Creating brand engagement on digital, social and mobile media Edward C Malthouse, Professor of Integrated Marketing Communications, Medill School, Northwestern University Bobby J Calder, Professor of Marketing, Kellogg School of Management, Northwestern University Mark Vandenbosch, Professor of Marketing, Ivey School of Business, Western University Key Words: life goals, elaboration, cocreation, brand value Introduction As digital, social and mobile media platforms become more common, customer engagement is becoming increasingly important. Consumers are no longer limited to a passive role in their relationships with firms. They can easily create their own brand-relevant, user-generated content (UGC) and distribute it to large audiences. The possibility of this sort of engagement is changing the way that firms interact with their customers (e.g., Malthouse et al. 2013). While most firms now react to UGC, especially when it is negative
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Creating brand engagement on digital, social and mobile media
Edward C Malthouse, Professor of Integrated Marketing Communications, Medill
School, Northwestern University
Bobby J Calder, Professor of Marketing, Kellogg School of Management, Northwestern
University
Mark Vandenbosch, Professor of Marketing, Ivey School of Business, Western
University
Key Words: life goals, elaboration, cocreation, brand value
Introduction
As digital, social and mobile media platforms become more common, customer
engagement is becoming increasingly important. Consumers are no longer limited to a
passive role in their relationships with firms. They can easily create their own brand-
relevant, user-generated content (UGC) and distribute it to large audiences. The
possibility of this sort of engagement is changing the way that firms interact with their
customers (e.g., Malthouse et al. 2013). While most firms now react to UGC, especially
when it is negative (e.g., van Noort and Willemsen 2012), companies increasingly have
the opportunity to engage proactively with consumers.
It is sometimes assumed that digital, social, and mobile platforms foster engagement in
and of themselves. That is, by their very nature these platforms are more interactive, so
they must by this very fact increase engagement. But such a view rests on an overly
simplistic view of engagement.
There has been extensive work on defining engagement as an important new marketing
construct that is not synonymous with the properties of any particular media platform
(Brodie et al. 2011, 2013; Hollebeek 2011a; Hollebeek 2011b; Hollebeek et al. 2014).
There is agreement that “Customer engagement (CE) is a psychological state that occurs
by virtue of interactive, cocreative experiences with a focal agent/object (e.g., a brand) in
focal service relationships … [and is a] dynamic, iterative process” (Brodie et al. 2011, p.
260). In other words, engagement should be thought of as the result of a certain kinds of
experiences with a brand, where some of these experiences may be designed for the
express purpose of creating stronger engagement. But how does this process of creating
engagement work if it is not simply a matter of exposing brands “on” certain kinds of
platforms?
This chapter deals with the process of creating engagement. Specifically we hypothesize
that engagement results from experiential contact points that prompt cognitive
elaboration about how a brand helps consumers achieve goals in their lives. We give the
results of one natural experiment testing this hypothesis in detail and cite other work that
tests it in different ways. We discuss how to harness this insight in digital, social and
mobile environments and cite relevant examples from marketing practice.
Creating engagement through experiences that connect to goals or values
Our analysis of engagement expands on the framework proposed by Calder and
Malthouse (2008, Figure 2) and Calder, et al. (2009, Figure 1). Consistent with the
Brodie et al. (2011) definition above, Calder and Malthouse think of engagement as being
rooted in one or more experiences that reflect consumers’ goals or values, and that
engagement causes various consequences such as product purchase and usage. It should
thus be an understanding of experiences that shows the way to creating engagement. As
Calder and Malthouse (2008) discuss, these experiences reflect the individual’s
interaction with the product over time as a way of accomplishing personal life goals or
manifesting larger values. Isaac, et al. (2015) extends this notion further by explicitly
defining engagement as “a multilevel, multidimensional construct that is reflected by the
thoughts and feelings consumers have about one or more rich experiences involved in
reaching a personal life goal or value.” Experiences are at the intersection of personal
goals and values in consumers’ lives and their connection to the brand in a way that adds
value to the brand. A brand that connects with consumers’ lives in this way will foster
engagement and thus attract the customer’s loyalty, prompting repurchase and use. As
distinct from just the notion of an experiential brand, the architect of an engaging brand
must understand and articulate how the brand contributes to specific goals or values.
The Evaluation versus the Design Perspective
It is important to note that engagement can be approached from two useful perspectives.
From one perspective, a marketer can measure how engaging a brand is at a point in time.
A marketer could employ generic questions about experiences with a brand. For example,
Brakus et al. (2009) suggested that dimensions of experience- sensory, affective,
intellectual, and behavioral. Hollebeek et al. (2014) measure cognitive, activation and
affection dimensions. The Calder-Malthouse approach employs direct measures of
experiences based on potential consumer beliefs and feelings about the extent to which a
brand links to a goal or value in a qualitatively meaningful way. Two examples of
goal/values and associated measurement items are given below.
Utilitarian Experience
It (brand) shows me how to do things the right way.
You learn to improve yourself from this (brand).
It (brand) helps me to make up my mind.
I learn a lot about things to do or places to go from this (brand)
Social Experience
I often bring up things I have seen about this (brand) in conversations with other people.
I especially like to follow what other people post about this (brand).
I frequently send things or links to things regarding this (brand) on the web to friends and
family
These items are generally crafted to reflect the way some consumer might talk about their
actual experiences with the specific brand in question.
Based on agree-disagree responses to assessment scale items like these Calder and
Malthouse use second-order factor measurement model. Overall engagement is derived
from a set of goal/value relevant experiences as measured by the items associated with
each particular experience. A number of studies provide evidence for the validity of this
approach to evaluating engagement.
The Design Perspective
Apart from evaluating consumers’ present level of engagement with a brand, a marketer
may be more action-oriented. The concern is with designing some marketing activity that
produces greater engagement with the brand. From this perspective, a marketer wants to
design experiences that create engagement. Here engagement can be thought of more as
an independent variable to be manipulated rather than as an outcome variable to be
measured.
The marketer wants to design activities in which consumers will experience the
connection of the brand with a life goal or value. The objective is to have consumers
actually or virtually engage with the brand in the pursuit of the goal or value. The
experience is one of active goal pursuit. It is the engagement created by this pursuit that
produces a stronger relationship with the brand.
To anticipate the sort of marketing activity used in the research reported below, the
marketer could, for example, design a contest using digital, or social, or mobile media.
This contest could be one in which consumers describe how a brand is linked to some
goal or value that is important in the consumer’s life. The contest is thus designed to
create an activity in which the consumer incrementally experiences engagement with the
brand.
Both the evaluation and the design perspective are consistent with the view that
engagement flows from experiences that connect the consumer with a goal or value. With
the design perspective, however, we are applying this theory as a guide to constructing
the marketing activity. The intention is have the consumer experience the brand in this
special way.
The process of employing a marketing activity (activation) in this way can be diagramed
as in Figure 1 (Calder and Malthouse, 2003). From this perspective it is apparent how and
why digital, social, and mobile useful can be very powerful tools for producing
engagement. The capability to have consumers involved in goal-relevant, user-generated
content fits perfectly with this perspective.
As a footnote to this discussion, we point out that the design perspective clarifies a
common point of confusion about engagement. Engagement is often confused with the
act of consuming the brand. Engagement is defined in term so activities such as paying
more attention to the brand, or “leaning forward”, or “engaging” in search behaviors. But
these should be regarded as potential consequences of engagement. Engagement itself
comes from experiences in which the brand is actually or virtually connected with a
goal/value. Marketers should focus on designing these experiences, experiences that in
turn produce positive changes in brand behaviors, where these behaviors may reflect an
increase in engagement.
Two Example Activities from Marketing Practice
The marketing task becomes one of designing experiences in which the consumer makes
the connection between the life goal and the brand. As noted digital, social and mobile
media are especially well suited to design interactive, co-creative experiences. They
enable customers to actively make, and cognitively elaborate on, the connection of the
brand to a life goal. In effect, customers actively elaborate on, or think through their
relationship with the brand. Such elaboration is known to have powerful effects on the
formation of attitudes (e.g., Petty and Cacioppo 1986). Again, the act of elaborating
reinforces the role of the brand experience in connecting to goals/values and thereby
producing a greater sense of engagement.
We again emphasize the importance of the intersection of brand and life goals/values in
an actual experience. A contact point could simply ask consumers to elaborate on the
positives of a brand, but this would not produce engagement as defined here. There must
be an experience that leads to an active connection of the brand to a life goal/value.
Kit Kat example of engagement activation
We begin with a practical example to make these ideas more concrete. To illustrate the
process of creating engagement by designing brand-life goal connection experiences
consider some examples from the Kit Kat candy bar’s use of Facebook. We contrast two
actual Kit Kat promotions. The first promotion was called the Fan of the Month.
Customer were invited to post a picture of themselves taking a break with Kit Kat, or
comment on, like or share other pictures. For example, a fan named Robin posted a
picture (Figure 2) of her eating a Kit Kat on top of a mountain she had just climbed. This
is a perfect example of the intersection. The Kit Kat brand has long associated itself with
the event of taking a break. Inviting customers to post pictures causes elaboration in that
the customer must think about the benefit of having a Kit Kat when they need it most.
Robin has customized the Kit Kat benefit—for her it provides much-needed energy after
an arduous climb. The picture reminds her and others of the value of taking a break with
Kit Kat.
Figure 2 about here
The second promotion, called the Game Time Give Away, invited Kit Kat customers to
give their email address on Facebook in order to win an NFL beverage pail, t-shirt, pen,
or other NFL merchandise. It is easy to find other examples of such promotions, where
customers get something in return for giving their contact information. Such promotions
may be an effective way of gathering names to build a database of customers and may
produce benefits by associating Kit Kat with the NFL brand, but we argue that it is
ineffective at creating engagement. The act of providing an email address itself does not
cause the consumer to make and elaborate on a connection between the brand and a life
goal.
Mein burger example of engagement activation
As a second example, consider a marketing program from McDonalds designed to
engage German customers. The McDonalds “Burger Battle” invited customers to design
their own sandwich. An interactive website provides consumers with a set of possible
ingredients that can be dragged onto a canvas to create a new sandwich. The consumer
names the creation, shares it on social media, and encourages friends to vote for it. Those
with the most votes are made in test kitchens and evaluated by juries, and the finalists are
made and sold in McDonalds stores in Germany for a period of time. This is a clear
example of cocreation, where consumers contribute to the design of the product and
participate in the creation of value, but consider the contest from the perspective of
personal goals. One can imagine a German consumer being attracted to certain aspects of
the McDonalds brand such as the convenience, speed and image, but not like the food
itself. This contest asks consumers to elaborate on how the food offerings could be made
more palatable to them. For example, one finalist was the McBrezel (“McPrezel”),
featuring leberkäse (“liver cheese”), a kind of Bavarian meatloaf that Germans like.
Participation also facilitates social interaction, since consumers must get their friends to
vote for their sandwich.
Testing the Goal-elaboration Hypothesis
Our central hypothesis is that cognitive elaboration about how a brand helps consumers
achieve goal(s) in their lives will create engagement and thus affect purchase behavior.
We test this hypothesis with data from the Air Miles Reward Program (AMRP), which
has been operating in Canada since 1992, and is not affiliated with Air Canada or any
other airline. As a coalition loyalty program, members collect miles from over 100
sponsors spanning nearly all purchase categories, including groceries, gasoline, apparel,
and credit card purchases. Members collect “miles” by swiping their AMRP card at the
time of purchase. Sponsors compensate AMRP on the basis of the number of miles issued
to members, so mile accumulation by members is proportional to AMRP’s revenue.
Collected miles can be exchanged for rewards such as travel (e.g., airline tickets, hotels),
merchandise (e.g., toasters, blenders) and gift cards (e.g., gas, movies, groceries, home
improvement). AMRP pays for the reward.
Activity on AMRP’s Social Media Page
AMRP maintained a social media website for members to discuss the program and its
benefits, which they discontinued in 2013. Posts made by members can be linked to their
mile accumulation, which is proportional to spending. Thus, this data set provides a
unique opportunity to measure the effect of participation in social media forums on actual
purchase behaviors with the brand. Figure 3 shows a histogram of the number of posts to
the AMRP social media site from October 1 through January 17, 2010. Posting to the site
is sporadic, with few posts on most days—the median number of posts per day for the
date range (Oct 1-Jan 17) is only 6. There are, however, large spikes in activity. The
maximum number of posts on a single day was 6,455, which occurred on 12/7/2010 and
was caused by an email sent on that day announcing what we call the winter contest.
Figure 3 about here
The winter contest invited members to “simply share with us what rewards you are
redeeming for this winter season and we’ll give you 10 bonus reward miles.” The contest
specifically asks respondents to specify and elaborate on a personal goal. There was a
limit of one reward per customer ID, and the offer was valid until 12/20/2010. Out of
11,740 total posts between December 7-20, 9,911 were in response to the winter contest.
These posts were made by 7,089 unique customer ID values that could be matched with
the transaction history, with 82% posting only one time.
We study the effects of the winter contest on mile accumulation during the six-week
period from 12/7/2010-1/17/2011. AMRP did not have any other UGC contests during
this period, nor in the two-month period before the contest, which reduces the risks of
confounds. Figure 3 shows that there was little activity on the site except for the winter
contest. We study the effects separately on each of the six week-long periods, indexed by
t, after the email. This will enable us to assess the immediate effect (e.g., week t=1) as
well as longer-term effects during subsequent weeks. The period 1/1-12/6/2010 is the pre
period (t=0), used to establish a baseline measure of mile accumulation per week and
account for customer heterogeneity.
Responses to the winter contest varied in length. Some wrote one or two words, e.g.,
“blender” or “digital camera.” One poster even wrote that he wanted “a new wife.”
Others wrote several sentences explaining why they wanted a particular reward, e.g.,
“haven’t seen my mother in several years and I want to get a ticket to visit her on her 80 th
birthday.” Summing words across posts, the average number of words written by each
participant was 17.2 and the quartiles were 8, 13 and 21 words.
Variable Operationalization and Study Design
AMRP provided the mile accumulation history for 141,308 members, which included all
promotion participants, a random sample of non-participants who had received the email
advertising the promotion, and a random sample of non-participants who had received no
information on the promotion. The vertical lines in Figure 3 demarcate seven study
periods. For each member i, we obtained the average number of miles per week
accumulated during the pre period (t=0) and each of the six post-email periods (t=1, …,
6). The number of miles accumulated is labeled yit. Note that Christmas occurs during
week 3 and New Year Eve during week 4. A total of 279,016 people who received the
email opened it. Elaboration is measured by the number of words written by a member.
Matching with Propensity Scoring Models to Control for Selection Bias
To evaluate the effects of entering the social media contest, we have before-after-with-
control-group quasi-experimental designs. Pre-measures (miles per week) account for
heterogeneity across customers, and the control group accounts for confounds in the
future periods such as history. While this design is robust to most threats to internal
validity, a potential problem with it is that members self-select into participating in the
contest. It turns out that those who elect to enter (treatment group) have systematically
higher levels of log mile accumulation during the pre-period than those who do not
(control group). It is not surprising that customers with higher purchase activity are more
likely to engage on AMRP’s social media site. Having a design with pre-measures of
mile accumulation addresses this selection bias to some extent, but the design can be
strengthened further through matching with propensity scoring models.
The goal is to identify a comparable control group that is as similar as possible to the
treatment group as of the time of the contest. Propensity scoring models achieve this by
predicting whether or not a member is in the treatment group from relevant member
measures known at the time of the contest using logistic regression. The predicted
probabilities of those in the treatment group are then matched with those not treated to
identify a “twin.” Details of our propensity scoring model can be found in Malthouse et
al. (2015). We used a wide variety of variables from the pre period including the level of
spending at different types of sponsors and the number of rewards earned.
Results
For each of the six future time periods, we regress yt (t=1, …, 6) on a dummy variable
indicating whether or not the member entered the contest, elaboration measured by the
total number of words written (0 for those who did not enter), and the total pre-period
miles to control for customer heterogeneity. We estimate the following multiplicative
model with least squares using the 7,089 who entered into the contest and their matched
controls, selected from the 37,350 who opened the email but did not participate, for a