Modeling Markets as They Are: How to Build Market Intelligence Data Models that Drive Impact

Data scientists have the responsibility not only to engage with upstream data sources but also to shape them to meet the organization’s needs. This talk will discuss the practical aspects of building a data model for market intelligence (MI) research. We will also share lessons learned from recent MI data modeling efforts within Intel’s Corporate Strategy Office. The development of actionable business strategies requires segmenting customers and products. Sound segmentation patterns account for multidimensionality and accommodate analysis at multiple levels of granularity. Building a data model that adequately reflects both realities can be challenging, and the use of spreadsheet software quickly becomes impractical. Modern MI calls for purpose-built data models that allow analysts to think about markets the way they actually are. This material covers three topics: 1) Modeling taxonomies, 2) capturing data, and 3) using the data model to drive impact.