A Journey from Zero to Hero

2017/10/11 Tableau Conference 2017での、清水の講演資料になります

Transcript

1. A Journey from Zero to Hero (and What Not to Do on that Journey!)
3. over 200 brands
5. Hiroki Maeda Hiroki Maeda Data Product Strategist
6. 2011 2012 2013 2014 2015 2016 2017 2018 100 1,000
8. Travel data Ballooning of data Complexification of data Number of cases doubled in past 3 years Need for analysis that combines diverse data • Behavior data (Adobe) • Inventory data • Point data • Ad data • Competitor data • etc. Requests for data processing were concentrated on the dedicated analysis team, lowering the speed of decision-making
9. 2011 2012 2013 2014 2015 2016 2017 2018 100 1,000
14. Executive A B C … CRM Salesperson Producer UI / UX CRM Salesperson Producer UI / UX CRM Salesperson Producer UI / UX Product A Product B Product C … 7 ◎ 3 ◦ 4 ◎ 1 × 2 ◦ 5 ◎ 6 ◎
15. 2011 2012 2013 2014 2015 2016 2017 2018 100 1,000
17. Ryusuke Shimizu Recruit Technologies Co., Ltd. BigData Department Chairman Recruit Tableau Users’ Group
19. Comprehensive Tableau Function Education (free) Follow-up after Education (Tableau Doctor) Step-Up Approach (paid) Receipt of inquiries on Tableau Portal
20. • Adobe Analytics
22. Search box
23. Ordered by popularity
24. Information on the reservation table
25. Comments from engineers Comments from engineers
26. Meaning of various codes Code name Code value Code meaning
27. SQL statements of the referenced table
28. Created and saved SQL Saved SQL
29. 2012 2013 2014 2015 2016
30. Search Find Acquire Knowledge Update Expansion
31. 2011 2012 2013 2014 2015 2016 2017 2018 100 1,000 Over 1,000 licenses
33. Team for each business area Leader Leader Leader
34. “Data storytelling is so important in completing a profit-generating Viz with overwhelming high-quality”
35. Viz requirements: "Develop a Dashboard to serve as a tool for client retention" Challenge: 38 iterations Viz Value: About ¥600 million (6 million US dollars) / year Key Success Factor: Data storytelling
36. Profit-generating Viz Appropriate KPIs Appropriate Design
37. Level 0 Level 1 Level 2 KPI Unclear Too Many Appropriate Design Sales Designer Designer (with Appropriate Design) Data storytelling Cannot Not Good Perfect Data storytelling is "Cannot"
38. Profit-generating Viz Appropriate KPIs Appropriate design
39. Data storytelling is "Not Good" Level 0 Level 1 Level 2 KPI Unclear Too Many Appropriate Design Sales Designer Designer (with Appropriate Design) Data storytelling Cannot Not Good Perfect
40. Profit-generating Viz Appropriate KPIs Appropriate Design
41. Level 0 Level 1 Level 2 KPI Unclear Too Many Appropriate Design Sales Designer Designer (with Appropriate Design) Data storytelling Cannot Not Good Perfect
44. The speed of business Total optimality Basic User Training Data Catalog Custom User Training Self-Service-ETL
45. CoE LoB Monthly training class (hands-on) Tableau Basic General (super store) Data Source Production real-data Data connection Production DWH SQL From basic to window function Scenarios Real objectives Certification Exam with real data
46. AWS-S3 DWH DWH
47. AWS-S3 DWH