Best Practices for Data Migration

For successful business migration, here is a list of critical points to consider:

Perform test migrations with production data before moving large chunks of data

Prepare a product rationalization plan and map it to the best plans and price points in the new stack to attract and enable customer-initiated migration

Have a well-documented and contextualized migration cookbook

Conduct functional testing over migrated data to assess both the completion and the correctness of migration

Build key capabilities, tools, accelerators, and metrics to monitor, measure, and test the reconciliation or operational corrections, which help in stabilizing and reducing customer care needs

Ensure the right level of communication with stakeholders to avoid panic and subsequent flood of queries on customer service representatives

Perform a volumetric assessment, and evaluate license needs, platform limits, as well as scaling capabilities before commencing migration, with reassessments from time to time.

Have additional customer care support to hand-hold customers on their queries during migrations to ensure a smooth and successful process

Start with a strategy. A public cloud requires investment in time and resources. Understand why you need cloud migration and follow up by monitoring and validating the results

Revisit and review the licensing agreement for the application from three vectors – performance per user, migration throughput, and scalability

Think big and start small. In a phased migration approach, opt for low-risk customers or low-priority data. This can help substantially reduce risks and adopt a framework to adjust or correct the parameters based on the customer context

Automate repetition for large volume workloads using Azure, Amazon, JMeter, microservices, UI Bots, or robotic process automation (RPA) tools

Proactively monitor, report, and alert for auditability using platform tools like Salesforce’s event monitoring/Splunk reports and dashboards, Azure Monitor, as well as Google cloud monitoring. These tools help assess platform limits, resources, and storage consumption with increased usage over time

Focus on legacy platform capabilities to support quicker and large volume migrations

Designs must achieve consistent user experience during the moving average convergence/divergence (MACD) journey for migrated and new services. They must enable the system to perform delta migrations for inflight orders as necessary

Design a rollback strategy for customers with fallouts that cannot be remediated or were pre-migrated but cannot be included in the cutover due to business reasons

Have the right tools and establish industry benchmarking, using it as a North Star in implementation, performance goals, and business outcomes

Establish well managed, responsible, accountable, consulted, and informed (RACI) roles with executive sponsorship and inclusiveness of all parties

Precise estimation of customer-initiated changes on product, plan, value-added services (VAS), and applicable discounts during the migration period helps in planning delta migration intervals

Control and orchestrate capability releases in the new stack, which will unlock customers eligible for migration

Prioritize customer experience over volumes where feasible, as a lack of focus on customer experience as well as stability and quality of migration can impact migration plans and timelines

Plan with the predictability of migration volumes at line level, including predictability of fallouts and exclusions. This can be achieved with the best-defined and implemented testing framework emulating the production world

Do not rush into high-volume migrations immediately after a functional release. Conduct a pilot for all scenarios in production to check operational stability and then consider the migration of newly eligible customers for the new functionalities unlocked in the target system