The general problem of test data creation and use is extremely complex (perhaps a candidate to classificationas a wicked problem - https://en.wikipedia.org/wiki/Wicked_problem#:~:text=In%20planning%20and%20policy%2C%20a,are%20often%20difficult%20to%20recognize.).
For any non-trivial activity, I would say addressing the problem of test data creation and use requires sitting down and explicitly conceiving a Test Data Strategy as part of your Test Plan. This should consider several factors:
- The complexity of the data;
- The existence of Production data, or parts of the data;
- The confidentiality of the data, or parts of the data (or other requirements; ex: GDPR);
- Correct and incorrect data;
- Frequency of test execution;
- Interference of test execution in terms of data use;
- Time required to restore test data to its state prior to test execution;
- Isolation of testers activities in terms of data use;
- Cost of test data creation from scratch;
The are just a few essential topics that come to mind.
One crucial aspect is that you will not find a test strategy that will be optimal, or even good, in every scenario. The right solution, if this exists at all, much like in software architecture, is very sensitive to many factors.