The document discusses the differences between data-driven companies and data-justified companies. Data-driven companies start with a business question and use a scientific process to capture data, form hypotheses, run tests, make recommendations, and implement changes based on the results. Data-justified companies start with pre-existing beliefs and cherry-pick data that supports those beliefs while discarding other data. The document advocates that analytics professionals should only work for genuinely data-driven companies and outlines strategies for identifying data-justified versus data-driven companies and managers.
6. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
7. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
8. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
9. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
10. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
11. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
This what data-driven
looks like
12. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
This what data-driven
looks like
• Start with beliefs and the idea you
want to support
13. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
This what data-driven
looks like
• Start with beliefs and the idea you
want to support
• Request more data than what you
really need
14. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
This what data-driven
looks like
• Start with beliefs and the idea you
want to support
• Request more data than what you
really need
• Cherry-pick the data that justifies
your prior beliefs, discard the rest
15. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
This what data-driven
looks like
• Start with beliefs and the idea you
want to support
• Request more data than what you
really need
• Cherry-pick the data that justifies
your prior beliefs, discard the rest
• Make your business case
16. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
This what data-driven
looks like
• Start with beliefs and the idea you
want to support
• Request more data than what you
really need
• Cherry-pick the data that justifies
your prior beliefs, discard the rest
• Make your business case
• Implement changes
17. Data-driven vs data-justified
• Start with a business question
• Brainstorm about the data we will
need
• Implement the data capture
• Formulate an hypothesis
• Run a test
• Make a recommendation
• Implement changes
This what data-driven
looks like
• Start with beliefs and the idea you
want to support
• Request more data than what you
really need
• Cherry-pick the data that justifies
your prior beliefs, discard the rest
• Make your business case
• Implement changes
This is what data-justified
looks like
27. Analytics talent shortage
Let’s build an analytics team with internal employees
• Great understanding of the business
• Poor data literacy and objectivity
28. Analytics talent shortage
Let’s build an analytics team with internal employees
• Great understanding of the business
• Poor data literacy and objectivity
Let’s hire external talent
29. Analytics talent shortage
Let’s build an analytics team with internal employees
• Great understanding of the business
• Poor data literacy and objectivity
Let’s hire external talent
• Hard to find, expensive
30. Analytics talent shortage
Let’s build an analytics team with internal employees
• Great understanding of the business
• Poor data literacy and objectivity
Let’s hire external talent
• Hard to find, expensive
• Lack of domain knowledge
42. Making your own luck
• Data-justified companies can only recruit by taking advantage of ill-
informed analysts
43. Making your own luck
• Data-justified companies can only recruit by taking advantage of ill-
informed analysts
• The analysts, sick of boritoring, start building networks and
44. Making your own luck
• Data-justified companies can only recruit by taking advantage of ill-
informed analysts
• The analysts, sick of boritoring, start building networks and exchange
opinions and information about better places to work for and
45. Making your own luck
• Data-justified companies can only recruit by taking advantage of ill-
informed analysts
• The analysts, sick of boritoring, start building networks and exchange
opinions and information about better places to work for and how
much they are worth
46. Making your own luck
• Data-justified companies can only recruit by taking advantage of ill-
informed analysts
• The analysts, sick of boritoring, start building networks and exchange
opinions and information about better places to work for and how
much they are worth
• Experienced analytics practitioners know which managers have a
proven data-driven record
47. Making your own luck
• Data-justified companies can only recruit by taking advantage of ill-
informed analysts
• The analysts, sick of boritoring, start building networks and exchange
opinions and information about better places to work for and how
much they are worth
• Experienced analytics practitioners know which managers have a
proven data-driven record. Anybody else, the answer is нет (nyet)
50. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
51. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
• Geeks deemed as unsuitable
candidates for managerial roles
52. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
• Geeks deemed as unsuitable
candidates for managerial roles
• Hard to replace an analyst once
promoted because of talent
scarcity
53. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
• Geeks deemed as unsuitable
candidates for managerial roles
• Hard to replace an analyst once
promoted because of talent
scarcity
I can’t become a head of analytics?
Oh well, hello data science!
54. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
• Geeks deemed as unsuitable
candidates for managerial roles
• Hard to replace an analyst once
promoted because of talent
scarcity
I can’t become a head of analytics?
Oh well, hello data science!
• Every year Big Four consultants
look for client-side manager roles
55. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
• Geeks deemed as unsuitable
candidates for managerial roles
• Hard to replace an analyst once
promoted because of talent
scarcity
I can’t become a head of analytics?
Oh well, hello data science!
• Every year Big Four consultants
look for client-side manager roles
• They will then rotate every couple
of years until a CXO role
opportunity comes
56. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
• Geeks deemed as unsuitable
candidates for managerial roles
• Hard to replace an analyst once
promoted because of talent
scarcity
I can’t become a head of analytics?
Oh well, hello data science!
• Every year Big Four consultants
look for client-side manager roles
• They will then rotate every couple
of years until a CXO role
opportunity comes
• Therefore prior analytics
experience is irrelevant and
perhaps even bad
57. Managers with analytics skills?
• Many top-performing employees
fail their transition to
management
• Geeks deemed as unsuitable
candidates for managerial roles
• Hard to replace an analyst once
promoted because of talent
scarcity
I can’t become a head of analytics?
Oh well, hello data science!
• Every year Big Four consultants
look for client-side manager roles
• They will then rotate every couple
of years until a CXO role
opportunity comes
• Therefore prior analytics
experience is irrelevant and
perhaps even bad
Head of analytics? What the heck is
that? I will rotate in 2 years, right?
60. Expert leadership
More and more experienced analytics practitioners are finally getting
promoted Head of Analytics and implement a genuinely data-driven
approach and transform the analytics department into a profit centre
61. Expert leadership
More and more experienced analytics practitioners are finally getting
promoted Head of Analytics and implement a genuinely data-driven
approach and transform the analytics department into a profit centre
Expert leaders are a great motivator for more junior analysts who can
look up to someone who was just like them 5 or 10 years ago
62. Expert leadership
More and more experienced analytics practitioners are finally getting
promoted Head of Analytics and implement a genuinely data-driven
approach and transform the analytics department into a profit centre
Expert leaders are a great motivator for more junior analysts who can
look up to someone who was just like them 5 or 10 years ago
In cities where flats are ridiculously expensive, expert leadership could
help a mid-weight analyst stop renting and get a mortgage instead
65. Remember this?
What if all the web and data analysts worked only for data-driven
companies?
66. Remember this?
What if all the web and data analysts worked only for data-driven
companies?
If you are working in a data-justified department
67. Remember this?
What if all the web and data analysts worked only for data-driven
companies?
If you are working in a data-justified department, this
department only exists
68. Remember this?
What if all the web and data analysts worked only for data-driven
companies?
If you are working in a data-justified department, this
department only exists because you and your
colleagues took their job
69. Remember this?
What if all the web and data analysts worked only for data-driven
companies?
If you are working in a data-justified department, this
department only exists because you and your
colleagues took their job instead of the same job but
at a data-driven company
72. Nobody wants to work for us?
• I told him “That’s how we do web analytics here”. A week later, he
handed me his resignation, he had three job offers elsewhere. He was
still in his probation period
73. Nobody wants to work for us?
• I told him “That’s how we do web analytics here”. A week later, he
handed me his resignation, he had three job offers elsewhere. He was
still in his probation period
• I don’t understand what’s going on, I’m only getting junior candidates
from the career pages and the recruiters say that nobody is interested
74. Nobody wants to work for us?
• I told him “That’s how we do web analytics here”. A week later, he
handed me his resignation, he had three job offers elsewhere. He was
still in his probation period
• I don’t understand what’s going on, I’m only getting junior candidates
from the career pages and the recruiters say that nobody is interested
• I thought the interview went well, she was a strong candidate. Then
the recruiter said she told him after that I could not name one single
thought-leader in analytics and she won’t work for us
79. Identify data-justified companies
• Find other people in analytics
• Figure out how much you are really worth
• Identify the companies and managers who are data-driven in
our field
80. Identify data-justified companies
• Find other people in analytics
• Figure out how much you are really worth
• Identify the companies and managers who are data-driven in
our field
• When a company is hiring, try to find the name of the
manager and check their credentials and reputation
81. Identify data-justified companies
• Find other people in analytics
• Figure out how much you are really worth
• Identify the companies and managers who are data-driven in
our field
• When a company is hiring, try to find the name of the
manager and check their credentials and reputation
• A company had Adobe Analytics and migrated to Google
Analytics = symptom of a company that could not deliver
value from analytics
84. At your next interview, ask them
• So, what’s your definition of analytics?
85. At your next interview, ask them
• So, what’s your definition of analytics?
• Can you name one thought-leader in the field of analytics?
86. At your next interview, ask them
• So, what’s your definition of analytics?
• Can you name one thought-leader in the field of analytics?
• What’s the last analytics blog or book you have read in the
past 3 months?
87. At your next interview, ask them
• So, what’s your definition of analytics?
• Can you name one thought-leader in the field of analytics?
• What’s the last analytics blog or book you have read in the
past 3 months?
If they answer wrong, they fail the interview