Skip to main content

Build a growth mindset with retrospectives (Microsoft Fabric Migration)

Retrospectives reinforce the principles of a growth mindset: experimentation, testing, learning, sharing, growing, and empowering. They also provide a safe space for team members to share the challenges they faced during the Fabric migration and help teams establish sustainable, data-driven practices for improvement.

Retrospective structure​

For the migration of the Data Estate to Microsoft Fabric, conduct a structured retrospective by having each team member reflect on three key questions:

  1. What went well?
  2. What could have been better?
  3. What did we learn?

This session marks the end of a release or Fabric iteration (e.g., wave 1: ingestion setup, wave 2: Power BI alignment, wave 3: Data Activator integration).

Conducting the retrospective​

To ensure meaningful discussion:

  • Facilitator: Assign a retrospective coach to guide the session and foster open dialogue.
  • Tooling: Use tools like Azure DevOps Boards, Miro, or Microsoft Whiteboard to track outcomes.
  • Structure: Organize insights according to the three questions above. Cluster by theme (e.g., governance, pipeline performance, semantic layer).

Example retrospective output (Fabric Migration)​

What went well?What could have been better?What did we learn?
Seamless deployment of Data Pipelines with Dataflows Gen2.Delay in semantic model refresh after Lakehouse changes.Use CI/CD to sync Lakehouse and semantic layer changes.
Clear governance policies for domain teams.Ambiguity in data product ownership roles.We need stronger onboarding for Fabric Product Owners.
Data Activator alerts triggered as expected.Some DQ checks not integrated before cutover.Include DQ pipeline validation in Definition of Done.

Applying insights​

  • Add improvement actions to the Fabric Migration backlog.
  • Integrate findings into upcoming sprint planning or onboarding materials.
  • Share lessons learned with other domains and mesh product owners.

Contributors