Two different approaches have emerged: “top-down” and the “bottom-up”.
A top-down approach “institutionalizes” a company’s vision through a comprehensive long-term strategy. It emphasizes organization-wide coordination, economies of scale, and alignment with the business model. It aims to provide an enterprise-wide prioritization of a “single version of the truth”, and leverages cloud platforms to rearchitect data and provide access to leading analytics services.
The advantage of a top-down approach is that it enables strategic enterprise-wide decisions, can help attract and retain top-tier talent, and provides a truly leading-edge architecture. However, the sheer scale of the transformation required means it can take longer to bear fruit—and can be more susceptible to losing focus.
A bottom-up approach is characterized by organic agility and speed to insight. Business units or individual teams build or buy their own solutions to meet their immediate data and analytics requirements, faster.
Bottom-up approaches offer more sustained innovation. Insights are typically more tightly integrated with processes, and more closely aligned with the needs of business users. However, it can mean the organizational approach to analytics lacks overall coherence, resulting in a fragmented/siloed architecture that limits opportunities for data-led innovation at scale.