Social Network Analysis and Marketing

1
Social Network Analysis and Marketing

• ANZMAC Doctoral Colloquium Skills
  Workshop
• Dr. Cynthia Webster
• 29 December 2003

2
Standard Marketing Approach

• concentrates on attributes of actors/relations
• type of firm/industry, frequency of interaction,
  trust
• aggregates individuals into categories
• destroying structural linkages
• retaining information about internalised norms
• expects category members to behave alike
• focuses on correlates within a social category

3
Network Approach

• reveals actors differential access to such
  resources as
• information, wealth and power
• highlights how patterns of ties provide
• opportunities as well as constraints
• provides alternative explanations
• does not treat social processes as the sum
• of attributes internalised norms

4
Social Network Analysis

• interdisciplinary
• pioneers in sociology social psychology
  (Moreno, Cartwright, Newcomb, Bavelas)
• and anthropology (Barnes, Mitchell)
• provides formal measures of social structure
• explicit mathematical definitions not metaphors
• not simply a collection of analytic procedures
• alternative approach to assuming independence
• grounded in social processes and concepts
• diffusion, centrality, group, social role

5
How Networks are Used in
Marketing

• theory acknowledged
• exchange, embeddedness interdependence
  (Cook 1982 Emerson 1981 Granovetter 1985)
• roles positions (Burt 1976 1992)
• diffusion of innovations (Rogers 1962 1979)
• theory development
• Modelling Networks (Iacobucci Hopkins 1992)
• Networks as a Marketing Strategy
• (Arabie Wind 1994)
• Consumer Networks (Martin Clark 1996)
• Strategic Alliances (Hakansson Sharma 1996)

6
How Networks are Used in
Marketing

• little empirical work conducted
• Brand-switching Behavior
• (Iacobucci, Henderson, Marcati Chang 1996)
• Business Network Structure Density
• (Cadeaux 1997)
• limited structural analyses
• actors, activities, resources described
• graphs used for illustrative purposes
• focal actors perspective
• undirected binary relations

7
Why are Networks Important?

• Customers can hardly hear you
• but they listen to their friends
• Customers are skeptical
• but they trust their friends
• Customers share information
• and now it is easier with e-mail and the internet

8
What are Networks ?

• Networks consist of actors connected to one
  another by relations
• Airline System Example

• relations
• informal
• advice, trust, respect,
• information exchange
• formal
• exchange of money,
• information exchange
• multiplex

• actors
• persons
• groups
• organisations
• countries

9
Basic Components of Social Network
Analysis

• actors
• persons
• groups
• organisations
• countries

• relations
• informal
• advice
• trust
• respect
• information exchange
• formal
• exchange of money
• information exchange
• multiplex

10
Data Collection

• Same Types
• Interviews
• Questionnaires
• Observations
• Diaries
• Different Questions
• Name Generators Role Generators

11
Network Questions

• Name-generating questions
• Q1. Suppose you need sugar or something like
  that and the
• shops are closed, or you need a
  piece of equipment.
• Who would you ask to lend you these
  sort of things?
• Q2. Suppose you need help with jobs in or
  around the house,
• for instance holding a ladder or
  moving furniture.
• Who would you ask for this kind of
  help?
• Q3. Suppose you have problems with filling
  informs,
• for instance tax forms.
• Who would you ask for help with
  such problems?
• Q4. Most people from time to time discuss
  important matters with
• others. Looking back over the last
  six months, who are the
• people with whom you discussed
  matters important to you?

12
Network Questions (continued)

• Q5. Suppose you need advice with a major change
  in your life,
• for instance changing jobs or
  moving to another area.
• Who would you ask for advice if
  such a major change
• occurred in your life?
• Q6. Suppose you have the flu and must stay in
  bed for a
• couple of days. Who would you ask
  to take care of you
• or do some shopping?
• Q7. Suppose you need to borrow a large sum of
  money.
• Who would you ask?
• Q8. Suppose you have serious problems with your
  partner
• which you cannot discuss with him
  or her.
• With whom would you talk about such
  problems?

13
Network Questions (continued)

• Q9. Suppose you are feeling depressed and you
  want
• to talk to someone about it.
• With whom would you talk about
  such problems?
• Q10. With whom do you go out once in a while,
  for
• instance shopping, going for a
  walk, going to a
• restaurant, or to a movie?
• Q11. With whom do you have contact at least once
  a
• month, by visiting each other for
  a chat, a cup of
• coffee, a drink, or a game of
  cards?
• Q12. Is there anybody else who is important to
  you, not
• mentioned so far? Relatives, or
  co-workers who
• are important to you?

14
Recording Social Relations

• strength
• frequency
• duration
• intimacy

• binary
• tie is simply either present or absent
• valued
• tie strength is recorded using some scale
• undirected
• tie is recorded as symmetric / mutual
• directed
• tie is recorded as asymmetric

15
Types of Networks

• egocentric or personal networks
• relations defined from focal individuals
• compare relational structures of actors from
  different organisations
• sociocentric or whole networks
• relations linking members of a single, bounded
  population
• examine internal structures and positioning of
  actors within one organisation

16
Differences in Network Types

• sociocentric networks
• require total list of all actors in population
• not all networks are bounded
• multiple sources of information which increases
  reliability and validity

• egocentric networks
• information can be obtained through large-scale
  surveys
• one source of information assumes focal actor
  provides accurate information for all relations

17
Representations of Networks

• graphs
• nodes represent actors
• lines represent relationships between actors
• actor-by-actor sociomatrices
• rows and columns indicate the same actors
• entries in each cell record related actors
• actor-by-event sociomatrices
• rows represent actors, columns are events
• entries in each cell record which actors attended
  which events

18
Communication Network)
19
Network Theory
20
Strength of Weak Ties (Granovetter 1973, 1983)

• Actors gain novel information from less intimate
  ties than close ties
• actors who are strongly connected share
  information directly, therefore they possess the
  same knowledge
• new information comes from external connections
  which are likely to be weak

21
Tie Strength Argument
Actors B D have a weak link thus are in a more
likely position to acquire novel information
which gives them a better opportunity for early
adoption
E
B
D
A
F
C
22
Structural Equivalence (Burt 1987)

• Actors who are identically positioned within a
  network use one another as a frame of reference
  to evaluate their relative adequacy
• actors are not necessarily directly connected
• actors are in competition with one another
• actors who are in the same equivalence class
  (status) tend to adopt at the same time

23
Structural Equivalence Argument
Actors B D are structurally equivalent and
have an indirect awareness of one another thus
they will adopt similarly
B
A
C
D
24
Structural Holes (Burt 1992)

• An actor is in a more advantageous position to
  gain and control novel information if s/he is
  connected to others who themselves are not
  directly connected to one another.
• Nonredundant sources of information allow
  efficiency with little constraint
• Developed to explain interpersonal communication
  in a competitive environment

25
Structural Holes Argument
Both actors A B are connected to 4 others.
Actor A is highly constrained with only 1
nonredundant tie. Actor B has 4 nonredundant
ties and is in a better position to access novel
information and adopt early.
A
X
F
W
B
C
E
D
Y
Z
26
Small World (Stanley Milgram 1967)

• Six degrees of separation
• The Oracle of Bacon at Virginia
• The Oracle says Nicole Kidman has a Bacon
  number of ?
• Kidman was in Peacemaker, The (1997) with Steve
  Altes Steve Altes was in Hollow Man
  (2000) with Kevin Bacon
• The Oracle says Hugh Jackman has a Bacon number
  of ?
• Jackman was in Someone Like You (2001) with
  Colleen Camp Colleen Camp was in Trapped (2002)
  with Kevin Bacon

27
Measures of Network Structure

• cohesion
• distance, reachability, geodesics, max flow
• diffusion of information / communication
• regions
• components, bi-components, k-cores
• partitioning of the network into distinct parts
• subgroups
• cliques, clans, k-plexes, lambda sets, factions
• partitioning of the network into subsets -
  overlap possible
• strategic alliances, collaborators / competitors

28
Measures of Network Structure

• ego networks
• density, structural holes
• constructs ego network for every actor in network
• size, density, wkcomp, reach, efficiency,
  constraint
• centrality
• degree, closeness, betweenness, ..
• identifies actors with most access/control over
  resources
• provides overall measure of network
  centralisation

29
Measures of Network Structure

• core/periphery
• categorical, continuous
• identifies actors in core and those in periphery
• roles positions
• structural equivalence, regular equivalence
• identifies actors who are similar in their
  relations with respect to actors in other
  positions
• network properties
• density, transitivity
• measures of complete networks

30
10 Principles of SNWs

• 1) Networks are Invisible
• privacy is protected but marketers are in the
  dark
• 2) People link with others who are Similar
• homophily matters
• 3) People talk with those who are Physically
  Close
• proximity matters
• 4) People who are Similar Close form Clusters
• 5) Info quickly Spreads within Dense Clusters
• people in the same clique know the same info

31
10 Principles of SNWs

• 6) Information gets Trapped in Clusters
• 7) Bridging Ties assist Info Flow between
  Clusters
• opinion leaders have connections to multiple
  clusters
• 8) Weak Ties are Surprisingly Strong
• acquaintances are important sources of novel
  information
• 9) The Net Nurtures Weak Ties
• explains why information travels much faster
  today
• 10) Networks go Across Markets
• users and non-users talk about multiple product
  categories