Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Data & Design Like PB&J

C. Todd Lombardo
February 15, 2018

Data & Design Like PB&J

A 2015 PwC survey of 1,300 CEOs in 77 countries, ranked data mining and analytics as the second most important digital technology and organizational capability. What does this mean for designers? How can designers be “data literate?” Designers who understand data will be the designers who make a bigger impact with their work. Design solves problems, we know this well. Data helps inform the choices you make to solve those problems. Taking this a step further into product and experience design, this talk higlights how data can be used to help make design choices that produce better experiences for our users. We’ll present an approach to follow along with examples in the field to draw inspiration for your work.

C. Todd Lombardo

February 15, 2018
Tweet

More Decks by C. Todd Lombardo

Other Decks in Design

Transcript

  1. DATA & DESIGN
    LIKE PB & J
    C. TODD LOMBARDO — @IAMCTODD
    HEAD OF PRODUCT & EXPERIENCE @ WORKBAR
    [email protected]

    View full-size slide

  2. WHO WORKS WITH
    DATA TODAY?

    View full-size slide

  3. FAST COMPANY

    View full-size slide

  4. “DATA SCIENCE IS AN ACT OF
    INTERPRETATION — WE TRANSLATE THE
    CUSTOMER’S ‘VOICE’ INTO A LANGUAGE
    MORE SUITABLE FOR DECISION-MAKING.”
    Riley Newman, Head of Data Science @ Airbnb

    View full-size slide

  5. “DATA SCIENCE IS AN ACT OF
    INTERPRETATION — WE TRANSLATE THE
    CUSTOMER’S ‘VOICE’ INTO A LANGUAGE
    MORE SUITABLE FOR DECISION-MAKING.”
    Riley Newman, Head of Data Science @ Airbnb
    DESIGN
    NEEDS
    TAKING ACTIONS

    View full-size slide

  6. I II III IV
    x y x y x y x y
    10.0 8.04 10.0 9.14 10.0 7.46 8.0 6.58
    8.0 6.95 8.0 8.14 8.0 6.77 8.0 5.76
    13.0 7.58 13.0 8.74 13.0 12.74 8.0 7.71
    9.0 8.81 9.0 8.77 9.0 7.11 8.0 8.84
    11.0 8.33 11.0 9.26 11.0 7.81 8.0 8.47
    14.0 9.96 14.0 8.10 14.0 8.84 8.0 7.04
    6.0 7.24 6.0 6.13 6.0 6.08 8.0 5.25
    4.0 4.26 4.0 3.10 4.0 5.39 19.0 12.50
    12.0 10.84 12.0 9.13 12.0 8.15 8.0 5.56
    7.0 4.82 7.0 7.26 7.0 6.42 8.0 7.91
    5.0 5.68 5.0 4.74 5.0 5.73 8.0 6.89
    99.00 82.51 99.00 82.51 99.00 82.5 99.00 82.51
    9.00 7.50 9.00 7.50 9.00 7.50 9.00 7.50
    3.32 2.03 3.32 2.03 3.32 2.03 3.32 2.03

    View full-size slide

  7. I II III IV
    x y x y x y x y
    10.0 8.04 10.0 9.14 10.0 7.46 8.0 6.58
    8.0 6.95 8.0 8.14 8.0 6.77 8.0 5.76
    13.0 7.58 13.0 8.74 13.0 12.74 8.0 7.71
    9.0 8.81 9.0 8.77 9.0 7.11 8.0 8.84
    11.0 8.33 11.0 9.26 11.0 7.81 8.0 8.47
    14.0 9.96 14.0 8.10 14.0 8.84 8.0 7.04
    6.0 7.24 6.0 6.13 6.0 6.08 8.0 5.25
    4.0 4.26 4.0 3.10 4.0 5.39 19.0 12.50
    12.0 10.84 12.0 9.13 12.0 8.15 8.0 5.56
    7.0 4.82 7.0 7.26 7.0 6.42 8.0 7.91
    5.0 5.68 5.0 4.74 5.0 5.73 8.0 6.89
    99.00 82.51 99.00 82.51 99.00 82.5 99.00 82.51
    9.00 7.50 9.00 7.50 9.00 7.50 9.00 7.50
    3.32 2.03 3.32 2.03 3.32 2.03 3.32 2.03

    View full-size slide

  8. 1) WHO ARE MY USERS?
    2) WHAT ARE THEY DOING? WHAT WILL THEY DO?
    3) WHAT ARE THE LIMITATIONS TO THE DATA?
    4) WTF SHOULD I DO?

    View full-size slide

  9. 1 WHO ARE MY USERS?

    View full-size slide

  10. K-MEANS
    CLUSTERING

    View full-size slide

  11. FIND GROUPS WHICH HAVE NOT BEEN
    EXPLICITLY LABELED IN THE DATA.

    View full-size slide

  12. 3
    2
    1
    PICK RANDOM DATA POINTS

    View full-size slide

  13. 3
    2
    1
    PICK RANDOM DATA POINTS
    FIND NEAREST POINTS

    View full-size slide

  14. 3
    2
    1
    PICK RANDOM DATA POINTS
    MAKE CLUSTERS OF NEAREST DISTANCE
    FIND NEAREST POINTS

    View full-size slide

  15. 1
    3
    2
    PICK RANDOM DATA POINTS
    MAKE CLUSTERS OF NEAREST DISTANCE
    FIND NEAREST POINTS
    FIND NEW CENTER OF CLUSTER

    View full-size slide

  16. 1
    3
    2
    PICK RANDOM DATA POINTS
    MAKE CLUSTERS OF NEAREST DISTANCE
    FIND NEAREST POINTS
    FIND NEW CENTER OF CLUSTER
    CALCULATE NEW CLUSTERS

    View full-size slide

  17. 1
    3
    2
    PICK RANDOM DATA POINTS
    MAKE CLUSTERS OF NEAREST DISTANCE
    FIND NEAREST POINTS
    FIND NEW CENTER OF CLUSTER
    CALCULATE NEW CLUSTERS

    View full-size slide

  18. 1
    3
    2
    PICK RANDOM DATA POINTS
    MAKE CLUSTERS OF NEAREST DISTANCE
    FIND NEAREST POINTS
    FIND NEW CENTER OF CLUSTER
    CALCULATE NEW CLUSTERS
    STOP WHEN POINTS DON’T CHANGE CLUSTERS

    View full-size slide

  19. LISTING OF OFFERS
    SOURCE: DATASMART

    View full-size slide

  20. CUSTOMERS WHO HAVE
    PURCHASED OFFERS
    SOURCE: DATASMART

    View full-size slide

  21. PIVOT
    TABLES
    SOURCE: DATASMART

    View full-size slide

  22. MATRIX
    SOURCE: DATASMART

    View full-size slide

  23. DISTANCE
    TO CLUSTER
    CENTERS
    SOURCE: DATASMART

    View full-size slide

  24. USE SOLVER
    OBJECTIVE: MINIMIZE DISTANCE TO CLUSTER CENTERS
    DECISION VARIABLES: DEAL VALUES OF EACH ROW
    CONSTRAINTS: CLUSTER CENTERS BETWEEN 0 AND 1
    SOURCE: DATASMART

    View full-size slide

  25. TOP DEALS PER CLUSTER

    View full-size slide

  26. CLUSTER 1
    TOP DEALS
    WHO LOVES PINOT NOIR?
    SOURCE: DATASMART

    View full-size slide

  27. CLUSTER 2
    TOP DEALS
    WHO LOVES A GOOD DEAL?
    WHO’S NOT BUYING BIG?
    SOURCE: DATASMART

    View full-size slide

  28. CLUSTER 3
    TOP DEALS
    HMMM…..?
    HOLIDAY CHAMPAGNE?
    SOURCE: DATASMART

    View full-size slide

  29. CLUSTER 4
    TOP DEALS
    SUMMER CHAMPAGNE LOVERS!
    SOURCE: DATASMART

    View full-size slide

  30. PINOT LOVERS DEALHUNTERS SEASONAL
    BUYER
    SUMMER
    CHAMPAGNE LOVER

    View full-size slide

  31. 2 WHAT ARE THEY
    DOING? WILL THEY DO?

    View full-size slide

  32. WHAT ARE THEY
    DOING?

    View full-size slide

  33. WHAT WILL THEY DO?

    View full-size slide

  34. IMAGE: ESHAAN KAUL

    View full-size slide

  35. IMAGE: ESHAAN KAUL

    View full-size slide

  36. DATA NOT REAL, FOR EXAMPLE ONLY

    View full-size slide

  37. 27% 43% 10% 6%
    DATA NOT REAL, FOR EXAMPLE ONLY
    24%

    View full-size slide

  38. WHAT’S THE OBJECTIVE?

    View full-size slide

  39. WHAT’S THE OBJECTIVE?
    HOW DOES THE CURRENT DESIGN REACH THAT OBJECTIVE?

    View full-size slide

  40. WHAT’S THE OBJECTIVE?
    HOW DOES THE CURRENT DESIGN REACH THAT OBJECTIVE?
    WHAT WAYS CAN WE BETTER REACH THAT OBJECTIVE?

    View full-size slide

  41. 3 WHAT ARE THE
    LIMITS OF THE DATA?

    View full-size slide

  42. DATA CAN MISLEAD

    View full-size slide

  43. H T T P : / / W W W. T Y L E R V I G E N . C O M / S P U R I O U S - C O R R E L AT I O N S

    View full-size slide

  44. CORRELATION

    CAUSATION

    View full-size slide

  45. YOU ARE BIASED
    AND SO IS THE DATA

    View full-size slide

  46. A N C H O R I N G
    S TAT U S Q U O
    S E L E C T I O N
    N E G AT I V E
    C O N F I R M AT I O N
    I N - G R O U P
    P R O B A B I L I T Y
    R AT I O N A L I Z E
    G A M B L E R ’ S
    B A N D WA G O N
    P R O J E C T I O N
    C U R R E N T
    M O M E N T

    View full-size slide

  47. INCONVENIENT TRUTHS OF DATA SCIENCE
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  48. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  49. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  50. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  51. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    ‣ In 90% of cases generalized linear regression will do the trick.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  52. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    ‣ In 90% of cases generalized linear regression will do the trick.
    ‣ You should embrace the Bayesian approach.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  53. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    ‣ In 90% of cases generalized linear regression will do the trick.
    ‣ You should embrace the Bayesian approach.
    ‣ No one cares how you did it.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  54. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    ‣ In 90% of cases generalized linear regression will do the trick.
    ‣ You should embrace the Bayesian approach.
    ‣ No one cares how you did it.
    ‣ Academia and business are two different worlds.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  55. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    ‣ In 90% of cases generalized linear regression will do the trick.
    ‣ You should embrace the Bayesian approach.
    ‣ No one cares how you did it.
    ‣ Academia and business are two different worlds.
    ‣ Presentation is critical. Context makes the story
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  56. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    ‣ In 90% of cases generalized linear regression will do the trick.
    ‣ You should embrace the Bayesian approach.
    ‣ No one cares how you did it.
    ‣ Academia and business are two different worlds.
    ‣ Presentation is critical. Context makes the story
    ‣ All models are false, but some are useful.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  57. INCONVENIENT TRUTHS OF DATA SCIENCE
    ‣ Data is never clean.
    ‣ You will spend most of your time cleaning and preparing data.
    ‣ 95% of tasks do not require deep learning.
    ‣ In 90% of cases generalized linear regression will do the trick.
    ‣ You should embrace the Bayesian approach.
    ‣ No one cares how you did it.
    ‣ Academia and business are two different worlds.
    ‣ Presentation is critical. Context makes the story
    ‣ All models are false, but some are useful.
    ‣ There is no fully automated Data Science. You need to get your hands dirty.
    SOURCE: KAMIL BARTOCHA (LASTMINUTE.COM)

    View full-size slide

  58. 4 WHAT CAN I DO AS
    A DESIGNER?

    View full-size slide

  59. UNDERSTAND THE
    PROBLEM:
    ASK MORE QUESTIONS!

    View full-size slide

  60. WHAT ARE WE TRYING TO ACCOMPLISH?
    WHAT DO WE KNOW TODAY?
    WHAT DO WE WANT TO KNOW?
    WHAT DATA DO WE HAVE?
    WHAT DATA DO WE NEED?

    View full-size slide

  61. PROBLEM SOLUTION
    THANKS: W. BRÜNING

    View full-size slide

  62. PROBLEM SOLUTION
    Water on the floor Mop
    THANKS: W. BRÜNING

    View full-size slide

  63. PROBLEM SOLUTION
    Water on the floor Mop
    WHY? Leaky pipe Replace pipe
    THANKS: W. BRÜNING

    View full-size slide

  64. PROBLEM SOLUTION
    Water on the floor Mop
    WHY? Leaky pipe Replace pipe
    WHY? Too much pressure Lower pressure
    THANKS: W. BRÜNING

    View full-size slide

  65. PROBLEM SOLUTION
    Water on the floor Mop
    WHY? Leaky pipe Replace pipe
    WHY? Too much pressure Lower pressure
    WHY? Pressure regulator Replace regulator
    THANKS: W. BRÜNING

    View full-size slide

  66. PROBLEM SOLUTION
    Water on the floor Mop
    WHY? Leaky pipe Replace pipe
    WHY? Too much pressure Lower pressure
    WHY? Pressure regulator Replace regulator
    WHY? Maintenance schedule More frequent inspection
    THANKS: W. BRÜNING

    View full-size slide

  67. BRING DATA INTO YOUR
    DESIGN PROCESS

    View full-size slide

  68. K N O W T H E
    A U D I E N C E
    K N O W T H E
    D ATA
    U N D E R S TA N D
    C O N T E X T
    D E S I G N
    S O L U T I O N
    E VA L U AT E

    View full-size slide

  69. “WHEN WE DON’T WORK WITH REAL
    DATA, WE DECEIVE OURSELVES.”
    Josh Puckett, Design Partner @ Combine VC

    View full-size slide

  70. REFRESH YOUR
    MINDSET

    View full-size slide

  71. DESIGNER SCIENTIST
    EMPATHIC
    MAKER
    FORWARD LOOKING
    RIGOROUS
    EXPERIMENTAL
    HYPOTHESIS DRIVEN

    View full-size slide

  72. DESIGN / UX PRODUCT
    DEVELOPMENT

    View full-size slide

  73. DATA
    DESIGN / UX PRODUCT
    DEVELOPMENT DATA

    View full-size slide

  74. ¡GRACIAS!
    C. TODD LOMBARDO — @IAMCTODD
    HEAD OF PRODUCT & EXPERIENCE @ WORKBAR
    [email protected]

    View full-size slide