Best practices for ERP Test Data Management
1. Matching ERP data volume to demand. LEAN copy versus EXACT copy
When providing data for testing ERP systems, it is important to consider the types of testing that take place. If data are established for unit testing during development, the volume may be small and isolated to the process being tested. It makes no sense to put large amounts of data into a development branch that is not intended or suitable for performance testing.
2. Ensure that production data to be copied does not impact other scenarios
Many of our clients, particularly those in Consumer Electronics and Retail, come to us when they need to update test data. Without damaging other already present data or scenarios.
Some parts of a material master have been updated on the test environment, such as expanding the material to a new plant. This cannot be overwritten, but testing cannot continue without the latest pricing conditions, new suppliers or customers recently created in production. This requires laser-like precision in identifying data and the filter options chosen. Production-driven data virtualization can be a solution to this problem, as it is easy to create and restore snapshots before (or after) test execution.
3. Sensitive data should be masked without reducing its testing value
By intelligently masking labels and identifiers, you can use real data for testing without losing its value. When that capability is lost, companies fall back on asking the tester to create the data. The danger here is that the person testing brings their biases with them. While intelligently masked data often surprises us the most by revealing otherwise hard-to-discover flaws.
4. Good testing processes are repeatable and where possible automated
Manual testing is costly, prone to error, and simply too time consuming. To enable business agility, testing must become faster and easier to schedule and execute. With 35% of large enterprises’ IT budgets going to software testing, there are clear benefits to be gained through automation.
5. Create simple, change-focused test scripts
Make the data and exact testing process identifiable and recordable. This ads consistency and repeatability to your test scenarios. Once your organization has a collection of repeatable test scripts, the reliability of development goes up. Besides that, automation becomes much simpler, cheaper, and it allows manual testing (often performed by Business Key Users) to focus on the new and/or complex and critical cases.