Recovering a cross-border CRM cloud migration
A custom CRM moved from on-premises to the Microsoft cloud, with a significant data migration, in two country go-lives. After the first go-live needed months of rework, I changed how the technical team worked, and the second go-live ran smoothly.
- Type
- Professional · client project (anonymised)
- Period
- Aug 2024 – Feb 2025
- Role
- Lead Consultant / Architect
- Sector
- Energy · Netherlands and Germany
Client names and financial figures are left out. Details are limited to what can be shared.
Context
A custom-built CRM was moving from on-premises infrastructure to the Microsoft cloud (Power Platform with Azure integration services). The migration included a significant data migration and went live per country: Germany first, then the Netherlands.
Problem
The first go-live in Germany led to significant rework, especially in the data migration. The rework continued for months and disrupted the normal project flow.
Constraints
Two countries with separate go-lives, a custom on-premises CRM as the source system, a large data migration, and a project that was already under pressure from the rework.
My role
Lead consultant and architect for the technical team. I changed the way the technical team worked and set the technical approach for the next go-live.
Decisions
I changed the management approach of the technical team, put a strong focus on best practices and design patterns, and streamlined the data migration and integration pipeline. One deliberate architectural choice: where Azure Data Factory could not cover a migration or integration requirement, we filled the gap with Azure Logic Apps instead of building workarounds inside Data Factory. This raised the migration budget, but it gave us pipelines that worked reliably and that the team could maintain, which mattered more after the rework of the first go-live.
Delivery
Built on Power Platform with ALM in Azure DevOps. Azure Data Factory and Azure Logic Apps for data migration and integration, and Power BI for reporting.
Outcome
The go-live in the Netherlands went smoothly, with almost no rework during the data migration. The more streamlined migration and integration pipeline saved the customer time and avoided rework costs.
IT considerations
- Reliability
- Repeatable migration and integration pipelines instead of manual fixes after go-live.
- Trade-off
- A higher migration budget, accepted in exchange for Logic Apps covering the functional gaps of Data Factory.
- Integration
- Azure Data Factory and Logic Apps between the new platform and surrounding systems.
- Maintainability
- Shared design patterns and solution-based ALM in Azure DevOps.