Organizing Enterprise Data Science

Recipes from the Global 2000 and Beyond

While the necessity to embed AI into the business is clear, the road to get there isn’t. One question many data science leaders wrestle with is how to organize data science teams to achieve the greatest impact.

• Do you centralize the function via a Data Science Center of Excellence, or COE, to build out economies of scale?
• Should data scientists be federated, working in the trenches with business staff where they can solve real problems faster?
• Or is a hybrid model that falls somewhere in between the best path to success?

In our conversations with nearly a dozen industry leaders building model-driven businesses, we found that there’s no one-size-fits-all answer. But there are a set of practices that leaders use to successfully:

• Create a discipline for managing the people and cultural changes necessary to embrace data science
• Establish scalable and repeatable processes, including metrics to measure success, across the end-to-end data science lifecycle
• Ensure data science teams have the right technology foundation to foster productivity and collaboration

Ultimately, these leaders demonstrate that regardless of how you organize data science, you need a strategy for scale and a path to get there quickly. Those lacking one will fall behind and likely struggle to deliver business impact.

In this report, we break down best practices across all three areas (discipline, process, technology) that these leaders of high-performing global data science teams shared with us. Whether you’re early in your journey or well underway and seeking to strengthen the impact of existing efforts, their insights can help you chart the right course for your organization.