Trouble with load development is there are so many variables to test that to control for all of them, it’d burn out a whole barrel just to find statistically valid results.
So the biggest factors, in order (in my opinion) are:
Try different bullets
Try different powders
Try major changes in seating depth (for example 15 thou off lands and 30 thou)
Try major changes in powder load (full grain difference)
I have a load development worksheet in the load development course that will walk you through the process step by step.