Big Data MBA

Chapter 9“By” Analysis Technique

Chapter 8, “Thinking Like a Data Scientist,” briefly introduced the “By” analysis as a technique around which the business subject matter experts (SMEs) and the data science team could collaborate to uncover new variables and metrics that might be better predictors of business performance. “By” analysis is a technique that was historically used during the data warehouse requirements gathering processes to ensure that the data warehouse schema was robust enough to support the full range of Business Intelligence queries and reports that business users might request. Data science builds on the “By” analysis to create a collaborative technique to drive alignment between the business users and the data scientists to identify and brainstorm variables and metrics that might be better predictors of business performance. The “By” analysis technique re-enforces the importance of the “thinking like a data scientist” process.

Remember the data science definition from Moneyball: The Art of Winning an Unfair Game covered in Chapter 5:

Data science is about finding new variables and metrics that are better predictors of performance.

The “By” analysis technique supports this data science objective by powering the partnership between the business users and the data scientists to leverage new sources of customer, product, operational, market, and competitive data, coupled with advanced analytics, to uncover metrics and variables that may be better predictors ...