I often work with large and complex reports. Sometimes making a minor change can have unexpected consequences. I like to be able to see that the only things that changed are the things I intended to change. My simple approach to this is described in this article.

R-tag has created a more sophisticated version of this process with their Unit Testing module.  It allows you to create a set of parameters and a validated output file (image, a PDF, a spreadsheet, etc.) for those parameters.  The parameter combinations are called theories, and the validated output file are called golden files.

After you make changes to the report, the data or the environment,  you can run a unit test for that theory.  The software will generate a new output file using the theory parameters, and will automatically compare that file with the golden file, listing all of the differences.  You can create multiple theories for each report with different different parameter combinations. You can then run the unit tests manually (e.g. right after a database update) or on a schedule.

My simple approach will only find values that change, but the unit testing feature can compare the output as two images, which will pick up formatting differences as well as data changes.

R-tag has a demonstration video here.