In software development, it's pretty normal to stage your deliverables from your Development environment to Test to Acceptation en finally to Production (DTAP). On it's way to production your software meets different data sets, hopefully improving in quality and relevancy.
In order to be more determinable when testing and accepting the software in the Test and Acceptation environments, your data should stage as well, but in the reverse direction.
On it's way you may need to obfuscate or anonymize the data.
It is a good idea to set up a way to continuously reverse stage the data. If you worry about the performance hit on your production machine, consider grabbing the data from a backup.