2.3_Steiner et al. (2016) - Platform Adoption in
System Markets.pdf
Steiner, Wiegand, Eggert & Backhaus (2016)
“Platform Adoption in System Markets: The Roles of Preference Heterogeneity and
Consumer Expectations”
The article investigates how consumers adopt platforms in system markets (e.g., video game
consoles), where the value of the product depends heavily on direct network effects (being
able to interact with other users) and indirect network effects (availability of high-quality
complementary goods, like game titles). Traditional research often measures these effects
using aggregate sales data, assuming consumers are homogeneous and either ignore
expectations or have perfect foresight about future network effects.
The authors challenge these assumptions by conducting two empirical studies with
individual-level survey data:
Study 1: Heterogeneity in Preferences (Choice-Based Conjoint)
● They measure how different attributes affect adoption intent: software variety,
software quality, availability of a favorite title, local installed base, software price,
hardware price.
● Using latent-class analysis, they find four distinct gamer segments with strongly
differing network-effect sensitivities:
1. Casual Gamers (44%)
weak preferences, moderately like variety/quality/price.
2. Social Gamers (24%)
care most about the local installed base (friends they can play with).
3. Habitual Gamers (15%)
mostly driven by whether their favorite title is available.
4. Hardcore Gamers (15%)
dominated by software quality and high involvement.
Study 2:The Role of Consumer Expectations (Experimental SEM model)
● Based on Zeithaml’s value model, they examine how expected direct and indirect
network effects influence adoption intentions.
● They test how perceived system value → expected installed base →
(expected direct & indirect effects) → adoption intention.
● Main finding: Expectations matter, but far less than previous theory claims.
○ Expected indirect network effects matter mainly for hardcore and habitual
gamers. Expected direct network effects matter almost exclusively for social
gamers.
○ For all groups, current system value (software quality, variety, and prices) is
the strongest driver of adoption.
Managerial insights
, ● Firms should not rely on expectations alone to drive early adoption, consumers don’t
trust or use expectations as much as theory suggests.
● Early targeting should focus on segments that do rely on expectations (hardcore +
habitual gamers).
● Offering high-quality flagship games early can be more effective than offering a large
but mediocre variety
Guiding questions
Theory
1. In which way does the article extend our understanding of network effects theory?
What is new?
The article extends network-effects theory in three important ways:
1. Individual-level evidence:
Prior research mostly uses aggregate market data. This article uses individual survey
+ conjoint + experimental data to capture how single consumers perceive network
effects. This reveals substantial heterogeneity not visible in aggregate models.
2. Joint examination of direct and indirect effects:
Prior work focused mostly on indirect effects (software availability). This article
measures both direct and indirect effects at the individual level and shows that direct
effects matter only for some segments.
3. Empirical measurement of expectations:
While theory emphasizes consumer expectations, no prior study directly measured
how expected direct/indirect network effects influence adoption. This article shows
that:
○ expectations matter much less than theory suggests,
○ their importance differs strongly across segments.
2. What are typical outcomes of competition in markets with strong network effects?
Under which conditions do which outcomes emerge?
Typical outcomes of system markets with network effects include:
● Tipping and dominance: One platform can become dominant if complements
(games) accumulate rapidly after reaching critical mass.
● Oligopolies: The game console market often doesn’t tip into one winner because
consumer preferences are highly heterogeneous, creating room for multiple platforms
(e.g., habitual gamers vs social gamers).
● Positive feedback loops: Larger installed bases attract more developers →
more and better games → increased attractiveness → more adopters.
Conditions:
● Homogeneous preferences + strong indirect network effects → market likely
tips to a single winner.
● Heterogeneous preferences (as in gaming) → multiple platforms can coexist
(“system wars”).
, ● High uncertainty about expectations → consumers rely less on expected
effects, reducing tipping.
Consumer Segmentation
3. How does the research carve out the different consumer segments? What method
is used? How does it roughly work?
The segmentation arises from Study 1, using:
● Choice-Based Conjoint Analysis (CBC) to estimate individual preferences for
platform attributes.
● Latent Class Analysis (LCA) to identify clusters of consumers with similar
preference patterns.
How it works:
1. Respondents choose preferred console profiles (varying quality, variety, prices, etc.).
2. CBC estimates attribute importance for each respondent.
3. LCA groups respondents with similar part-worth utilities.
4. Result = four distinct gamer segments.
4. Which variables in Study 1 reflect the indirect and direct network effects
dimensions?
Indirect network effects (complements):
● Software variety
● Software quality
● Availability of a favorite title
Direct network effects:
● Local installed base (share of friends/acquaintances who use the same console)
5. What are the results of the segmentation study? What segments exist and how do
they differ?
The four segments are:
1. Casual Gamers (44%)
● Low involvement, low expertise
● Moderately influenced by software features
● Little sensitivity to favorite titles or direct network effects
● Weak attitudes toward gaming
System Markets.pdf
Steiner, Wiegand, Eggert & Backhaus (2016)
“Platform Adoption in System Markets: The Roles of Preference Heterogeneity and
Consumer Expectations”
The article investigates how consumers adopt platforms in system markets (e.g., video game
consoles), where the value of the product depends heavily on direct network effects (being
able to interact with other users) and indirect network effects (availability of high-quality
complementary goods, like game titles). Traditional research often measures these effects
using aggregate sales data, assuming consumers are homogeneous and either ignore
expectations or have perfect foresight about future network effects.
The authors challenge these assumptions by conducting two empirical studies with
individual-level survey data:
Study 1: Heterogeneity in Preferences (Choice-Based Conjoint)
● They measure how different attributes affect adoption intent: software variety,
software quality, availability of a favorite title, local installed base, software price,
hardware price.
● Using latent-class analysis, they find four distinct gamer segments with strongly
differing network-effect sensitivities:
1. Casual Gamers (44%)
weak preferences, moderately like variety/quality/price.
2. Social Gamers (24%)
care most about the local installed base (friends they can play with).
3. Habitual Gamers (15%)
mostly driven by whether their favorite title is available.
4. Hardcore Gamers (15%)
dominated by software quality and high involvement.
Study 2:The Role of Consumer Expectations (Experimental SEM model)
● Based on Zeithaml’s value model, they examine how expected direct and indirect
network effects influence adoption intentions.
● They test how perceived system value → expected installed base →
(expected direct & indirect effects) → adoption intention.
● Main finding: Expectations matter, but far less than previous theory claims.
○ Expected indirect network effects matter mainly for hardcore and habitual
gamers. Expected direct network effects matter almost exclusively for social
gamers.
○ For all groups, current system value (software quality, variety, and prices) is
the strongest driver of adoption.
Managerial insights
, ● Firms should not rely on expectations alone to drive early adoption, consumers don’t
trust or use expectations as much as theory suggests.
● Early targeting should focus on segments that do rely on expectations (hardcore +
habitual gamers).
● Offering high-quality flagship games early can be more effective than offering a large
but mediocre variety
Guiding questions
Theory
1. In which way does the article extend our understanding of network effects theory?
What is new?
The article extends network-effects theory in three important ways:
1. Individual-level evidence:
Prior research mostly uses aggregate market data. This article uses individual survey
+ conjoint + experimental data to capture how single consumers perceive network
effects. This reveals substantial heterogeneity not visible in aggregate models.
2. Joint examination of direct and indirect effects:
Prior work focused mostly on indirect effects (software availability). This article
measures both direct and indirect effects at the individual level and shows that direct
effects matter only for some segments.
3. Empirical measurement of expectations:
While theory emphasizes consumer expectations, no prior study directly measured
how expected direct/indirect network effects influence adoption. This article shows
that:
○ expectations matter much less than theory suggests,
○ their importance differs strongly across segments.
2. What are typical outcomes of competition in markets with strong network effects?
Under which conditions do which outcomes emerge?
Typical outcomes of system markets with network effects include:
● Tipping and dominance: One platform can become dominant if complements
(games) accumulate rapidly after reaching critical mass.
● Oligopolies: The game console market often doesn’t tip into one winner because
consumer preferences are highly heterogeneous, creating room for multiple platforms
(e.g., habitual gamers vs social gamers).
● Positive feedback loops: Larger installed bases attract more developers →
more and better games → increased attractiveness → more adopters.
Conditions:
● Homogeneous preferences + strong indirect network effects → market likely
tips to a single winner.
● Heterogeneous preferences (as in gaming) → multiple platforms can coexist
(“system wars”).
, ● High uncertainty about expectations → consumers rely less on expected
effects, reducing tipping.
Consumer Segmentation
3. How does the research carve out the different consumer segments? What method
is used? How does it roughly work?
The segmentation arises from Study 1, using:
● Choice-Based Conjoint Analysis (CBC) to estimate individual preferences for
platform attributes.
● Latent Class Analysis (LCA) to identify clusters of consumers with similar
preference patterns.
How it works:
1. Respondents choose preferred console profiles (varying quality, variety, prices, etc.).
2. CBC estimates attribute importance for each respondent.
3. LCA groups respondents with similar part-worth utilities.
4. Result = four distinct gamer segments.
4. Which variables in Study 1 reflect the indirect and direct network effects
dimensions?
Indirect network effects (complements):
● Software variety
● Software quality
● Availability of a favorite title
Direct network effects:
● Local installed base (share of friends/acquaintances who use the same console)
5. What are the results of the segmentation study? What segments exist and how do
they differ?
The four segments are:
1. Casual Gamers (44%)
● Low involvement, low expertise
● Moderately influenced by software features
● Little sensitivity to favorite titles or direct network effects
● Weak attitudes toward gaming
