Building a Data Driven Culture
By Lucas Neo
If 1 book represents 1 petabyte of data…
50 Petabytes 24 Petabytes
If 1 book represents 1 petabyte of data…
50 Petabytes
All the worlds data (books, webpages,
videos, songs, etc.) between 4000BC
and 2006AD.
24 Petabytes
Amount of data processed daily by
Google alone in 2008.
[1]
[2]
“We are drowning in data but
starved for information”
~ Gary Cokins [3]
Influencers of Data-driven Growth
Brian Balfour, VP of Growth @
Hubspot
Andy Johns, VP of Growth @ Wealthfront
(previous Growth PM at Facebook,
Twitter, Quora)
Starting a Growth Team
SystemiseBrainstorm Prioritise Testing AnalyseImplement [4]
Building a Growth Team
SystemiseBrainstorm Prioritise Testing AnalyseImplement
Key Learning #1:
Have an executive champion
[4]
[5]
Encounters with Disparate Data
Multiple sources
of data from apps
Managed and used
by separate teams
To produce reports
and make decisions
Encounters with Disparate Data
Multiple sources
of data from apps
Managed and used
by separate teams
To produce reports
and make decisions
Key Learning #2:
Centralise your data
Generating Wins
Dashboards for
key metrics
Revamping
event tracking
Centralising data
sources
Creating
prediction models
Generating Wins
Key Learning #3:
Plan your value milestones
Dashboards for
key metrics
Revamping
event tracking
Centralising data
sources
Creating
prediction models
Build a Process
Standard data
science flow
Automated data
extraction and cleaning
Educating the company
on SQL and data
Build a Process
Standard data
science flow
Key Learning #4:
Make it replicable
Automated data
extraction and cleaning
Educating the company
on SQL and data
Recap
#1 Have an executive champion
#2 Centralise your data
#3 Plan your value milestones
#4 Make it replicable
What’s next for us?
Matrix Organisational
structure
Build a growth machine
Airbnb has a data
scientist in every team
Hubspot’s growth
machine by B. Balfour
Sources
[0] Icons from the Noun Project
Title slide: Data Planning and Data Network by Ivan Colic
Slide 6: Advocate by Lorie Shaull
Slide 7 & 8: Database by Shmidt Sergey, Group by Alexandr Cherkinsky, Graph by Vicons Design
Slide 9 & 10: App analytics by Sergey Novosyolov
Slide 11 & 12: Reporting by Alfredo Hernandez, Repeat by Pantelis Gkavos
[1] Kelly Kevin, 2006. http://www.nytimes.com/2006/05/14/magazine/14publishing.html?pagewanted=all&_r=0
[2] Dean and Ghemawat, 2008. http://dl.acm.org/citation.cfm?doid=1327452.1327492
[3] Gary Cokins, 2013. http://www.forbes.com/sites/jeffthomson/2013/10/30/why-cfos-are-drowning-in-data-but-
starving-for-information/#29968dd52623
[4] Brian Balfour, 2014. http://www.coelevate.com/essays/growth-process-first-tactics-second
[5] Josh Schwarzapel https://medium.com/android-news/how-to-start-a-growth-team-ff70cd29c0f2#.uls9ubea6
End / Fin

Building a Data-Driven Culture

  • 1.
    Building a DataDriven Culture By Lucas Neo
  • 2.
    If 1 bookrepresents 1 petabyte of data… 50 Petabytes 24 Petabytes
  • 3.
    If 1 bookrepresents 1 petabyte of data… 50 Petabytes All the worlds data (books, webpages, videos, songs, etc.) between 4000BC and 2006AD. 24 Petabytes Amount of data processed daily by Google alone in 2008. [1] [2]
  • 4.
    “We are drowningin data but starved for information” ~ Gary Cokins [3]
  • 5.
    Influencers of Data-drivenGrowth Brian Balfour, VP of Growth @ Hubspot Andy Johns, VP of Growth @ Wealthfront (previous Growth PM at Facebook, Twitter, Quora)
  • 6.
    Starting a GrowthTeam SystemiseBrainstorm Prioritise Testing AnalyseImplement [4]
  • 7.
    Building a GrowthTeam SystemiseBrainstorm Prioritise Testing AnalyseImplement Key Learning #1: Have an executive champion [4] [5]
  • 8.
    Encounters with DisparateData Multiple sources of data from apps Managed and used by separate teams To produce reports and make decisions
  • 9.
    Encounters with DisparateData Multiple sources of data from apps Managed and used by separate teams To produce reports and make decisions Key Learning #2: Centralise your data
  • 10.
    Generating Wins Dashboards for keymetrics Revamping event tracking Centralising data sources Creating prediction models
  • 11.
    Generating Wins Key Learning#3: Plan your value milestones Dashboards for key metrics Revamping event tracking Centralising data sources Creating prediction models
  • 12.
    Build a Process Standarddata science flow Automated data extraction and cleaning Educating the company on SQL and data
  • 13.
    Build a Process Standarddata science flow Key Learning #4: Make it replicable Automated data extraction and cleaning Educating the company on SQL and data
  • 14.
    Recap #1 Have anexecutive champion #2 Centralise your data #3 Plan your value milestones #4 Make it replicable
  • 15.
    What’s next forus? Matrix Organisational structure Build a growth machine Airbnb has a data scientist in every team Hubspot’s growth machine by B. Balfour
  • 16.
    Sources [0] Icons fromthe Noun Project Title slide: Data Planning and Data Network by Ivan Colic Slide 6: Advocate by Lorie Shaull Slide 7 & 8: Database by Shmidt Sergey, Group by Alexandr Cherkinsky, Graph by Vicons Design Slide 9 & 10: App analytics by Sergey Novosyolov Slide 11 & 12: Reporting by Alfredo Hernandez, Repeat by Pantelis Gkavos [1] Kelly Kevin, 2006. http://www.nytimes.com/2006/05/14/magazine/14publishing.html?pagewanted=all&_r=0 [2] Dean and Ghemawat, 2008. http://dl.acm.org/citation.cfm?doid=1327452.1327492 [3] Gary Cokins, 2013. http://www.forbes.com/sites/jeffthomson/2013/10/30/why-cfos-are-drowning-in-data-but- starving-for-information/#29968dd52623 [4] Brian Balfour, 2014. http://www.coelevate.com/essays/growth-process-first-tactics-second [5] Josh Schwarzapel https://medium.com/android-news/how-to-start-a-growth-team-ff70cd29c0f2#.uls9ubea6
  • 17.