By Robert Morton July 21, 2009
Bruce Boston started using Tableau several years ago at C-Net. He now works at Apple Inc. in the AppleCare Analytics division, but he wasn't shy about insisting on using Tableau in his new role. Software such as BootCamp and Parallels eliminated the platform barrier, allowing Bruce to successfully deploy Tableau Desktop and Server within Apple.

The key to Bruce's evangelism of Tableau was his insight into the clear path to saving money within the company, while actually enhancing the quality of Apple's product. This quality-oriented focus to product management is known as Kaizen: "a Japanese business philosophy of continuous improvement of the total product experience."

Bruce focused his improvements on automating the number-crunching parts of analysis, while empowering human analysis suited to our strengths such as: metadata awareness, qualitative valuation and contextual interpretation. Using Tableau, he was able to offer timely analysis of product defects, exploring the number of reported incidents per product each week. Bruce couldn't share any of Apple's data, but he mocked up some dummy data sets to explain his approach, as in the attached image. He used time on both axes, showing the weeks since product launch on the horizontal axis versus the number of elapsed weeks from a customer's purchase to their reporting of product defects. Diving deeper into the data, statistical analysis exposed patterns in the reported defects and allowed analysts to confirm causation from the correlation - for example, by root-causing problems with suppliers, software glitches, or integration defects.

His guide to the division of labor for Kaizen 10-step analysis is:

  • Discovery (software automation)
  • Statistical validation (software automation)
  • Correlation (software automation)
  • Financial valuation (software automation)
  • Review results of discovery (Tableau)
  • Engineering evaluation (human evaluation)
  • Business explanation (Tableau)
  • Process improvement (human evaluation)
  • Engineering re-evaluation (human evaluation)
  • Verification and review (human evaluation)