“Is everything working the way it should now?” This question is often uttered while introducing translation memory, terminology management or similar systems. We can quite simply answer this question with “Yes!” If, and only if a conscientious test management has been carried out from start to finish. In this blog, I present errors and obstacles we have encountered in client projects when introducing new tools. And of course I also explain why a conscientious test management guarantees a successful system go-live!
Testing is good, test planning even better!
Frequently, extensive testing of newly introduced or adapted systems takes place shortly before the scheduled go-live. Go-live of a system being the point in time when a system is transferred from testing state to productive state. At this point, however, the project duration is usually already exhausted and delays are no longer justifiable. This can lead to go-live of systems or functions that have not or not sufficiently been tested. The fixing of system errors in productive state is usually only possible at great expense. This is why sufficient time and resources for several (!) test loops should be included when planning the project. This is the only way to ensure that all bugs (errors) that prevent a successful go-live are identified and, above all, eliminated.
A case for test cases
Actual test execution is always preceded by the creation of relevant test scenarios and test cases. These should be created on the basis of existing requirements (change requests, specifications), but should also have the necessary practical relevance. Ideally, several sources of information are taken into account when creating test cases. This prevents the creation of wrong test cases, too few test cases or, with regard to efficient test management, even too many.
The different roles in a system are very important for test case creation. This is logical, because completely different functions are relevant for the administrator of a system compared to the end user. It is therefore essential to consider target groups of test scenarios when creating test cases.
Data, data, test data!
For meaningful test results you need good test cases and the right test data. Oftentimes test data for certain test scenarios is incomplete or not available at all. During a test execution, they then have to be created from scratch or manipulated manually. This leads to additional effort in the test execution phase, and even worse can lead to a falsification of test results. For this reason, all relevant parties should be involved in the procurement of usable, valid test data before testing begins.
More test standards for more success!
If a deviation from the desired system behavior or an error is found during test execution, it is necessary to estimate how serious the deviation or error really is. For this purpose, it is advisable to categorize the test scope, standardize the test documentation and assign test priorities that apply to all parties involved. That way a seemingly small deviation from the target behavior does not turn out to be a showstopper in the productive state. And an error that is classified as preventing go-live, which could easily have been eliminated by making minor adjustments to the configuration in production operation.
Good testing – happy end!
Test plan, test scenarios and test cases, test data, test documentation – these are all essential components of successful test management! We have made the experience that the positive effects of meticulous test management are often underestimated and that too little or no resources are allocated for this. This leads to additional cost and time delays during system introduction.
Conclusion: Meticulous test management removes all obstacles on the road to successful system go-live.
Are you planning to introduce a new translation management, machine translation or terminology system? Or planning new functions for your systems?
Please contact us! We’d love to support you with professional test management.
We also help with your system evaluations, pilot projects and proof-of-concepts.
Picture: Photo by Fotis Fotopoulos on Unsplash
Picture: Photo by Kelly Sikkema on Unsplash