Algorithmic Persuasion
in Online Adver sing
ti
, Online behavioral adver sing – Boerman et al., 2017
• De ni on: Online behavioral adver sing: Monitoring people’s online behavior and using
the informa on collected to show people individually targeted adver sements
• Example: an adver sing network (i.e., a company that serves adver sing on thousands
of websites) tracks a consumer’s website visits. If a consumer visits several websites
about cars, the network assumes the consumer is interested in cars
• collec ng, using, and sharing personal data ! raises consumer privacy concerns
• aims at personal relevance
• involves tailoring adver sing based on online behavior
Adver ser controlled factors
Ad characteris cs
• Level of personaliza on:
amount (one search term or combina on)
type of data used (browsing / search history
• Accuracy - When consumers know an ad is based on their past online behavior, they
understand that the marketer has made inferences about them
OBA Transparency
• Privacy statements and informed consent – rarely informing due to expert language use
& seldom read
• Disclosure
Consumer controlled factors
Knowledge and abili es
• Have li le knowledge into the extent to which their online behavior is tracked, even less
on privacy protec on
• Informa on asymmetry: Companies know much about consumers, yet consumers know
li le about what happens to their personal data
• Only minority of consumers try to control personal data by dele ng cookies
• Blocking third-party cookies can e ec vely limit OBA
• Opt-out op ons limit receiving behaviorally targeted ads but may not limit being tracked
Percep ons
• Percep ons are mixed: Relevant but creepy
2
ttfi tititt
titi titi ti titititi ti ffti ti ti ti ti ti
, • Privacy concerns & trust play important role in acceptance and e ec veness of OBA
• Privacy calculus: Response to OBA is outcome of weighing costs and bene ts – if its
relevant react more posi ve
• if consumers’ personal data are not transmi ed to third par es, consumers are less
concerned about their privacy and more sa s ed with the content of a smartphone app
about products
Consumer characteris cs
• Low level of privacy concern = more posi ve towards OBA
• E ects of OBA on purchase inten ons and behavior are more posi ve when ad ts
consumers need
• Responses are related to age, educa on and online experience
OBA Outcomes
Adver sing e ects
• level of personaliza on in OBA in uences click-through inten ons and click-through
rates
• companies covertly collect informa on, click-through rates are not in uenced or are
even reduced
• people seem to appreciate company transparency, and an icon can func on as a cue to
trust the adver ser and even posi vely a ect adver sing outcomes
• OBA were mediated by the consumers’ experience of vulnerability
• personaliza on appears to increase click- through rates and inten ons, but only when
consumers know that data are collected
• When consumers are unaware that data are collected, they feel more vulnerable when
confronted with personalized adver sing, which decreases their inten on to click on an
ad
• Personaliza on ! + Relevance ! + Click through rates
• Personaliza on ! + feelings of vulnerability ! - Click through rates
Purchases and purchase inten ons
• the e ects of the di erent ads, depended on the decision stage of the consumer, with
OBA being more e ec ve when consumers had narrowly construed preferences and
thus had a greater focus on speci c and detailed informa on. When preferences were
s ll broad and people were in the early stages of a purchase decision, generic ads led to
a higher likelihood of purchases
3
tiff ffti ti ffti fftiff titi ti ti tifl
fititititi tiff titt fi ti ti titi fftiti tiflti ti fi fi
, • more personaliza on increases feelings of intrusiveness and thus nega vely a ects
purchase inten ons
• people with high concerns had lower inten ons to purchase
OBA Acceptance and resistance
• privacy concerns and ad irrita on both increase ad skep cism, which consequently leads
to more avoidance of OBA
• reciprocity argument increased consumers’ acceptance of OBA compared to the
relevance argument, and consumers were more likely to opt in and disclose personal
data for OBA purposes. This nding suggests that consumers believe that receiving web
services for “free” in return for use of their personal data is an acceptable trade-o
• When people perceive a lack of freedom of choice, it could lead to heightened irrita on
and skep cism toward OBA, as well as more concerns about privacy
Privacy Paradox: people say they reject OBA, they take few measures to protect their data from
it
4
ti ti ti fiti ti ti ti ff ff ti
in Online Adver sing
ti
, Online behavioral adver sing – Boerman et al., 2017
• De ni on: Online behavioral adver sing: Monitoring people’s online behavior and using
the informa on collected to show people individually targeted adver sements
• Example: an adver sing network (i.e., a company that serves adver sing on thousands
of websites) tracks a consumer’s website visits. If a consumer visits several websites
about cars, the network assumes the consumer is interested in cars
• collec ng, using, and sharing personal data ! raises consumer privacy concerns
• aims at personal relevance
• involves tailoring adver sing based on online behavior
Adver ser controlled factors
Ad characteris cs
• Level of personaliza on:
amount (one search term or combina on)
type of data used (browsing / search history
• Accuracy - When consumers know an ad is based on their past online behavior, they
understand that the marketer has made inferences about them
OBA Transparency
• Privacy statements and informed consent – rarely informing due to expert language use
& seldom read
• Disclosure
Consumer controlled factors
Knowledge and abili es
• Have li le knowledge into the extent to which their online behavior is tracked, even less
on privacy protec on
• Informa on asymmetry: Companies know much about consumers, yet consumers know
li le about what happens to their personal data
• Only minority of consumers try to control personal data by dele ng cookies
• Blocking third-party cookies can e ec vely limit OBA
• Opt-out op ons limit receiving behaviorally targeted ads but may not limit being tracked
Percep ons
• Percep ons are mixed: Relevant but creepy
2
ttfi tititt
titi titi ti titititi ti ffti ti ti ti ti ti
, • Privacy concerns & trust play important role in acceptance and e ec veness of OBA
• Privacy calculus: Response to OBA is outcome of weighing costs and bene ts – if its
relevant react more posi ve
• if consumers’ personal data are not transmi ed to third par es, consumers are less
concerned about their privacy and more sa s ed with the content of a smartphone app
about products
Consumer characteris cs
• Low level of privacy concern = more posi ve towards OBA
• E ects of OBA on purchase inten ons and behavior are more posi ve when ad ts
consumers need
• Responses are related to age, educa on and online experience
OBA Outcomes
Adver sing e ects
• level of personaliza on in OBA in uences click-through inten ons and click-through
rates
• companies covertly collect informa on, click-through rates are not in uenced or are
even reduced
• people seem to appreciate company transparency, and an icon can func on as a cue to
trust the adver ser and even posi vely a ect adver sing outcomes
• OBA were mediated by the consumers’ experience of vulnerability
• personaliza on appears to increase click- through rates and inten ons, but only when
consumers know that data are collected
• When consumers are unaware that data are collected, they feel more vulnerable when
confronted with personalized adver sing, which decreases their inten on to click on an
ad
• Personaliza on ! + Relevance ! + Click through rates
• Personaliza on ! + feelings of vulnerability ! - Click through rates
Purchases and purchase inten ons
• the e ects of the di erent ads, depended on the decision stage of the consumer, with
OBA being more e ec ve when consumers had narrowly construed preferences and
thus had a greater focus on speci c and detailed informa on. When preferences were
s ll broad and people were in the early stages of a purchase decision, generic ads led to
a higher likelihood of purchases
3
tiff ffti ti ffti fftiff titi ti ti tifl
fititititi tiff titt fi ti ti titi fftiti tiflti ti fi fi
, • more personaliza on increases feelings of intrusiveness and thus nega vely a ects
purchase inten ons
• people with high concerns had lower inten ons to purchase
OBA Acceptance and resistance
• privacy concerns and ad irrita on both increase ad skep cism, which consequently leads
to more avoidance of OBA
• reciprocity argument increased consumers’ acceptance of OBA compared to the
relevance argument, and consumers were more likely to opt in and disclose personal
data for OBA purposes. This nding suggests that consumers believe that receiving web
services for “free” in return for use of their personal data is an acceptable trade-o
• When people perceive a lack of freedom of choice, it could lead to heightened irrita on
and skep cism toward OBA, as well as more concerns about privacy
Privacy Paradox: people say they reject OBA, they take few measures to protect their data from
it
4
ti ti ti fiti ti ti ti ff ff ti
