Of chameleons and consumption:
the impact of mimicry on choice and preferences
Investigation of: effect of mimicry on consumer product consumption and appraisal
 We test 2 paths via which mimicry may influence product preferences:
1. Mimicking consumer path: individuals mimic consumption behaviour (automatically)
 Such mimicry effects preferences toward product(s) consumed
2. Mimicked consumer path: being mimicked leads to increased prosociality
 Affects preferences for products in dyadic interactions
INTRODUCTION
Behavioral mimicry occurs automatically + can influence actions and attitudes toward others
 Has the potential to influence choices made in social contexts
 Aim of current research: explore potential ways in which consumers’ consumption
and preferences might be influenced by behavioral mimicry.
(examine consequences for mimicker (automatically mimicking others) and the
mimicked (being mimicked by others)
THEORETICAL DEVELOPMENT
Automatic Behavioral mimicry of others
Generally accepted: individuals automatically mimic many aspects of their interaction partners (also
complete strangers)
e.g. Babies (1 month) already mimic facial expressions
e.g. interaction partners mimic accents, rate and rhythm of speech
Proven in experiments: mimicry appeared to be automatic and nonconscious
 Chronic tendency to dynamically alter behavioral mannerisms to blend in with the prevailing
social surroundings
 Called “chameleon effect”
Almost all research to date has focused on the mimicry of behaviours that occur in the context of
social interactions
 Less attention: mimicry of behaviours occurring outside of direct social interactions (e.g. 2
consumers browsing independently in the same aisle in a store)
 Important to explore extent to which automatic behavioral mimicry extends to
imitable consumption-orientated behaviours that occur outside of context of direct
interactions.
Consequences of automatically mimicking others
If a mimicking individual consumes something, she may construct her preferences to be in line with
her behaviour (regardless of whether she is aware of the underlying motivation for that
consumption)
+
To the extent that preferences are positively (negatively) changed as a result of engaging in such
mimicry, an individual could be more (less) likely to choose the item in the future.
 Automatic mimicry of the consumption behaviours of others could be an important
nonconscious source of unintentional preference and behaviour shifts
This leaves several unanswered questions for marketing:
a. Will the effect of mimicry on consumption carry over to downstream preferences?
b. If an individual observes another person consuming something but is without means to
mimic that consumption (e.g. seeing someone eating mints but you don’t have any yourself),
will preferences be changed to the same extent?
Consequences of being mimicked
Being mimicked:
-
creates feelings of rapport, affiliation and closeness toward the mimicker
-
has been shown to engender helping towards others
 creates prosocial behaviour
By fostering communication and rapport
 mimicry provides social glue
 helping to bind social groups together and create harmonious relationships
Mimicry of mannerisms and postures: might lead to more liking and rapport between individuals.
(see study: p. 3 of paper, paragraph 2: Chartrand and Barght (1999, study 2) tested … )
Recent research:
Explore whether positive prosocial effects of mimicry influence behaviour towards others:
Van Baaren et al. showed that verbal mimicry could influence customer tipping
behaviour in a restaurant environment
+ Van Baaren et al. demonstrated that participants who had previously been
mimicked by an experimenter picked up more pens dropped by the experimenter
than did the non-mimicked participants.
(see paper p. 3, paragraph 3)
BUT: increased prosociality has consequences beyond helping per se
e.g. it may have implications for persuasion
(motor movements can influence persuasiveness and product evaluations outside of
conscious awareness)
study: participants who were mimicked subsequently reported higher levels
of agreement with the message
Research overview
Explore the potential for mimicry to influence product consumption and appraisal
Particularly: investigate two distinct paths via which this influence may manifest itself
a. Consumer may mimic another individual
Relies on:
Consumer’s automatic mimicry of observed consumption behaviours.
Condition:
Consumer must have opportunity to mimic interaction partner (must
have equivalent access to the consumed product)
Hypothesis:
Increased consumption of the snack selected by the interaction
partner will influence preferences of mimicking consumer.
b. Consumer may be mimicked by another individual
Relies on:
Prosocial emotions (feelings of rapport and liking) being generated in
a consumer when he or she is mimicked by an interaction partner.
Hypothesis:
-
A consumer would display greater liking for a product that was
introduced by a mimicker than a non-mimicker.
-
Effect of mimicry should be stronger when the mimic’s need is
made more transparent.
EXPERIMENT 1
Exp. 1 examined mimicry of consumption behaviour and its subsequent effect on the mimicker’s
preferences.
Participants:
Story: Participants believed they were engaging in a study concerning memory for
advertisements. They had to watch a video of other participants (actually a
confederate) who describes a series of advertisements and they would be asked
about their memory for and impressions of those ads.
Exp.:
Confederate ate only one of two available snacks during experiment.
Hypothesis:
a. Participants would mimic the snacking behaviour of confederate
b. Preference ratings provided during an ostensibly unrelated second study would be
consistent with participants’ snacking behaviour
c. The effect of the confederate’s snacking behaviour on participants’ preferences would be
mediated by participants’ mimicry behaviour. Participants’ snacking behaviour, which is
the direct outcome of mimicry, will shift preferences.
Possible alternative explanation:
Merely observing someone’s consumption behaviour affects preferences toward the
consumption objects, which then determine behaviour

Not important whether mimicry occurs
This explanation was ruled out! (paper p. 4, paragraph 3 of title ‘experiment 1’)
METHOD
Participants
147 undergraduates from Duke university. Data from 113 participants was eventually used.
Materials
2 x 2 design: confederate’s snacking behaviour (goldfish vs. animal crackers) x food presence
(food vs. no-food on table in front of participant)
Prior preferences
Survey 3 days before experiment to indicate on 9-point scale how much they liked variety of
items (including goldfish and animal crackers)
Procedure
Story: participants were told they would complete two unrelated studies:
a. Examining the effects of hearing ad descriptions on ad memory
b. Examining personality and product preferences
They were told that students recruited earlier had been videotaped describing a
series of ads and that they would watch the video of one of these participants (the
confederate).
Just prior to playing the first ad, the experimenter (on the videotape) mentioned the snacks (that
were placed in front of the confederate) to the confederate in passing.
Food condition: confederate was told that he could help himself to the snacks and water at
any point.
No-food cond.:
the confederate was told that the snacks were left over from a prior study
and that he could help himself to the snacks and water at any point (this
was done so that participants in the no-food conditions would not wonder
why they themselves did not have any snacks)
 Participants in the food condition were given the same instruction regarding the
snacks as the confederate.
During the task, the confederate exclusively ate either goldfish or animal crackers. After the
video, participant was moved into a different room to complete the ostensibly unrelated second
study (rating how much they liked 30 different snacks)
Measures
Trainer coder recorded number of instances in which the participant took at least one snack
from either bowl. (Snack selection measure was percentage of times goldfish crackers were
selected)
 Mimicry: if participant took more goldfish crackers (> 50% of time) in goldfish-only
condition and vice versa.
Snack preference ratings were measured on a nine-point scale. The preference measure was
calculated as the difference between the goldfish and animal cracker ratings.
RESULTS AND DISCUSSION
Prior preferences
Goldfish was rated significantly higher than animal crackers (M’s = 6,33 and 5,67 respectively, p
= .01). Important: prior preferences did not differ across snacking conditions or food presence.
Main analysis
 Participants in goldfish-cracker-only condition selected goldfish more often and
participants in the animal-cracker-only condition selected goldfish crackers less often
(M’s= 71% and 44% respectively)
 The greater the positive value, the higher the preference rating for goldfish crackers relative
to animal crackers.
Food condition:
Goldfish-cracker only: rated goldfish higher than animal crackers (M(difference) = 1.61)
Animal cracker only: rated goldfish about the same as animal crackers (M(difference) = .21)
No-food condition:
Participants in both goldfish cracker-only and animal cracker-only conditions rated goldfish
higher than animal crackers and at the same levels (M(difference)’s= 1.23)
Significant relationship between
(a) Snacking condition and participant’s snack selection
(b) Snacking condition and preferences
(c) Participant’s snack selection and preferences
Not significant:
(d) Between snacking condition and preferences when controlling for participant’s snack
selection
 Combined with the lack of effect of snacking condition on preferences in the no-food
conditions, these results support the position that mimicry of consumption behaviour
led to adjustments in preferences.
Participants were asked to describe how they decided what to eat:
1. Properties of food: 45%
2. Pre-existing preferences: 64%
3. Behaviour of confederate: 11%
Participants were asked to indicate the extent to which an aspect influenced their cracker preference
(1 (not at all) to 9 (to a large extent)):
1. Eating behaviour:
No-food condition (past eating behaviour): mean response: 6,55
Food condition (during study session): 3,95
2. Prior preferences: mean response: 6,91
3. Confederate’s behaviour: mean response: 1,66
Findings:
-
When food present: participants tended to mimic confederate’s consumption
-
Participants’ preference ratings reflected their snack selection
-
Participants were not aware that confederate’s behaviour had influence on own snack
choice/preferences
-
The mere observation of snack choice did not differentially affect preferences
 Evidence that automatic mimicry of others can influence preferences of the person
engaging in mimicry
EXPERIMENT 2
Focus: mimicked consumer path  explores downstream effects of being mimicked by others
METHOD
Participants
39 participants from Duke university.
Mimic or no-mimic condition.
Data used of 37 participants.
Procedure
Story: Experiment concerned the impression formation process for new products and a trained
facilitator would be questioning them about their soft drink preferences and explaining some
features of a new sports drink called Vigor that was approaching market launch.
1.
Facilitator asked preferences questions about opinion soft drinks
2.
He explained 3 beneficial features of new sports drink
3.
Explanation of each feature followed by single related question
4.
Participants taste drink (“taste was finalised but colouring not yet”)
5.
The facilitator handed a feedback packet about the drink and left room
Mimic condition:
facilitator imitated body posture of participant + verbal mimicry was carried out
via the facilitator repeating back the key elements (using same syntax)
No-mimic condition:
facilitator anti-mimicked participant (e.g. if participant crossed legs, he kept
both feet flat on floor)  this ensured facilitator wouldn’t unintentionally
mimic participant + the facilitator used general confirmatory phrases to
respond to participant responses
 This condition was designed only to ensure an absence of automatic mimicry rather
than to create a situation where the participants felt excessively out of synch with
facilitator
Measures
Survey at end: asked participants to rate their agreement with following statements (0-10 scale):
-
“I really enjoyed taste of Vigor”
-
“I would buy Vigor when it goes on sale”
-
“I would expect Vigor to be successful when it is launched”
+ after participants finished survey, they recorded number of grams of beverage consumed to
provide a direct behavioural measure and an implicit measure of liking.
RESULTS AND DISCUSSION
Overall effect of mimicry across four dependent measures:
-
Enjoyment
-
Likelihood to buy
-
Expectations of success
-
Weight consumed
Independent variable: mimicry (mimic vs. no-mimic)
 Participants more positive toward vigor when the facilitator mimicked them than
when he did not.
 Initial evidence that mimicry can lead to more favourable attitudes toward a
product presented by the mimicker
EXPERIMENT 3
We argued:
Effect of exp. 2 is driven by mimicry-induced prosociality influencing the behaviour of
mimicked participants, such that those who are mimicked will respond more positively to a
product associated with the mimicker.
BUT: exp. 2 did not directly manipulate the degree to which facilitator expressed needed help
 We expect any prosocialty-driven effect to be more acute for a mimic who more
clearly exhibits a need.
 Exp. 3: Facilitator states:
a. He will directly benefit from the success of the product or marketing messages
b. He is not directly associated with the manufacturer of the product
2 (mimicry: mimic vs. no mimic) x 2 (facilitator need: invested vs. independent) X 2 (gender: m vs. f)
between-subject design.
METHOD
Participants
57 undergraduates (Duke university)  data from 52 participants used
Procedure
Story: a trained facilitator would be questioning them about snack product preferences and
soliciting their impressions of various marketing messages that were under consideration to
market chees straws in a potential national launch.
1.
Facilitator read 3 promotional messages for cheese straws
2.
Participants could taste cheese straw  placed full cup of cheese straws on table
and invited them to have a taste
3.
Further series of questions about flavour and texture
4.
Participants got feed back packet and invited them to take as many straws as they
wished.
Mimic condition + no-mimic condition:
see exp. 2
Facilitator need manipulation:
Invested:
the confederate told participants that he got more money from cheese straw
company if they believe marketing messages are better
Independent: the confederate told participants he had nothing to do with the company and
the cheese straws were chosen randomly
Measures
See exp. 2 + participants rated their mood on an 11-point scale (really bad – really good) +
manipulation check
RESULTS AND DISCUSSION
Manipulation check:
participants believed facilitator had more self-interest in invested condition
BUT: not observe any differences in his apparent motivation
Dependent measures:
-
Enjoyment
-
Likelihood to buy
-
Expectations of success
-
Weight eaten
Independent variables:
-
Gender: main effect due to male participants eating more
-
Mimicry: marginally significant effect across four dependent measures
 Effect was more acute when participants believed facilitator was invested
-
Facilitator
 First demonstration that positive effect of mimicry on attitudes are enhanced when
the mimicker is openly invested in the product in question
EXP 2 + 3:
Participants who were behaviourally mimicked
 Displayed stronger preferences than did participants not mimicked
 Effect enhanced when mimicking facilitator was invested (prosocial explanation)
GENERAL DISCUSSION
Consumption + preferences: influenced by behavioral mimicry
 Found support for both mimicked consumer path and mimicking consumer path
EXP. 1: preferences can be influenced by automatic mimicry of consumption of participants of
people they observe (only if consumers had chance to mimic consumption, not when only
observed) (= mimicking consumer path)
+ behavioral mimicry is a subtle mechanism (did not interact with confederate, just
observed)
EXP. 2: mimicked participants: more positive attitudes towards product
EXP. 3: Effect of mimicry particularly acute when mimicking facilitator was perceived as invested
(= mimicked consumer path)
+ mimicked consumers liked product more, expressed higher intent to purchase, …
 Mimicry has potential to be valuable tool in interpersonal persuasion and perhaps
particularly, in those cases where underlying motivation ans persuasive intent of the
persuader are transparent to target individual whom he or she is desirous of
persuading (much promise in sales domain)
Theoretical contribution
Current research extends our existing knowledge of mimicry in a number of important ways:
1. Focus on mimicking others AND on being mimicked
2. Effect of mimicry extends beyond explicitly social outcomes to encompass influences on
product preferences and choice
3. Greater transparency of facilitator need leads to larger effect of mimicry
Implications for marketing
Consumer behaviour can be driven by processes that occur outside of awareness, intent and control.
Potential roadblocks to mimicry’s relevance in consumer settings:
-
Experiments in controlled laboratory environments
BUT: many retail interactions may be of too brief a duration and/or involve style of
interaction that would render less applicable the mimicking techniques used in
current research.
(some situations are very much like in exp.: e.g. buying a car from a salesperson
sitting in front of you at a desk)
-
Use of deliberate mimicry  risk of being noticed by customer  may be upset
-
Studies do not speak to any possible moderation of effect by the various positive or negative
characteristics of the mimicker
Limitations and Future directions
Further questions for future research:
-
Facilitators used combination of verbal and physical mimicry
 Future research: tease out any differences in effect caused by the two mimicry types
-
Exp. 1: automatic mimicry can unintentionally shift one’s own preferences
 Potential for future research: e.g. mimicry may serve as a means of stabilizing and
reinforcing group thinking and behaviour
-
Could mimicry occur if observer’s prior preferences were very low (to the snack offered) (did
a little research and this suggest that it would still work) or if the item tasted bad (speculate
that this would significantly lower effects of mimicry)?
-
Would beneficial effects of mimicry on product appraisal transfer to products that are not
referenced to by the mimicker yet are present during mimicry?
+ Would a product
introduced by a different individual immediately after mimicry be more favourably regarded?
(recent research suggests: yes)
CONCLUSION
Behavioral mimicry can affect consumption behaviour of both mimicker and mimicked.
Automatic mimicry can influence behaviour outside of awareness.