empowering machine learning
play

EMPOWERING MACHINE LEARNING SOLVING THE FORECASTING DILEMMA Tom - PowerPoint PPT Presentation

EMPOWERING MACHINE LEARNING SOLVING THE FORECASTING DILEMMA Tom Stanek & Jonathan Prantner 2 Tom Stanek Jonathan Prantner President Chief Analytics Officer 3 Leading applied artificial intelligence and data science company from Ann


  1. EMPOWERING MACHINE LEARNING SOLVING THE FORECASTING DILEMMA Tom Stanek & Jonathan Prantner 2

  2. Tom Stanek Jonathan Prantner President Chief Analytics Officer 3

  3. Leading applied artificial intelligence and data science company from Ann Arbor, MI Services and solutions • Leading Domo implementation and consulting firm • Custom artificial intelligence kick-start program • RXA Studio • Media Optimization • Voice of Customer • Workforce Optimization Over 70 different customers across North America, Europe, and Asia. 2019 Domo Innovative Partner of the Year 4

  4. 3 KEY TAKEAWAYS Solve Provide Domo 6

  5. The Forecasting Dilemma Now what?

  6. WE NEED TO KNOW WHAT ARE SALES WILL BE NEXT QUARTER 8

  7. SALES ARE FORECASTED TO BE BELOW BUDGET... 9

  8. …WHY? 10

  9. …AND HOW CAN WE CHANGE THEM? 11

  10. Sales Decomposition A Primer

  11. WHAT IS A SALES DECOMPOSITION Market Mix Models decompose sales Volume Decomposition Due-to Analysis 13

  12. MARKETING MIX FORECASTING Forecasting Sales using: Sales • Inventory • Pricing • Marketing • External Influences • 14

  13. USE A PREDICTION MODEL TO DECOMPOSE SALES 15

  14. USE A PREDICTION MODEL TO DECOMPOSE SALES Predict with the overall model Isolate each variable – Set the variable equal to zero – Predict with the reduced model – Calculate difference Scale each component to the total GET THE CODE AT: https://rxa.io/Domo 16

  15. DEPLOYING IN DOMO 17

  16. DUE TO ANALYSIS 18

  17. DOMO GETS THE RIGHT INFORMATION INTO THE DECSION MAKERS HANDS 19

  18. MARKETING MANAGER What Matters to Them: Media Performance • Optimization • Web Traffic • ROI • Sales • 20

  19. MARKETING MANAGER What They Can Control: Budgets • Schedules • Mix • Strategy • 21

  20. MARKETING MANAGER VIEW 22

  21. MARKETING MANAGER VIEW 23

  22. MARKETING MANAGER VIEW 24

  23. MARKETING MANAGER VIEW 25

  24. MARKETING MANAGER VIEW 26

  25. MARKETING MANAGER VIEW 27

  26. MARKETING MANAGER VIEW 28

  27. PRODUCTION MANAGER What Matters to Them: Inventory • Costs • Workforce • Demand • Sales • 29

  28. PRODUCTION MANAGER What They Can Control: Inventory • Suppliers • Production Calendar • 30

  29. PRODUCTION MANAGER VIEW 31

  30. PRODUCTION MANAGER VIEW 32

  31. PRODUCTION MANAGER VIEW 33

  32. PRODUCTION MANAGER VIEW 34

  33. External Influences

  34. WHY CONTROL DATA Accounts for external pressures Improves forecast Provides context 36

  35. SIMPLE WITH DOMO 37

  36. EXTERNAL PRESSURES Unemployment Interest rates Construction Weather Consumer Confidence 38

  37. IMPROVES FORECAST Without Proper Control Data With Ready Signal Control Data Control data can significantly improve accuracy of models, but is often underutilized based on required effort and knowledge. 39

  38. DOES A 60-DEGREE DAY MEAN FEWER MOTORCYCLE SALES? 40

  39. PRODUCTION MANAGER VIEW 43

  40. PRODUCTION MANAGER VIEW 44

  41. EXECUTIVE TEAM What Matters to Them: Sales • Profitability • Growth • 45

  42. EXECUTIVE TEAM What They Can Control: Strategy • Resources • Budgets • Competitive Positioning • 46

  43. EXECUTIVE TEAM VIEW 47

  44. EXECUTIVE TEAM VIEW 48

  45. 3 KEY TAKEAWAYS Solve Provide Domo 49

  46. THANK YOU & join us at our virtual booth!

Recommend


More recommend