MEASURING THE EFFECT OF BRANDS AND CUSTOMER
REVIEWS ON UNCERTAINTY BY ELICITING CHOICE
PROBABILITIES
Relevance of Research
2
Consumers often face uncertainty in
purchase decisions, especially when adopting innovative products
Uncertainty arises from a lack of
information
If companies fail to provide
Research Objective
3
Investigating two marketing constructs on their
capability of reducing uncertainty for an innovative product
Brands: serve as quality signal (Erdem & Swait, 1998; Kirmani & Rao, 2000) or heuristic (Gigerenzer & Goldstein, 2011)
Customer reviews as social influence (Huang & Chen, 2006):
focus on valence and volume
Interaction: How does valence influence the effect of
Main Insights of
Literature Review (1/2)
4
Brands: reduce risk, engender trust and facilitate
choice
Brand credibility, reputation and prestige serve as quality
signals (Baek, Kim & Yu, 2010; Erdem & Swait, 2004; Herbig & Milewicz, 1993)
Brand recognition and brand familiarity serve as heuristics
(Campbell & Keller, 2003; Laroche, Kim & Zhou, 1996)
Customer reviews: one of the most popular information
source before purchase
Main Insights of
Literature Review (2/2)
5
Customer reviews: two metrics were chosen
Valence (average rating) and volume (number) of
reviews
Effects of valence and volume have led to mixed
findings so far (Kostyra et al., 2016; Ludwig et al., 2013)
Interaction between valence and brands:
research is limited and opposing views
Unclear whether stronger or weaker brands are more
affected by negative reviews
Chosen Innovation
iSkin
1 7 “…customizable touch sensor that can be worn directly on the
skin” (Weigel et al., 2015)
Categorized as technological breakthrough: new technology,
but low additional customer benefit (Chandy & Tellis, 1998)
Methodology (1/3)
Choice Probabilities & HB
8
It was opted for choice probabilities instead of
a traditional choice-based conjoint
+ Express uncertainty by differentiating between
strong and weak preferences
+ Gain information about all alternatives
- Dual-response option is not feasible
Hierarchical Bayes model was used instead of
Methodology (2/3)
Study Design
9
Two conjoint analyses in one survey
First five sets without customer reviews, followed by six sets with
reviews (including the attributes valence and volume)
Each set contained three alternatives and a no-choice option
Attributes and their levels selected in a way that they were
mutually exclusive and no number-of-levels effect could occur
Randomized design chosen: balanced overlap
Inclusion of control variables: gender, age, country of origin
Methodology (3/3)
Study Design
Results (1/3)
Conjoint Analysis
11
High relevance of no-choice option
18.4% of respondents distributed 100% in the
whole survey
Most preferred combination: Apple, music
player on the forearm, 4-star rating and 1000 reviews
Increase in choice share towards the no-choice
Results (2/3)
Hypotheses
12
Individual standard deviations were used from the CA (and
squared) to verify the hypotheses with paired t-tests
Hypothesis 1: variances decreased from Ink‘d to LG,
however increased again for the brand Apple
H1 was rejected
Hypothesis 2a & 2b: variances of product characteristics
increased for respondents in all last quartiles of valence and for the two lower levels of volume after the introduction of reviews
Results (3/3)
Hypotheses
13
Hypothesis 3: five out of nine pairs showed a
decrease in uncertainty with an increase in rating; effect was higher for fictional brand
H3 was partially supported
Differences in results could be observed if
Managerial Implications
14
48.21% increase in choice probability towards
no-choice option after introduction of reviews
Motivate customers to write reviews
Possible usage in offline setting
Reviews do not support in decreasing
uncertainty of product characteristics but change importance of attributes
Research Limitations &
Further Research Opportunities
15
Product chosen for the study is a prototype and not marketable
yet
Combination of music player on the forearm was shown twice in
the video
Other combinations were disadvantaged? Subconscious influence by repeated display?
Future research could address other metrics such as helpfulness or
Discussion
References (1/2)
17
Baek, T. H., Kim, J., & Yu, J. H. (2010). The differential roles of brand credibility and brand prestige
in consumer brand choice. Psychology & Marketing, 27(7), 662-678.
Campbell, M. C., & Keller, K. L. (2003). Brand familiarity and advertising repetition effects. Journal of Consumer Research, 30(2), 292-304.
Chandy, R. K., & Tellis, G. J. (1998). Organizing for Radical Product Innovation: The Overlooked Role
of Willingness to Cannibalize. Journal of Marketing Research, 35(4), 474-487.
Erdem, T., & Swait, J. (1998). Brand Equity as a Signaling Phenomenon. Journal of Consumer Psychology, 7(2), 131-157.
Erdem, T., & Swait, J. (2004). Brand credibility, brand consideration, and choice. Journal of Consumer Research, 31(1), 191-198.
Gigerenzer, G., & Goldstein, D. G. (2011). The recognition heuristic: A decade of research. Judgment and Decision Making, 6(1), 100-121.
Huang, J., & Chen, Y. (2006). Herding in online product choice. Psychology & Marketing, 23(5),
References (2/2)
18
Herbig, P., & Milewicz, J. (1993). The relationship of reputation and credibility to brand success. Journal of Consumer Marketing, 10(3), 18-24.
Kirmani, A., & Rao, A. R. (2000). No Pain, No Gain: A Critical Review of the Literature on Signaling
Unobservable Product Quality. Journal of Marketing, 64(2), 66-79.
Kostyra, D. S., Reiner, J., Natter, M., & Klapper, D. (2016). Decomposing the effects of online
customer reviews on brand, price, and product attributes. International Journal of Research in
Marketing, 33(1), 11-26.
Laroche, M., Kim, C., & Zhou, L. (1996). Brand familiarity and confidence as determinants of
purchase intention: An empirical test in a multiple brand context. Journal of Business Research, 37(2), 115-120.
Ludwig, S., de Ruyter, K., Friedman, M., Brüggen, E. C., Wetzels, M., & Pfann, G. (2013). More Than
Words: The Influence of Affective Content and Linguistic Style Matches in Online Reviews on Conversion Rates. Journal of Marketing, 77(1), 87-103.
Weigel, M., Lu, T., Bailly, G., Oulasvirta, A., Majidi, C., & Steimle, J. (2015). Iskin: flexible, stretchable
19
Demographic Descriptives
20
250 respondents completed the survey
66% females and 34% males
Average age: 30.56 years
Country of origin: Germany (72.4%), Netherlands (6%) and
others (21.6%), half of them coming from the USA
Level of highest education: main school/secondary school
Attribute Importance
21
Attributes Conjoint 1 (without reviews) Conjoint 2 (with customer reviews)2
Brand 32.05% 20.68% (43.82%)3 Function 39.11% 14.51% (30.75%) Placement 28.84% 12.00% (25.43%) Valence n.a. 30.94% Volume n.a. 21.86% Sum: 100% 100%
2 Does not add up to 100% due to rounding
Output H1
22
Pairs Mean t-value Df Sig.
(2-tailed) Supporting H1?4 Ink’d & LG 0.01189 2.136 249 0.034 (+) LG & Apple -0.31115 -15.350 249 0.000 (-) Apple & Ink’d 0.29926 16.154 249 0.000 (-)
4 (-) signifies that the value was either not significant or significant with the opposite sign
Output H2b
23
Brand Function Placement
# n Ink’d LG Apple Telephone Music Message Finger Hand Forearm 10 62 8 6 11 6 13 12 13 7 12 100 62 12 3 18 7 22 19 19 12 19 1000 62 37 9 32 15 31 41 34 24 32
Number of respondents in the fourth quartiles of volume who decreased their
Output H3
Valence as linear parameter
24
Pairs Mean t-value df Sig. (2-tailed) Supporting
H3?5
Ink’d x Val & LG x Val -0.11196 -8.007 249 0.000 (+)
LG x Val & Apple x Val 0.09563 8.193 249 0.000 (-)
Apple x Val & Ink’d x Val 0.01633 1.439 249 0.152 (-)
5 (-) signifies that the value was either not significant or significant with the opposite sign
Output H3
Valence as part-worth
25
Pairs Mean t- value df Sig. (2-tailed) Supporting H3? Ink'd x 2.2 and Ink'd x 3.1 0.25969 8.798 249 0.000 (+)
Ink'd x 3.1 and Ink'd x 4.0 0.02881 1.653 249 0.099 (-) Ink'd x 4.0 and Ink'd x 2.2 -028851 -14,108 249 0.000 (+)
CA Output with Aggregate Values
26
