In my previous blog post, “Do you test with Data or does it test you?” I had discussed some of the common pitfalls in selecting test data management solutions. In this blog, I will talk about avoiding these pitfalls for better business outcomes.
“A stitch in time saves nine” is an adage we have all heard. Software testing teams swear by this. Early bug detection can make the difference between success and failure in product launches. Defects at the production stage are not only expensive to fix but more importantly can cause loss of reputation. Hence, identifying and eliminating bugs even before they are coded is of paramount important. But with testing managers at cross-roads in selecting the right TDM solution, what testing strategies deliver the best ROI (Return On Investment)?
Here are the 5 ideas or best practices that can help you design and execute tests with the right test data:
- Version control the test data: Always create fresh set of data. Tag the batches with individual build numbers and store version control systems for future reference. Unless there is a very strong case for data persistency between two different builds, it is good to purge stale data and load a fresh set of data.
- Test with more data volume: Testing with large volumes of data helps capture various user and transactional profiles. Repetitive and small set of data quickly loose their ability to identify new defects.
- Analyze production defects: It is necessary to analyze production defects and identify failures that resulted from data deficiency. It should be a continuous process of updating your test data design to incorporate learnings from such analysis.
- Understand business functionality: Analyze the business and list functionality to understand the different workflows based on data. Focus on parameters such as geographic differences, exception processing, user profile based changes, user customization based changes and business rules.
- Optimize environment usage: Test environment is usually different from the production environment. This is a predicament testers face. Test managers should understand the differences and leverage data virtualization to bridge the gaps.
There you have it folks – some of the important points to consider before investing in a TDM solution. Having the right test data is important and a well-designed TDM strategy delivers the right test data. It is also important to ensure that the test data solution fully integrates with the DevOps processes – as this functionality automatically executes test scripts as soon as a build is ready for deployment on the QA environment. While commercial tools are expensive, with tools such as VirtusaPolaris’ Enterprise Data Generation Engine (EDGE) Solution, test managers can eliminate CAPEX, automate test data provisioning and test execution, all the while reducing dependency on development teams.