Data is both qualitative and quantitative in nature; hence, your data migration strategy should include a concrete and measurable success definition to determine when data migration can be considered complete. Now that you have a solid plan for configuration and data management, you should now explore all the available techniques that will assist you in accomplishing the plan.

The following is a suggested list of key activities you should factor into your data migration plan/strategy:

• Collect the requirements for data migration with measurable factors
• Identify all the data elements/entities and their sources
• Understand and keep the target solution/system schema in perspective
• Develop a governance strategy for leadership and direction
• Define data quality and integrity parameters
• Identify all the data validations and rules
• Identify and assign an owner for every type of data
• Define data conversion needs (if any)
• Agree on a data cleansing approach

• Collect data volumes per entity
• Identify when full data loads are needed and when the incremental approach needs to be taken
• Identify all the post data migration checkpoints
• Identify and leverage the tools provided by Microsoft/principal

Now, let's consider a list of items to factor in for your data migration plan:

• Environment: You need to plan for an environment to run the data migrations iteratively. You don't want the test environment to be messed with every week while the data migration team is still trying to stabilize the data migration processes.
• Cycles: You need a plan for multiple cycles of data migration that are a few weeks apart. This gives us time to validate the data, fix issues, and improve the performance of the migration processes.
• Resources: Business resources will be required to help extract and validate the data for each cycle. They may be needed to help cleanse the data if you run into issues with the legacy data. IT resources will be required to extract, import, and validate the data.
• Training: It is a good idea to train and utilize dedicated resources in the data conversion process as this is an iterative process. It also gives you experienced resources focusing on improving the process based on the feedback received from data validation.
• Verification: Data quality in the source system has a huge impact on the number of data migration iterations that you have to perform during tests.
• Testing: Complete a full data migration prior to starting system integration testing, UAT, and training. These migrations should be performed by following the data migration process documentation, and the time for each step needs to be recorded. As a part of this process, have the migrated data validated by the business prior to starting the tests in these environments.
• Automation: Come up with iterative/automated processes, including data extraction from legacy systems. This makes the cycle time for data migration shorter, improves its quality, and provides consistent results. For some extractions, you could use reports from the legacy system that the business uses. For example, if a business uses a detailed Accounts Receivable (AR) aging report, you can use that report as an input for migration rather than building a separate job for data extraction.

• Teamwork: The team should record the timing for each process and arrange dependencies and processes that can be run in parallel.
• Communication: Document the migration process end to end – from data extraction and intermediate validation to migration (the development team that writes the code should not be the one executing it). With a documented process, you can get more team members to execute the repetitive data migration processes.

The next step after planning is ensuring that you have a smooth and spot-on execution.