Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Data Science Salon: How to Build a Data-Driven Culture


Published on

The world of modern data teamwork isn’t one that can be created by software and business process alone. Individuals will need to alter their behavior, which is the hardest part about change. This workshop will show you not only how to help your team evolve, but the reasons that will make it clear as to why they should.

Next DSS MIA Event -
Next DSS AUS Event -

Published in: Data & Analytics
  • Be the first to comment

Data Science Salon: How to Build a Data-Driven Culture

  1. 1. DataPractices:DataPractices: 2.1HowtoBuildaData-Driven2.1HowtoBuildaData-Driven CultureCulture 1 BACK TO COURSEWARE
  2. 2. HistoryHistory 2 BACK TO COURSEWARE
  3. 3. ValuesandPrinciplesforDataPracticesValuesandPrinciplesforDataPractices What set of values and principles describes the most effective, ethical, and modern approach to data teamwork? 3 BACK TO COURSEWARE
  4. 4. 4corevalues4corevalues 12principles12principles 39authors39authors 1400+signatories1400+signatories 4 BACK TO COURSEWARE
  5. 5. FourcorevaluesFourcorevalues Inclusion Maximize diversity, connectivity, and accessibility, amoung data sources, colaboration, and outputs. Experimentation Ephasise continuously iterative testing and data analysis. Experimentation Ephasise continuously iterative testing and data analysis. Experimentation Ephasise continuously iterative testing and data analysis. 5 BACK TO COURSEWARE
  6. 6. SupportedbyleadersofthedatacommunitySupportedbyleadersofthedatacommunity 39 authors, including: Eric Colson, Chief Algorithms O icer, StitchFix Amy Gershkoff, former Chief Data Scientist, Fernando Perez, creator of iPython, Assistant Professor, Statistics, UC Berkeley Andrew Therriault, Chief Data O icer, City of Boston Therese Couture, Human Tra icking Data Analyst, Polaris Wes McKinney, BDFL, Pandas 1,300+ signatories, including: DJ Patil, former Chief Data Scientist of the United States Monica Rogati, former VP of Data Science, Jawbone Kirk Borne, Principal Data Scientist, Booz Allen Hamilton Tricia Wang, Fellow, Harvard Berkman Center Jonathan Albright, Research Director, Tow Center for Digital Journalism Gregory Piatetsky, founder, 6 BACK TO COURSEWARE
  7. 7. GettingStartedGettingStarted 7 BACK TO COURSEWARE
  8. 8. TopicsTopics CoveredCovered 1. Framing the Problem 2. Pillars of a Data-Driven Company 3. Data-Driven Leadership 4. Decision Making 5. Treat Data Like an Asset! 6. Data Governance isn't a Dirty Word 7. Break Down Silos 8. Ask Questions 9. The "Culture" of Your Data-Driven Organization 8 BACK TO COURSEWARE
  9. 9. Exercise0:OutlineyourchallengesExercise0:Outlineyourchallenges As it relates to data, write down what your primary organizational challenges are. Consider: InfrastructureInfrastructure ToolingTooling CultureCulture CreationCreation CurationCuration 5 minute exercise (solo) 9 BACK TO COURSEWARE
  10. 10. FramingtheProblemFramingtheProblem 10 BACK TO COURSEWARE
  11. 11. PlayBigger The book “Play Bigger” was targeted primarily at businesses and how they can build a category. Many of these lessons can be applied to the challenge of building good data practice within an organization. What do you need to solve? Key component to the problem? De ine the "villain" 11 BACK TO COURSEWARE
  12. 12. WinningWithData The book “Winning with Data explores the cultural changes big data brings to business, and shows you how to adapt your organization to leverage data to maximum effect. Authors Tomasz Tunguz and Frank Bien draw on extensive background in big data, business intelligence, and business strategy to provide a blueprint for companies looking to move head-on into the data wave.” 12 BACK TO COURSEWARE
  13. 13. DefiningYourDataProblemsDefiningYourDataProblems BreadlinesBreadlines ObscurityObscurity FragmentationFragmentation BrawlsBrawls 13 BACK TO COURSEWARE
  14. 14. data.worldPOV Data is dramatically shaping the future of how decisions are made. Sitting at the con luence of governments, industry, and community, has been watching the ecosystem evolve. Data + Community Data Governance & Tool Fatigue Transformation Through Data Change is People-Powered What's YOUR Version of Open? 14 BACK TO COURSEWARE
  15. 15. Exercise1:RefineyourchallengesExercise1:Refineyourchallenges Having heard how others frame the problem, re ine your original outline. Consider: Whatproblem(s)doesyourorgneedtosolve?Whatproblem(s)doesyourorgneedtosolve? (breadlines,obscurity,etc)(breadlines,obscurity,etc) WhatpercentageofyourworkforceusesdataonadailyWhatpercentageofyourworkforceusesdataonadaily basistomakeinformeddecisions?basistomakeinformeddecisions? Whatthingsdoesyourorganizationdowell?Whatthingsdoesyourorganizationdowell? Whattools/processesdoyouuse?Whattools/processesdoyouuse? Whatquestionsdoyouusedatatoanswer?Whatquestionsdoyouusedatatoanswer? WhatquestionsSHOULDyoubeusingdatatoanswer?WhatquestionsSHOULDyoubeusingdatatoanswer? 5 minute exercise (solo) 15 BACK TO COURSEWARE
  16. 16. PillarsofaData-PillarsofaData- DrivenCompanyDrivenCompany16 BACK TO COURSEWARE
  17. 17. DataInfrastructure There are many different aspects to data infrastructure, and each company will need to solve their own needs here (there is no silver bullet). The real key here is that it solves for issues like accessibility, interoperability, ease of use, etc. Data Sources Data lakes Data warehouses Databases Applications (streaming or API data) Spreadsheets / lat iles Dark data Data access / dictionary / single source or truth 17 BACK TO COURSEWARE
  18. 18. DataGovernance Data governance isn’t a dirty word! (more on that later) As your organization operationalizes data, there need to be considerations and processes built around the availability, usability, integrity and security of your data. While many people consider this to be unnecessary “gating” in front of your data, and the associated work around it, there are ways to achieve good data governance without getting in the way. 18 BACK TO COURSEWARE
  19. 19. DataLiteracy While having tools, processes, and infrastructure are important parts to any data solution, it is the workforce which truly de ines whether or not an organization is data- driven. You absolutely don’t have to be highly technical to be a data practitioner, but even if you aren’t a practitioner you should be in the habit of using data to make good, informed decisions. To do this there are some basic skills that you’ll need to have. Examples (Facebook, AirBnB) Speaking the same data language Who? 19 BACK TO COURSEWARE
  20. 20. Data-DrivenData-Driven LeadershipLeadership20 BACK TO COURSEWARE
  21. 21. LeadingtheChange While it is the practitioners who will typically do the bulk of the transformative work, leadership needs to be both aware of what transformations need to happen as well as lead by example. Be patient, change takes time Data isn’t everything, avoid paralysis Hold employees accountable Create quanti iable goals Avoid "expert syndrome" 21 BACK TO COURSEWARE
  22. 22. Exercise2:SocializeyourExercise2:Socializeyour challengeschallenges Now that we've thought about organizational structure and leadership, take advantage of a partner who can help act as a sounding board to focus your thinking. 1. Describeyourdataprocess(doyouhaveone?)toyourDescribeyourdataprocess(doyouhaveone?)toyour partnerend-to-end.Includeeverythingfromkickofftopartnerend-to-end.Includeeverythingfromkickoffto 2. Highlightwhatyouthinkyoudowell,andhowmanyHighlightwhatyouthinkyoudowell,andhowmany people/businessunits/disciplines/etcparticipateinpeople/businessunits/disciplines/etcparticipatein thisprocess.Isitsiloedorcollaborative?thisprocess.Isitsiloedorcollaborative? 3. Describeyourcurrentproblemsorinefficiencies.DoyouDescribeyourcurrentproblemsorinefficiencies.Doyou alreadyhavethoughtsonhowtofixthem?Doyoustillalreadyhavethoughtsonhowtofixthem?Doyoustill havequestions?Areyourproblemsmoretechnicalhavequestions?Areyourproblemsmoretechnical (tools/infrastructure/supply)orcultural(tools/infrastructure/supply)orcultural (literacy/collaboration/communication)?(literacy/collaboration/communication)? 30 minute exercise (Pairs) 22 BACK TO COURSEWARE
  23. 23. DataDecisionMakingDataDecisionMaking The trick of making meaningful and informed decisions from your data is making sure that you’re asking real questions, and actively pursuing the answers, rather than creating frivolous dashboards that aren’t used for actual decision making. PractitionersPractitioners Responsibilities: Data collection / pipeline Data prep / validation Deep analysis SME/Non-TechnicalSME/Non-Technical Responsibilities: Data catalog Data limitations Familiar analysis 23 BACK TO COURSEWARE
  24. 24. Treatdatalikeanasset Historically organizations have treated data as a cost center, and as a result it was never prioritized the way that it could have been. Most organizations now realize, whether on their own or from examples like Google or Facebook, that operationalized data can be an immense asset (or sometimes even a revenue stream). 24 BACK TO COURSEWARE
  25. 25. DataGovernanceisn'tadirty word. Data governance used to be a luxury (or complication) that only large organizations needed. With increasing focus and importance on data and analytics it’s becoming an organizational necessity for everyone. Building consistency Ensuring data quality Establishing common lexicon Regulatory requirements (ex: GDPR) Democratizing access 25 BACK TO COURSEWARE
  26. 26. UNUSEDUNUSED DARKDARK RESTRCTIEDRESTRCTIED 80% of data is never analyzed Data sitting on laptops Problem for reproducibility / auditability Most common. Leads to data breadlines / dark data BreakdownsilosBreakdownsilos 26 BACK TO COURSEWARE
  27. 27. AskQuestions Asking questions of your org/team/colleagues is the best way to ensure constant growth and focus. Do you have a data portal/platform? What percentage of your employees use it each week? What tools and training are available? What documentation do you have in place to support your data? 27 BACK TO COURSEWARE
  28. 28. CultureCulture 28 BACK TO COURSEWARE
  29. 29. TEAMTEAM ORGORG WORLDWORLD Small team collaboration Fight against "dark data" Push AND pull work lows Single pane of glass for all data Cross disciplinary self-service Uni ied data language Only for data that makes sense Multiplicitive potential / ROI Be careful of PHI/PII/PFI Defineyourversionof"open"Defineyourversionof"open" 29 BACK TO COURSEWARE
  30. 30. Buildingculturethroughpractice.Buildingculturethroughpractice. Organizational plumbing 30 BACK TO COURSEWARE
  31. 31. Exercise3:FinalizeyourplanExercise3:Finalizeyourplan Now that you have a good understanding of all the pieces, inalize your problem statement and action plan. Think in broad strokes, this isn't your inished strategy, but the framework to develop one from later. 1. Whattools/infrastructuredoyouhaveinplacenowforWhattools/infrastructuredoyouhaveinplacenowfor data?Considerany/allofthefollowing:data?Considerany/allofthefollowing: Data Catalog / Governance Data Science / Analysis Data Visualization / Dashboarding / Reporting Others 2. Whatisyourorganizationaldataprocesscurrently?DoWhatisyourorganizationaldataprocesscurrently?Do youhaveone?Whatneedstochange?youhaveone?Whatneedstochange? 3. WhatistheculturearounddatainyourorganizationWhatistheculturearounddatainyourorganization like?Woulditsupportchangeandevolution?Shouldit?like?Woulditsupportchangeandevolution?Shouldit? 4. Whattwochangeswouldmakethemostimpact?SelectWhattwochangeswouldmakethemostimpact?Select onethatisshortterm(ex:tool/process)andonethatonethatisshortterm(ex:tool/process)andonethat wouldbelongerterm(ex:education).wouldbelongerterm(ex:education). 30 minute exercise (Pairs) 31 BACK TO COURSEWARE
  32. 32. Exercise4:PresentyourplanExercise4:Presentyourplan Volunteer members of the workshop present their outlines / plans and participate in collaborative Q&A / brainstorming. Remainder (Group) 32 BACK TO COURSEWARE
  33. 33. Wanttorunaworkshoplikethisatyourcompany?Wanttorunaworkshoplikethisatyourcompany? 33 BACK TO COURSEWARE