individual difference representing consumers' items
propensity to include important brands as part
of how they view themselves. p. 92

Calder et al. (2009) Consumer engagement with a website 8 dimensions/32 1 quantitative

is dened as a collection of experiences items
(consumer's beliefs about how a site ts into
his/her life) with the site. p. 322

Algesheimer et al. Community engagement is the consumer's 1 dimension/4 1 quantitative

(2005) intrinsic motivation to interact and cooperate items
with community members (p. 21)

Hennig-Thurau et al. eWOM communication is dened as as any 8 dimensions/27 1 quantitative

(2004) positive or negative statement made by items
potential, actual, or former customers about a
product or company, which is made available to
a multitude of people and institutions via the
Internet p. 39

Note: MROCS = Marketing Research Online Communities.

Gremler, 2004), but very few have examined communities centered on research on brand communities, review the scale development process,
brands in the online domain. As a result, our paper is the rst to truly cap- and discuss the implications of this research.
ture the unique engagement dimensions for these communities that
must capture motivations tied to the channel, other consumers, and the 2. Scale development and validation procedure
brand simultaneously. Without considering all these elements, our un-
derstanding is incomplete and overly generic. While these broader con- To develop measures for online brand community engagement, we
ceptualizations certainly have a lot of value to the literature, they must began with a review of the engagement literature (Algesheimer,
be complemented with context-specic investigations that provide ac- Dholakia, & Herrmann, 2005; Calder et al., 2009; Hennig-Thurau et al.,
tionable insights at a very granular level. This is particularly important 2004; Hollebeek et al., 2014; Sprott, Czellar, & Spangenberg, 2009)
from an area of marketing investment as important as online brand followed by a grounded theory approach to establish baseline
communities. dimensions of engagement in online brand communities and then
We attempt to close this gap by conducting a comprehensive exam- proceeded with a modied scale development processes. Table 2 pro-
ination of consumer motivations to participate in a broad variety of vides an overview of the entire process, which entailed two qualitative
brand communities and developing a measure of online brand commu- data collections, item generation, expert review, an exploratory data
nity engagement following a grounded theory approach. In doing so, collection, and two validation studies. In the following sections, we
this research contributes to the marketing literature by providing a plat- provide details on the entire process and criteria used at each stage of
form for future investigations into how these motivations inuence the development. (See Table 1.)
consumer behavior in online brand communities and in the market-
place following interactions in the community. Accordingly, our prima- 2.1. Identication of engagement dimensions
ry research question is what motivations do consumers have for
interacting with an online brand community? Because there has been limited research into online brand commu-
Results of the scale development process and subsequent nomolog- nity engagement, we follow a grounded theory approach (Spiggle,
ical net testing suggest that online brand community engagement is not 1994) to explore the domain of engagement to develop the scale for on-
unidimensional, but multidimensional. Therefore, extant measures of line brand community engagement. Specically, we began the process
engagement are too narrow to capture online brand community using a series of qualitative research efforts (focus groups and qualita-
members' diverse motivations. Ultimately this research enables both re- tive surveys) to identify consumer motivations for interacting with
searchers and managers to better understand consumer motivations for brand communities. Consistent with our earlier denition, these moti-
participating in brand communities and provides a widely-applicable vations served as our primary engagement dimensions. When these di-
platform for future research on brand communities. In the following mensions aligned with prior literature, we labeled them accordingly
section, we briey review the evolution of brand communities, current and for dimensions that were unique, we created new labels and
980 B.J. Baldus et al. / Journal of Business Research 68 (2015) 978985

Table 2 and identied preliminary themes related to brand community engage-

Scale development process. ment. In order to rene the initial listing of motivational themes, a
Steps in the process Details follow-up qualitative survey was administered.
Study 1 focus groups Two focus groups
Qualitative analysis of focus group 2.1.2. Study 2 open-ended surveys
transcripts to identify motivations An online panel company was then used to collect a sample of very
Focus group interviews
Study 2 Open-ended surveys Open-ended surveys of active brand
active (i.e., participate in an online brand community more than 23
community members times a week) online brand community members. Of the 70 completed
Qualitative analysis of responses to open surveys returned, 44 surveys were screened because they were not very
Extension of focus group
ended survey questions to rene the active members of online brand communities (i.e., participate in an
questions into open-ended
listing of motivations
questionnaire format online brand community more than 23 times a week). In addition, 2
This process resulted in the identica
tion of 11 distinct motivations for surveys were dropped for data quality concerns (e.g., lack of elaboration
interacting with an online brand on qualitative questions and speeding). The remaining 24 responses
community were analyzed for themes surrounding the motivations brand commu-
Item generation Generation of 494 items by research nity members have to interact with the brand community (11 male and
team based on 11 constructs
Item reduction and expert review 138 items selected for expert review
13 female, median age 40 years old). Open-ended survey questions
2 Marketing Faculty Members and 6 based on the focus group questioning route were used to assess the mo-
Marketing Doctoral Students Judged Items tivations of respondents. The responses were analyzed for themes sur-
94 Items remained rounding brand community members' motivations to interact with
Study 3 initial validation study Online survey of online brand
the community. Following this analysis, the research team met and
community members
72 items remained reconciled the results of both qualitative research efforts and developed
Conrmatory factor analysis a preliminary listing of 11 dimensions of online brand community en-
o Dimensionality
gagement. Following Rossiter (2002), tentative construct denitions
o Factor loadings
Validity were created for each dimension. Table 3 provides an overview of the
o Convergent validity 11 engagement dimensions and operational denitions.
o Discriminant validity
2.2. Initial item generation and reduction
Study 4 validation study and Online survey of online brand
establishment of a short form scale community members
Establishment of a short form version of 2.2.1. Item generation
the scale including 42 items across the Following the identication of the engagement dimensions, the
Conrmatory factor analysis
11 dimensions research team generated items using the construct denitions. In
o Convergent validity total, 2 researchers and 2 research assistants independently genera-
o Discriminant validity ted a total of 494 items (approximately 10 items per dimension per
Reliability person) to measure the 11 dimensions of brand community member
Study 5 validation of a short form Online survey of online brand
scale and nomological validity community members
testing Validation of a short form version of the
scale including 42 items across the 11 2.2.2. Item reduction and expert review
dimensions After creating a large pool of potential items, the research team met
Conrmatory factor analysis
Establishment of nomological validity
Validity to rene construct denitions, eliminate redundant items, and select
o Convergent validity items that had good face validity for expert review. In total, 138 items
o Discriminant validity were selected for expert review. Given the large number of constructs,
experts were randomly assigned half of the total items. A review of
Nomological validity testing
via OLS regression the item pool was conducted by two marketing faculty members and
Study 6 testretest reliability Online surveys of a panel of online six marketing doctoral students who were familiar with the branding
assessment brand community members at two time literature, but unaware of the specic focus of this research project.
periods, 30 days apart
Each expert was presented 69 items one at a time in random order.
For each item, judges were asked to assign the item to the most repre-
sentative construct denition. If a majority of the experts correctly
assigned the item to its intended denition, then the item was retained
for further testing. In total, 94 items were retained based on the expert
operational denitions that captured the essence of the motivation. De-
tails on the focus groups and the open ended survey are provided next. 2.3. Preliminary assessment of the scale

2.1.1. Study 1 focus groups 2.3.1. Study 3: initial validation

Focus group participants were recruited from a large Midwestern For study 3, an online panel company was contracted for the next
university's undergraduate marketing courses. Thirty students round of data collection to get a sample of 2839 US Internet users who
volunteered to participate in the focus groups for extra credit. 11 stu- are 18 and older. 1190 respondents indicated that they were not mem-
dents (6 male and 5 female, median age 21 years old) were identied bers of brand communities and were screened immediately from the
as active members of online brand communities and included in this study. 737 of the remaining responses were screened by the panel
phase of the research process. Focus group sessions were conducted company for speeding through initial survey questions. In total, 911
based on a questioning route developed specically for this study completed surveys were returned. Of the completed surveys returned,
(Krueger & Casey, 2009). Following the focus group sessions, the 282 responses were dropped for data quality (e.g., straight-lining and
researchers and assistant moderators (two research assistants not gibberish responses to open-ended questions). Two independent raters
aware of the theoretical background) met to discuss the transcripts. coded whether or not the remaining respondents were actually mem-
The researchers and assistant moderators reviewed the transcripts bers of a brand community (86% agreement on classifying members
B.J. Baldus et al. / Journal of Business Research 68 (2015) 978985 981

Table 3 of inspections and iterative model estimations, we removed a total of 23

Online brand community engagement dimensions. items due to large standardized residuals (N .25) (Gerbing & Anderson,
Engagement Denitions 1988, p. 189), lambdas below .707 (lambda's below .707 indicate
dimensions that random error determines more variation in the item than
Brand inuence The degree to which a community member wants to inuence what is determined by the latent construct), and signicant cross-
the brand. loadings as detected through an examination of Lagrange Multipli-
Brand passion The ardent affection a community member has for the brand. er indices. After deleting these items, a nal measurement model
Connecting The extent to which a community member feels that being a
was estimated that offered improved t ( 2 = 7168, df = 3662;
member of the brand community connects them to some good
thing bigger than themselves. CFI = 0.897; SRMR =0.054; RMSEA = 0.053).
Helping The degree to which a community member wants to help Moreover, using the results of this nal CFA, we assessed validity and
fellow community members by sharing knowledge, reliability based on the recommendations for Fornell and Larcker
experience, or time. (1981). The results provided support for convergent validity as each
Like-minded The extent to which a community member is interested in
discussion talking with people similar to themselves about the brand.
the average variance extracted (AVE) for each construct was greater
Rewards The degree to which the community member wants to gain than 0.50. Discriminant validity was also supported as the AVE for
(hedonic) hedonic rewards (e.g., fun, enjoyment, entertainment, friend- each scale exceeded the squared correlation between the construct
ly environment, and social status) through their participation and all other constructs in the measurement model. Finally, all construct
in the community.
reliabilities for the constructs exceeded 0.70, providing evidence of
Rewards The degree to which the community member wants to gain
(utilitarian) utilitarian rewards (e.g., monetary rewards, time savings, reliability.
deals or incentives, merchandise, and prizes) through their
participation in the community. 2.4. Study 4 validation and establishment of a short-form scale
Seeking The degree to which a community member wants to receive
assistance help from fellow community members who share their
2.4.1. Validation of the measurement model
knowledge, experience, or time with them.
Self-expression The degree to which a community member feels that the A validation dataset was collected to conrm the measurement
community provides them with a forum where they can model and establish a short-form version of the scale. This dataset was
express their true interests and opinions. collected through Amazon's Mechanical Turk service and all respon-
Up-to-date The degree to which a community member feels that the
dents were screened to ensure that they were active participants in
information brand community helps them to stay informed or keep up-to-
date with brand and product related information online brand communities. Of the 376 completed surveys returned, 19
Validation A community member's feeling of the extent to which other respondents indicated that they were not members of brand communi-
community members afrm the importance of their opinions, ties and were screened immediately from the study. 151 of the remain-
ideas, and interests. ing responses were screened for speeding through initial survey
questions and data quality (e.g., straight-lining, lack of elaboration and
gibberish responses to open-ended questions). Two independent raters
coded whether or not the remaining respondents were actually mem-
and nonmembers of brand communities). 285 responses were screened bers of a brand community (90% agreement on classifying members
because the respondents were not active members of brand communi- and nonmembers of brand communities). 45 responses were screened
ties. This data collection effort resulted in a usable sample of 344 of because the respondents were not active members of brand communi-
online brand community members (38% of respondents were male, ties. In total, 198 respondents (49% of respondents were male, median
median age 47 years old and median education 2 year college degree). age 31 years old, and median education 4 year college degree) provided
To analyze the results, we adopted Gerbing and Anderson (1988) data suitable for analysis.
updated approach to the Churchill (1979) paradigm that leverages Once again, we assessed the scales by iteratively estimating
the benets of structural equation modeling as opposed to item- models to identify any items that may be negatively affecting the
to-total correlation analysis. In line with similar scale development scale for each dimension by assessing standardized residuals, lamb-
efforts in marketing (Brocato, Voorhees, & Baker, 2012; Seiders, da loadings, and cross-loadings. The initial measurement model pro-
Voss, Godfrey, & Grewal, 2007), we conceptualize brand community vided good t to the data (2 = 4487, df = 2429; CFI = 0.85;
engagement as a second order formative construct making SEM an SRMR = 0.07; RMSEA = 0.07). As a result of the nal screening 5
ideal method to examine the loadings of the reective dimensions items were removed. Following the removal of these items, the mea-
at the rst level. surement model offered good t (2 = 3764, df = 2089; CFI = 0.87;
Specically, we iteratively estimated a series of measurement SRMR = 0.06; RMSEA = 0.06). Moreover, we found evidence for the
models where the 11 engagement dimensions were estimated as rst validity and reliability of each scale based on Fornell and Larcker
order reective scales. In these models, bad items were removed (1981) criteria.
and then the entire model was re-estimated and re-assessed. During
this process, items with large standardized residuals were removed 2.4.2. Development of a short-form scale
as they negatively affected the unidimensionality of each dimension Following the nal validation of the scale, we had identied 11 di-
(Gerbing & Anderson, 1988). In order to enhance the rigor of this mensions that were measured using 67 items. The original goal of this
initial investigation, we include measures for two related variables: research was to develop a short-form scale that adequately measured
expectations for community member behavior (e.g., I think that reg- all dimensions in a manner that both future researchers and managers
ular participation is necessary to be an active member of this brand could use in practice. As a result, we undertook one nal assessment
community) and intentions to share information from the communi- and revision to the measurement model with the goal of identifying
ty with others (e.g., I encourage my friends and family to participate an ideal subset of the best three to four indicators for each dimension
in this brand community). These variables were included to provide that would serve as the nal items in the short-form version of this
a more rigorous assessment of discriminant validity and ensure that scale. In order to accomplish this, for dimensions with more than four
items for the engagement dimensions were strictly measuring en- items retained, we selected the four items with the highest lambda
gagement and not expectations or outcomes of brand community loadings and if a construct had four or fewer items remaining, we
participation. retained all items. The measurement model for the short form version
Initial overall model t for the CFA was modest (2 = 12,380, df = of the scale had offered excellent t to the data (2 = 1248, df = 764;
5916; CFI = 0.844; SRMR = 0.075; RMSEA = 0.056). Through a series CFI = 0.94; SRMR = 0.05; RMSEA = 0.06). Once again, all dimensions
982 B.J. Baldus et al. / Journal of Business Research 68 (2015) 978985

exhibited construct reliability, convergent validity, and discriminant items retained for the short-form version of the scale as well as their
validity based on Fornell and Larcker's (1981) criteria (AVE's ranged respective ranges, means, standard deviations, and lambda loadings to
from 0.82 to 0.89). In Table 4 we provide a complete listing of all assist in the establishment of scale norms.

Table 4
Scale items, descriptive statistics, and factor loadings for the short form version of the scale.

Factor item Study 4 establishing Study 5 validating

scale scale

Range Mean SD Range Mean SD

Brand inuence (Study 4: CR = .87; Study 5: CR = .84)

1. I am motivated to participate in this brand community because I can help improve the brand and its products 111 7.02 2.84 .91 411 9.64 1.49 .94
2. I like to know that my comments and suggestions can inuence the brand and its products 111 7.35 2.78 .84
3. Increasing the inuence I have on the brand and its products makes me want to participate more in this 111 6.98 2.75 .89 211 9.37 1.63 .84
brand community
4. I hope to improve the brand or product through my participation and expression in this brand community 111 7.18 2.79 .91 411 9.75 1.45 .90

Brand passion (Study 4: CR = .87; Study 5: CR = .88)

1. I am motivated to participate in this brand community because I am passionate about the brand 111 8.75 2.20 .91 111 8.39 2.16 .95
2. I participate in this brand community because I care about the brand 211 8.90 2.12 .89 111 8.88 1.88 .87
3. I would not belong to a brand community if I did not have passion for the brand 111 8.79 2.31 .81
4. My passion for this brand's products makes me want to participate in this brand community 111 8.76 2.25 .89 111 8.45 2.09 .95

Connecting (Study 4: CR = .78; Study 5: CR = .82)

1. Increasing the strength of the connection I have with this brand community makes me want to participate 111 7.89 2.17 .90 211 8.92 1.79 .88
more in the community
2. Being part of this brand community makes me feel more connected to the brand 111 8.41 1.91 .74 411 9.37 1.63 .88
3. Being part of this brand community makes me feel more connected to other consumers of the brand 111 8.53 1.92 .82 211 8.89 1.91 .85

Helping (Study 4: CR = .86; Study 5: CR = .84)

1. I like participating in the brand community because I can use my experience to help other people 111 7.58 2.46 .82 211 9.06 1.73 .87
2. I like to share my experience and knowledge with others in this brand community to help them be more 111 7.74 2.39 .82
educated about the brand
3. I really like helping other community members with their questions 111 7.85 2.40 .91 211 8.84 1.82 .92
4. I feel good when I can help answer other community member's questions 111 8.34 2.29 .90 311 9.19 1.71 .90

Like-minded discussion (Study 4: CR = .86; Study 5: CR = .85)

1. I look forward to discussing my opinions about the brand with others who share the same interest as me 111 8.12 2.35 .87 111 8.88 1.88 .89
2. I enjoy conversing with people similar to myself in this brand community 111 8.37 2.16 .86
3. I look to this brand community when I want to discuss a topic with people who have similar interests 111 7.99 2.64 .85 111 8.37 2.14 .90
4. Having conversations with people in this brand community who share the same views about this brand 111 7.51 2.45 .87 111 8.36 2.16 .92
is important to me

Rewards (Hedonic) (Study 4: CR = .87; Study 5: CR = .87)

1. I like participating in this brand community because it is entertaining 111 8.29 2.23 .91 211 8.95 1.71 .92
2. Having fun is my main reason for participating in this brand community 111 7.58 2.71 .83
3. I participate in this brand community because I think it is fun 111 8.44 2.22 .89 211 8.94 1.78 .89
4. I nd participating in this brand community to be very entertaining 111 8.17 2.18 .88 111 8.73 1.90 .95

Rewards (utilitarian) (Study 4: CR = .82; Study 5: CR = .78)

1. I am motivated to participate in this brand community because I can earn money 111 2.83 2.84 .91 111 8.15 2.76 .94
2. If it weren't for the money, I wouldn't participate in this brand community 111 2.47 2.42 .89 111 5.46 3.06 .58
3. Receiving more money makes me want to participate more in this brand community 111 3.62 3.28 .78 111 8.41 2.77 .89

Seeking assistance (Study 4: CR = .88; Study 5: CR = .89)

1. I am motivated to participate in this brand community because I can receive help from other community members 111 7.86 2.49 .90 111 7.76 2.42 .95
2. I am motivated to participate in this brand community because community members can use their knowledge 111 8.09 2.50 .90 111 8.07 2.31 .94
to help me
3. I like participating in this brand community because it gives me an opportunity to receive help from other 111 7.95 2.47 .94 111 8.00 2.37 .94
community members
4. It is important to me to be able to use this community to nd answers to my questions about the brand 111 8.23 2.42 .83

Self-expression (Study 4: CR = .86; Study 5: CR = .85)

1. I feel that I can freely share my interests in the brand community 211 8.78 1.96 .89 511 9.78 1.49 .93
2. I would express any opinion or idea I had about this brand in this brand community 111 8.64 2.28 .85
3. I can always be myself when interacting with others in this community 111 8.55 2.32 .86 311 9.73 1.57 .90
4. This community makes it easy for me to express my true beliefs about the brand 111 8.74 2.08 .87 411 9.75 1.48 .89

Up-to-date information (Study 4: CR = .86; Study 5: CR = .84)

1. This brand community is my critical connection for new and important information about the brand and its products 111 8.60 2.26 .86 111 8.24 2.31 .91
2. When I want up-to-date information about this brand, I look to this brand community 111 8.95 2.15 .89 111 8.24 2.33 .89
3. This community keeps me on the leading edge of information about the brand 111 9.00 1.98 .85
4. This community is the best way to stay informed about new developments with this brand 211 8.86 2.11 .88 111 8.68 2.19 .89

Validation (Study 4: CR = .85; Study 5: CR = .85)

1. Receiving more afrmation of the value of my comments, makes me want to participate more in the brand community 111 8.00 2.31 .77
2. I feel good about myself when other community members share my ideas 111 8.00 2.10 .91 111 8.71 1.94 .91
3. I appreciate when others agree with the ideas I express in this brand community 111 8.26 2.12 .85 211 8.83 1.83 .90
4. When others support my ideas and opinions in this brand community, I feel better about myself 111 7.54 2.30 .87 111 8.35 2.18 .90

Note: All scales measured on a 010 Likert-type scale with anchors 0 = Strongly Disagree and 10 = Strongly Agree. Prior to analysis all values were recoded to a 111 range, which is
presented in all results tables. CR = Construct Reliability.
B.J. Baldus et al. / Journal of Business Research 68 (2015) 978985 983

2.5. Study 5 nal validation of the short-form scale and nomological va- community engagement dimensions do an excellent job predicting con-
lidity testing sumer motivations for participating in online brand communities.

2.5.1. Validation of the short form scale 2.6. Study 6 testretest reliability assessment
A nal validation dataset was collected to conrm the measurement
model for the short-form version of the scale and assess the predictive In the nal study, we assess the testretest reliability of the scales.
validity or nomological net properties of the short-form scale. The Specically, a few weeks after the original survey administered in
data for this study was gathered through a partnership with a consult- Study 5, 651 community members were re-contacted and asked to
ing rm specializing in the creation and management of online brand complete a brief follow-up survey. This survey included single items
communities and eight of their clients. Due to survey length require- for each dimension of engagement measured in Study 5. In total, 160
ments set by the data collection partner, three items were used to respondents completed the survey for a 25% response rate. To assess
measure each dimension of engagement. A link to our survey was testretest reliability, we conducted a number of tests. Specically, we
posted in each community and community members were offered gift conducted paired sample t-tests to assess changes in means across
card sweepstakes calibrated to community norms in exchange for time as well as correlations and Cronbach's alpha estimates for each
their participation in the survey effort. Ultimately, this process resulted dimension. The results of the paired t-tests revealed relative stability
in an average response rate of 21% across the eight communities, pro- in community member motivations over time. Specically, with the
viding us with 620 usable responses (47% of respondents were male, exception of Brand Inuence, Seeking Assistance, Self-Expression, and
median age 41 years old, and median education some college). Up-to-date Information all other engagement dimensions did not sig-
In this phase of the scale development process, we conducted one nicantly differ across time periods. Moreover, the average testretest
nal measurement model where all items from the short-form version correlation across all 11 dimensions was .60. Estimates for Cronbach's
of the scale were allowed to load on their respective constructs. The alpha calculated using scores from each time period ranged from .65
measurement model provided good t to the data (2 = 1153, df = (Validation) to .84 (Helping) with an average value of .74 across all
440; CFI = 0.97; SRMR = 0.06; RMSEA = 0.05). Moreover, all items dimensions. Taken together, the results suggest that engagement is
loaded highly and signicantly on their respective constructs and the relatively stable over time.
model results provided evidence of both convergent and discriminant
validity based on Fornell and Larcker's (1981) criteria. Table 5 presents 3. Discussion
the correlations and scale statistics for the data collected as part of
Study 5. There is some productive overlap between our study and extant lit-
erature, but the vast majority of our factors are completely distinct and
advance these early discussions. First, we found similarities with the
2.5.2. Assessing the nomological net broader conceptualizations of motivations to interact with an online
Once the measurement model was conrmed, we assessed the abil- brand community. Hennig-Thurau et al. (2004) and Nambisan and
ity of the 11 dimensions of online brand community engagement to pre- Baron (2007) found that there are different types of engagement that
dict intentions to participate in the brand community (Algesheimer propel people to interact with an online community. Broadly, these fac-
et al., 2005). According to regulatory engagement theory, higher levels tors are social status enhancement, social interaction, learning more
of engagement should lead to higher value perceptions of the communi- about using the product and having fun. Research in this stream places
ty (Higgins & Scholer, 2009). Therefore, we expect that online brand an overly strong emphasis on the uses and gratication paradigm which
community engagement should be positively related to participation viewed online communities from a one-way (i.e., marketer to consum-
intentions. At a basic level, all engagement dimensions correlated signif- er) point of view. These studies offered important insight into the moti-
icantly with participation intentions. To extend this assessment further, vations of community members as the internet evolved from an
we regressed participation intentions on the 11 engagement dimen- information storage and retrieval system to an interactive environment.
sions and dummy variables developed to reect community member- Our research is the rst dedicated effort to develop a multi-
ship for the eight brands. dimensional measure of online brand community engagement and the
The independent variables explained 44% of the variance in partici- rst substantial investigation into online brand community engagement
pation intentions and 8 of the 11 dimensions had a signicant impact since signicant advances in online communities such as the launch of
on the dependent variable. Specically, Brand Passion, Utilitarian Facebook, YouTube, and a litany of other Internet-based technologies
Rewards, and Validation were not drivers of participation intentions in and applications revolutionized online brand communities. Across six
this sample. Complete results of the regression analysis are presented studies, we developed and validated a short-form scale for online
in Table 6. Ultimately, the results suggest that the online brand brand community engagement across a variety of communities and

Table 5
Study 5 results of measurement model assessment and scale statistics nal validation study short form scale.

1 2 3 4 5 6 7 8 9 10 11

1. Brand inuence
2. Brand passion 0.67
3. Connecting 0.79 0.78
4. Helping 0.62 0.61 0.76
5. Like-minded discussion 0.59 0.69 0.75 0.73
6. Rewards (Hedonic) 0.60 0.60 0.72 0.58 0.65
7. Rewards (utilitarian) 0.09 0.04 0.09 0.03 0.03 0.14
8. Seeking assistance 0.37 0.49 0.54 0.60 0.69 0.49 0.07
9. Self-expression 0.68 0.54 0.67 0.60 0.55 0.59 0.09 0.36
10. Up-to-date information 0.63 0.66 0.72 0.54 0.67 0.62 0.04 0.53 0.51
11. Validation 0.51 0.55 0.67 0.63 0.61 0.51 0.20 0.45 0.46 0.53
Average variance extracted 0.80 0.86 0.75 0.80 0.81 0.85 0.67 0.89 0.82 0.80 0.81
Construct reliability 0.84 0.88 0.82 0.84 0.85 0.87 0.78 0.89 0.85 0.84 0.85

Note: All correlations greater than 0.11 and less than 0.11 are signicant.
984 B.J. Baldus et al. / Journal of Business Research 68 (2015) 978985

Table 6 3.2. Managerial implications

Study 5 results of regression analysis.

Independent variables The online brand community engagement scale is a reliable and valid
Brand inuence 0.24
tool for measuring consumer motivations for interacting with an online
Brand passion 0.07 brand community. Our results reveal that the engagement dimensions
Connecting 0.13 can explain up to 44% of the variance in participation intentions. However,
Helping 0.14 the directionality of the effects reveals a more complex story. Specically,
Like-minded discussion 0.09
two of the engagement dimensions (Seeking Assistance and Up-to-
Rewards (Hedonic) 0.20
Rewards (utilitarian) 0.05 Date Information) had signicant, negative effects on participation.
Seeking assistance 0.07 Ultimately, it appears that consumers primarily motivated by a
Self-expression 0.13 need for information are less inclined to participate, suggesting
Up-to-date information 0.10 that they may only log in to answer a question. This suggests that
Validation 0.02
community managers may need to nd new ways to engage their in-
R2 0.44 formation seekers to increase their connections with the brand if
Note: they want to draw them into becoming more active community
p b .10. members. In the following three sections, we will highlight three
p b .05.
important potential applications of the engagement scale.
p b .01.

3.2.1. Proling of community members

The scales introduced in this research would allow rms to prole
their current and potential community members. Armed with an im-
contexts. Moreover, in the fth study, we demonstrated that these proved understanding of what compels consumers to actions within
engagement dimensions predict intentions to participate in a brand their specic community, they could develop strategic marketing ac-
community. Finally, in Study 6, the results reveal that these motivations tions to best engage their community base. For example, a community
are relatively stable over time. that is comprised of members primarily motivated by inuencing
the brand should place an emphasis on providing feedback on how
3.1. Theoretical implications suggestions from the community are shaping the trajectory for the
brand. Alternatively, if members are driven by hedonic rewards, the
Our ndings extend initial studies on online brand community community should offer members special recognition and access to
engagement (e.g., Algesheimer et al., 2005; Dholakia et al., 2004; exclusive events.
Madupu, 2006) by providing an updated and comprehensive investiga- Moreover, consumers could be pre-screened and recruited for com-
tion into engagement in online brand communities. Without an up- munity interaction based on their motivational proles. For example, if
dated examination into the motivations of online brand community a rm sought to develop a sub-group within their community focused
users, marketing researchers must rely on dated operationalizations on idea generation for new products, they could productively isolate
that were vetted by lead users and may no longer apply to the current consumers who are driven by a need for up-to-date information and
landscape. Many of the prominent studies into brand communities brand inuence and likely experience stronger involvement, improved
to date have focused on relatively few highly visible and very unique idea generation, and better results overall than if they just pulled com-
communities (e.g., Harley Owners Group). As brand communities are munity members or their customers at random.
being created around an increasing variety of brands and continue to
migrate online, a more comprehensive investigation of online brand 3.2.2. Targeted communication efforts
communities is needed to account for the technological leaps that As we started this project, it was clear that the landscape for online
have occurred as well as the shifts in the membership in online brand communities had changed drastically since the initial investiga-
brand communities. For example, utilitarian rewards was primarily tions into this phenomenon, but the results of our research revealed
viewed as time savings or giving and receiving information in the past an even more complex set of engagement dimensions than we had
(e.g., Dholakia et al., 2004), with the evolution of the internet from an anticipated. As communities have evolved so has the membership
informational to a transactional medium, utilitarian rewards also in- base. Communities are no longer comprised of lead users who can be
clude monetary rewards, deals or incentives, merchandise, and prizes. activated exclusively by product information and access to new prod-
The short form scales developed in this manuscript provide marketing ucts. With a more heterogeneous audience with a range of motivations,
researchers with a versatile set of motivations that are suitable for rms must better understand the motivational composition of their
future investigation. communities before launching communication efforts within the com-
Another important contribution of this study is to delineate the munity (e.g., Kozinets, de Valck, Wojnicki, & Wilner, 2010). Specically,
domain of engagement. It is important to note that the engagement di- it appears prudent for community manager to micro-segment their
mensions developed here represent a comprehensive set of motivations community members to understand the motivations of each member
across a wide range of online brand communities and consumers. Prior at the individual member and then develop tailored communication
research exploring few communities to varying levels of depth has failed members for each individual that are designed to activate their specic
to capture the domain of engagement that we were able to in this study. motivations. Failing to tailor these messages could result in alienating
Finally, the provision of a complete battery of engagement dimensions some community members with poorly calibrated messages.
that have been validated across a range of brand communities and con-
sumers allows marketing researchers to extend the work on ofine 3.2.3. Lead user campaigns
brand communities to the online realm. This is an important Finally, while members of brand communities are certainly more
advancement as given the tremendous cost savings and scalability rms heterogeneous than the initial forums launched in the late 1990s,
are quickly shifting their focus and investments to harvesting these online there are still many lead users who actively participate in more main-
brand communities. By better understanding the motivations for using stream communities. The online brand community engagement scale
online communities, better parallels can be drawn to prior studies on introduced in this research can help managers identify these individuals
brand communities, customer to customer interactions, and brand at- within their community and activate them for specialized marketing ef-
tachment that were derived in ofine contexts. forts that rely on lead users as seeds for their messages (e.g., Kozinets
B.J. Baldus et al. / Journal of Business Research 68 (2015) 978985 985

et al., 2010). For example, online campaigns like the Fiesta Movement Barkholz, D., & Rechtin, M. (2012, May 21). Net worth: GM's paid-ad pullback high-
lights a key question for auto makers: Can Facebook sell cars? Automot News
launched by Ford required signicant investments in recruiting and (1 & 42).
screening agents to participate in this program. Leveraging the mea- Brocato, E.D., Voorhees, C.M., & Baker, J. (2012). Understanding the inuence of cues from
sures introduced in this research, rms like Ford could better identify other customers in the service experience: A scale development and validation.
Journal of Retailing, 88(3), 384398.
their own community members who are passionate about the brand, Brodie, R.J., Ilic, A., Juric, B., & Hollebeek, L. (2013). Consumer engagement in a virtual
want to help others understand the brand, want to connect with some- brand community: An exploratory analysis. Journal of Business Research, 66(1),
thing bigger than themselves, and ultimately shape the future of the 105114.
Calder, B.J., Malthouse, E.C., & Schaedel, U. (2009). An experimental study of the relation-
brand to immediately activate them as part of this movement. In ship between online engagement and advertising effectiveness. Journal of Interactive
doing so, they would experience signicant cost savings associated Marketing, 23(4), 321331.
with independent recruiting efforts and would likely experience less se- Churchill, G.A. (1979). A paradigm for developing better measures of marketing
constructs. Journal of Marketing Research, 16(1), 6473.
lection errors as brand community members have established histories
Cova, B., & Pace, S. (2006). Brand community of convenience products: New forms of cus-
with brand and the community as opposed to individuals recruited off tomer empowermentthe case My Nutella the Community. European Journal of
the street. Marketing, 40(910), 10871105.
Dholakia, U.M., Bagozzi, R.P., & Pearo, L.K. (2004). A social inuence model of consumer
participation in network- and small-group-based virtual communities. International
3.3. Limitations and future research Journal of Research in Marketing, 21(3), 241263.
Fornell, C., & Larcker, D.F. (1981). Evaluating structural equation models with
unobservable variables and measurement error. Journal of Marketing Research,
This research makes a substantial departure from the dominant 18(1), 3950.
paradigm of small sample case analysis pervasive in the brand commu- Gerbing, D.W., & Anderson, J.C. (1988). An updated paradigm for scale development
nity research by sampling from as large and broad a pool of online brand incorporating unidimensionality and its assessment. Journal of Marketing Research,
25(2), 186192.
community members as possible. In capturing such a broad view of Hennig-Thurau, T., Gwinner, K.P., Walsh, G., & Gremler, D.D. (2004). Electronic
online brand community members caution needs to be used when word-of-mouth via consumeropinion platforms: What motivates consumers
attempting to apply the norms from brand community engagement to articulate themselves on the internet? Journal of Interactive Marketing,
18(1), 38.
scales to individual online brand communities. Similar to Zaglia Higgins, E.T., & Scholer, A.A. (2009). Engaging the consumer: The science and art of the
(2013), future research should work to categorize the diverse sub- value creation process. Journal of Consumer Psychology, 19(2), 100114.
types of brand communities (e.g., research oriented online brand com- Hollebeek, L.D., Glynn, M.S., & Brodie, R.J. (2014). Consumer brand engagement in social
media: Conceptualization, scale development, and validation. Journal of Interactive
munities, fan sponsored online brand communities) to better under- Marketing, 28(2), 149165.
stand how stable these motivational proles are across different Kozinets, R.V., de Valck, K., Wojnicki, A.C., & Wilner, S.J.S. (2010). Networked narratives:
brands and types of communities. Moreover, researchers could examine Understanding word-of-mouth marketing in online communities. The Journal of
Marketing, 74(2), 7189.
situations in which certain motivations dominate participation inten-
Krueger, R.A., & Casey, M.A. (2009). Focus groups: A practical guide for applied research. Los
tions and if certain engagement dimensions are stronger under different Angeles, CA: SAGE.
situational circumstances. In addition, future research should explore Madupu, V. (2006). Online brand community participation: Antecedents and consequences.
Dissertation. The University of Memphis.
the mechanisms through which online brand community engagement
McAlexander, J.H., Schouten, J.W., & Koenig, H.F. (2002). Building brand community. The
affects loyalty to the brand outside of the community and if marketing Journal of Marketing, 66(1), 3854.
actions that are calibrated based on brand community engagement Nambisan, S., & Baron, R.A. (2007). Interactions in virtual customer environments: Impli-
can provide better results than incongruent tactics. cations for product support and customer relationship management. Journal of
Interactive Marketing, 21(2), 4262.
Rossiter, J.R. (2002). The C-OAR-SE procedure for scale development in marketing.
Acknowledgments International Journal of Research in Marketing, 19(4), 305335.
Seiders, K., Voss, G.B., Godfrey, A.L., & Grewal, D. (2007). SERVCON: Development and val-
idation of a multidimensional service convenience scale. Journal of the Academy of
Thank you Jacci Weber, Sam Herzing, & Cody Potter. Marketing Science, 35(1), 144156.
Spiggle, S. (1994). Analysis and interpretation of qualitative data in consumer research.
The Journal of Consumer Research, 21(3), 491503.
References Sprott, D., Czellar, S., & Spangenberg, E. (2009). The importance of a general measure of
brand engagement on market behavior: Development and validation of a scale.
Algesheimer, R., Dholakia, U.M., & Herrmann, A. (2005). The social inuence of brand Journal of Marketing Research, 46(1), 92104.
community: Evidence from European car clubs. The Journal of Marketing, 69(3), Zaglia, M.E. (2013). Brand communities embedded in social networks. Journal of Business
1934. Research, 66(2), 216223.