Accurate benchmarking is hard and there are lots of things that can skew your results. Compilers are very clever and will optimize, which can make trivial benchmarks less valuable. The compiler may output very similar code for different implementations.
What you put in your data structures will also have a large effect on their performance. It's usually best to test or profile your actual application and not optimize prematurely. Readability of code is very valuable and shouldn't be sacrificed for runtime efficiency unless there is a significant performance problem (or if it's already unreadable).
There are benchmarking frameworks that you can use to help with your testing, such as BenchmarkDotNet, which is available ...
