$30 off During Our Annual Pro Sale. View Details »

Social Network Analysis Made Easy

Jeff Mohr
October 28, 2014

Social Network Analysis Made Easy

More than ever, we need to learn how to harness the power of networks to tackle the complex issues we're facing as a society. Here's a quick guide to the basics of social network analysis.

Interested? Sign up at http://kumu.io

Jeff Mohr

October 28, 2014
Tweet

More Decks by Jeff Mohr

Other Decks in Education

Transcript

  1. SOCIAL NETWORK
    ANALYSIS
    ( made easy )

    View Slide

  2. happiness is best predicted
    by the breadth & depth
    of one’s social connections.
    - Robert Putnam

    View Slide

  3. WHY NETWORK ANALYSIS?

    View Slide

  4. reason #1
    The challenges we face are so complex
    they can’t be solved by any one
    organization.

    View Slide

  5. The urgency and scale of social
    problems, coupled with the limited
    results to date, cry out for
    new approaches.
    - Jane Wei-Skillern, Nora Silver and Eric Heitz “Cracking the Network Code”

    View Slide

  6. over
    1.5 million
    non-profits in the US

    View Slide

  7. Organizations have been the lever
    through which we try to create social
    change for far too long.
    !
    We have to bring people together across
    sectors, from within and outside
    government, and from all walks of life.

    View Slide

  8. Reason #2
    Even within organizations, hierarchies
    aren’t accurate representations of how
    work actually gets done.

    View Slide

  9. Reason #2
    Even within organizations, hierarchies
    aren’t accurate representations of how
    work actually gets done.
    org charts lie!

    View Slide

  10. Information doesn’t flow along
    organizational hierarchies.
    !
    Networks are a far more accurate picture
    of how work gets done.
    org charts lie!

    View Slide

  11. Reason #3
    We must understand the
    status quo to overcome it.

    View Slide

  12. The status quo is a result of the web of
    relationships and incentives among
    stakeholders (including us).
    !
    It’s not that we’re “stuck” — it’s that
    competing interests provide a balancing
    effect that resists change.

    View Slide

  13. hi there!
    Jeff Mohr
    Cofounder & CEO of Kumu

    View Slide

  14. Jeff Mohr
    Cofounder & CEO of Kumu
    my background
    systems
    networks
    social change

    View Slide

  15. So… WHERE can SNA help?

    View Slide

  16. • Identifying change leaders
    • Breaking down silos
    • Evaluating progress
    • Driving innovation
    social impact
    increasing

    View Slide

  17. • Weaving stronger connections
    • Bridging across silos
    • Reducing crime
    • Improving resilience
    stronger communities
    building

    View Slide

  18. ORGANIZATION PERFORMANCE
    • Promoting effective collaboration
    • Avoiding burn out
    • Selecting new leaders
    • Uncovering informal structures
    improving

    View Slide

  19. Great! How do i start?

    View Slide

  20. 3 STEPS
    collect + interpret + act

    View Slide

  21. Step 1
    COLLECT THE DATA

    View Slide

  22. Data can be collected via survey, pulled
    from existing data sources or populated
    via personal knowledge.
    surveys Data Knowledge
    Pull from spreadsheets,
    CRMs, public data, email
    traffic, social networks and
    more
    Surveys ask participants
    both relational and
    demographic questions
    Use the wisdom in the room
    to identify stakeholders and
    key relationships

    View Slide

  23. • Who do you work with?
    • Who do you turn to for new ideas?
    • Who do you turn to for advice?
    • How does working with this person affect your
    energy levels?
    Examples of Relational
    Survey Questions

    View Slide

  24. Examples of DEMOGRAPHIC
    Survey Questions
    • What is your age?
    • What sector do you work in?
    • What is your job title?
    • How many years experience do you have?

    View Slide

  25. GREAT RESULTS ARE DRIVEN
    FROM GReAT QUESTIONs.

    View Slide

  26. GREAT RESULTS ARE DRIVEN
    FROM GReAT QUESTIONs.
    !
    CHoose WISELY.

    View Slide

  27. And Don’t be afraid to simulate
    holes in the data.

    View Slide

  28. And Don’t be afraid to simulate
    holes in the data.
    just because they didn’t
    respond doesn’t mean they
    aren’t part of the network.

    View Slide

  29. Step 2
    INTERPRET

    View Slide

  30. Metrics provide an unbiased way to interpret relationships.
    You’ve got a few to choose from…
    degree INdegree OUTdegree ties pairs CLOSENESS
    farness reach betweenness eigenvector katz
    pagerank percolation cross-clique

    View Slide

  31. Metrics provide an unbiased way to interpret relationships.
    You’ve got a few to choose from… but we’ll focus on these
    three for now.
    degree INdegree OUTdegree ties pairs CLOSENESS
    farness reach betweenness eigenvector katz
    pagerank percolation cross-clique

    View Slide

  32. Understanding the
    Core Metrics
    degree + closeness + betweenness

    View Slide

  33. Degree
    Identifies local connectors and hubs

    View Slide

  34. Degree
    Identifies local connectors and hubs
    by counting the number of connections
    for a given element

    View Slide

  35. Degree
    WARNING
    Not necessarily the most influential or
    best connected to the wider network

    View Slide

  36. Closeness
    Identifies those with high visibility about
    what’s happening across the network

    View Slide

  37. Closeness
    Identifies those with high visibility about
    what’s happening across the network
    by measuring the distance from one
    element to all other elements

    View Slide

  38. These people can quickly spread
    information (good or bad)
    across the network
    Closeness
    WARNING

    View Slide

  39. Identifies key bridges and those who
    control the flow of information
    Betweenness

    View Slide

  40. Identifies key bridges and those who
    control the flow of information
    by counting the number of times an
    element lies on the shortest path
    between two other elements
    Betweenness

    View Slide

  41. These people may be bottlenecks
    or single points of failure
    Betweenness
    WARNING

    View Slide

  42. Metrics are people too

    View Slide

  43. Metrics are people too
    Each one reveals its own personality

    View Slide

  44. let’s Focus
    on the extremes

    View Slide

  45. two types of overly
    CENTRAL
    people

    View Slide

  46. bottlenecks
    Play central role
    to maintain information
    or power advantage
    OR
    people whose jobs have
    grown too big

    View Slide

  47. UNSUNG
    HEROES
    Engage selflessly to help
    the group in ways that
    often go unnoticed

    View Slide

  48. people at the
    Borders
    of the network

    View Slide

  49. Share different types of expertise, broker
    information and connect across geographies
    bridges

    View Slide

  50. OUTSIDERS
    Stuck on the periphery
    with no idea how to
    work their way inside
    intentionally peripheral
    OR

    View Slide

  51. Metrics only get
    You Started
    Use them to identify potential influencers
    and then validate with common sense
    WARNING

    View Slide

  52. Step 3
    DO SOMETHING

    View Slide

  53. This guy was obsessed
    with pretty pictures.

    View Slide

  54. This guy was obsessed
    with pretty pictures.
    !
    You’re better than that.

    View Slide

  55. go beyond the pretty
    picture and get shit done.
    Use strong visualizations, compelling
    narrative, and convincing arguments
    to make your impact.

    View Slide

  56. Use strong visualizations, compelling
    narrative, and convincing arguments
    to make your impact.
    Kumu helps you do all three
    shameless plug

    View Slide

  57. a few caveats to
    Network Analysis
    • be data-informed, not data-driven
    • take results with a grain of salt
    • validate using common sense

    View Slide

  58. let’s recap
    1. SNA helps tackle complex social problems.
    2. Use surveys, data, and local knowledge to build the network.
    3. Calculate metrics to identify key players within the network.
    4. Apply what you’ve learned to make a difference.
    5. Don’t forget to use common sense!

    View Slide

  59. If you want to go quickly, go alone.
    If you want to go far, go together.
    - African Proverb

    View Slide

  60. join us at Kumu.io

    View Slide

  61. Jeff Mohr is the cofounder & CEO of Kumu, a web-based platform that
    gives influencers the tools to track, visualize and leverage relationships
    to overcome their toughest obstacles.
    !
    Learn more at kumu.io or say hi @kumupowered
    Thanks!

    View Slide