MetPy is a toolkit for meteorological and atmospheric science applications in Python, building on the extensive, vibrant suite of libraries built by the scientific Python community. MetPy features domain-specific tools around calculations, plotting, and reading in data from a variety of formats. The current focus for the MetPy project centers around performance and scalability, with a goal to support analysis of larger datasets. This work leverages many existing tools, such as Cython and Dask. This work discusses our experience building up infrastructure around benchmarking the project, which is a necessary step to improving the performance of the project, and ensuring those improvements persist.