This document discusses social desirability effects in market research and how digital and social data can provide additional predictive value compared to traditional surveys. It presents two case studies that show digital behaviors are better predictors of purchase outcomes for younger audiences in movie box office sales and automotive sales compared to surveys, which may be impacted by social desirability bias. The conclusions suggest digital and social data provide more predictive insights for brands and situations where self-monitoring of responses is likely due to social norms.
7. ADDRESSING THE REPRESENTATIVENESS OF SOCIAL MEDIA
While social is not
representative of
the population, the
population is
represented on
social
of people in
developed markets
have active accounts
on their country’s most
popular social
platform (We Are
Social)
of Americans 65+
report using social
media on a regular
basis
Social media
provides insight into
younger audiences,
which have been
more difficult to
understand in the past
8. ADDRESSING THE REPRESENTATIVENESS OF SOCIAL MEDIA
While social is not
representative of
the population, the
population is
represented on
social
of people in
developed markets
have active accounts
on their country’s most
popular social
platform (We Are
Social)
of Americans 65+
report using social
media on a regular
basis
Social media
provides insight into
younger audiences,
which have been
more difficult to
understand in the past
9. ADDRESSING THE REPRESENTATIVENESS OF SOCIAL MEDIA
While social is not
representative of
the population, the
population is
represented on
social
of people in
developed markets
have active accounts
on their country’s most
popular social
platform (We Are
Social)
of Americans 65+
report using social
media on a regular
basis
Social media
provides insight into
younger audiences,
which have been
more difficult to
understand in the past
10. ADDRESSING THE REPRESENTATIVENESS OF SOCIAL MEDIA
While social is not
representative of
the population, the
population is
represented on
social
of people in
developed markets
have active accounts
on their country’s most
popular social
platform (We Are
Social)
of Americans 65+
report using social
media on a regular
basis
Social media
provides insight into
younger audiences,
which have been
more difficult to
understand in the past
11. ADDRESSING THE REPRESENTATIVENESS OF SOCIAL MEDIA
While social is not
representative of
the population, the
population is
represented on
social
of people in
developed markets
have active accounts
on their country’s most
popular social
platform (We Are
Social)
of Americans 65+
report using social
media on a regular
basis
Social media
provides insight into
younger audiences,
which have been
more difficult to
understand in the past
12. PEOPLE POST A RANGE OF POSITIVE & NEGATIVE EXPERIENCES ONLINE
23. SOCIAL AND DIGITAL DATA IS SOMEWHAT PREDICTIVE OF…
BOX OFFICE RECEIPTS
(Asur & Huberman, 2010;
Noseworthy, 2013; Panaligan
& Chen, 2013)
24. SOCIAL AND DIGITAL DATA IS SOMEWHAT PREDICTIVE OF…
ELECTORAL OUTCOMES
(DiGrazia et. al., 2013; Maldonado
& Sierra, 2016; O’Connor et. al.,
2010; Tumasjan et. al., 2010)
25. SOCIAL AND DIGITAL DATA IS SOMEWHAT PREDICTIVE OF…
INFLUENZA OUTBREAKS
(Ginsberg & Mohebbi, 2009)
26. SOCIAL AND DIGITAL DATA IS SOMEWHAT PREDICTIVE OF…
ECONOMIC INDICES
(Choi & Varian, 2012)
27. SOCIAL DESIRABILITY MEANS DIGITAL AND SOCIAL DATA MAY
PROVIDE ADDITIONAL VALUE FOR YOUNGER AUDIENCES
28. SOCIAL DESIRABILITY MEANS DIGITAL AND SOCIAL DATA MAY
PROVIDE ADDITIONAL VALUE FOR YOUNGER AUDIENCES
Young & McCoy, 2015
29. SOCIAL DESIRABILITY MEANS DIGITAL AND SOCIAL DATA MAY
PROVIDE ADDITIONAL VALUE FOR YOUNGER AUDIENCES
Pearson, Henryks & Jones, 2011
Young & McCoy, 2015
30. SOCIAL DESIRABILITY MEANS DIGITAL AND SOCIAL DATA MAY
PROVIDE ADDITIONAL VALUE FOR YOUNGER AUDIENCES
Pearson, Henryks & Jones, 2011
Reifman, Klein & Murphy, 1989
Visser & Krosnick, 1998
Tesch, 1983
Young & McCoy, 2015
39. PREDICTIVE VALUE OF SOCIAL AND DIGITAL METRICS ACROSS
MOVIE-GOING AGE BANDS
Digital is a better predictor of purchase behavior among those 18-44
39%
28% 28%
35%
54%
37%
55%
49%
41%
19%
13 - 17 18 - 24 25 - 34 35 - 44 45 - 54
Digital Metrics
Non-Digital Metrics
PREDICTIVE(R2)CONTRIBUTION
40. CASE STUDY #2: AGE EFFECTS IN
PREDICTIONS FOR AUTOMOTIVE SALES
51. SITUATIONS WHERE SELF-
MONITORING IS LIKELY
BRANDS THAT FALL INTO
LOWER PRICE TIERS
SITUATIONS WHERE DIGITAL AND SOCIAL CAN ADD ADDITIONAL
PREDICTIVE VALUE