The Secrets of AI Value Creation

13Data Management

Companies have tons and tons of data, but [success] isn't about data collection, it's about data management and insight.

—Prashanth Southekal (business analytics author, professor, head of the Data for Business Performance Institute)

When it comes to AI value creation requirements, data emerges as a crucial element, demanding a significant investment. However, data alone isn't enough; effective data life cycle management is crucial. The 2020 ESI ThoughtLab study of 1,200 organisations revealed that nearly all AI‐mature organisations excel in data management, whereas most followers consider it their main challenge (Celi and Miles 2020). In terms of investment, the study shows that data management requires about 35% of AI spending or in absolute numbers, on average $13.3 million out of $38 million per year (Celi and Miles 2020).

However, AI beginners tend to proportionally invest more in data management compared to AI achievers (Celi and Miles 2020).

The challenge of data management goes beyond data security and quality. A 2020 McKinsey study on data management costs reveals the allocation breakdown: 38% for data sourcing, 13% for data governance, 29% for data infrastructure, and 20% for data preparation and consumption (Grande et al. 2020).

Data Management

The main objective of data management is to ensure that plans, policies, programs, and practices are effectively developed, implemented, and managed to increase, control, protect, and enhance the value ...