Coaching for Data, Analytics, AI Leaders (1–10)

Principles and patterns to build winning performance+team

If you are leading teams responsible for Data, Analytics, Data Science, or AI — this blog series is for you! Throughout my career leading teams in Data Engineering, Data Analytics/Science, and AI in both startup and large company environments, I have uncovered principles and patterns that have proven to be effective. In this series, I will be sharing these principles and patterns.

#1 Look beyond titles to find the right talent

In hiring for your team, look beyond titles! There is currently a deluge of titles in the industry: Business/Technical Data Analyst, Analytics Engineer, Data Engineer, Data Scientist, BI Engineer, DataOps/MLOps engineer, etc. These titles do not have a consistent definition and mean different things based on the data, analytics, and AI maturity of the org. Instead, focus on matching past projects to the jobs-to-be-done within your team. Prioritize on learning grit given the roles of today will be significantly different in the future.

#2 For ad-hoc analytics requests, develop the muscle to ask “so what”

Ad-hoc data exploration requests from business partners can take a significant percentage of the team’s bandwidth. Develop the muscle within the analytics and data science teams to ask not only the “what” but also the “so what” for such requests i.e., if we discovered the specific data pattern or created a prediction model, how would it translate into actionable business decisions and how impactful would that be to business-specific KPIs. Having the upfront alignment on “so what” can ensure the effort spent by the team on ad-hoc requests has a clear line-of-sight to business impact.

#3 The Modern Analytics charter is extended

Given the fast pace of changes within the business, the charter of modern analytics needs to be extended. Traditionally, the analytics charter started only after data is ingested within the lake, and stopped once the insights were created and shared. The scope of modern analytics starts at the data source — making sure data is in the right format before being ingested in the lake (instead of endless pipeline jungles to transform and fix it later). It…