Without much idea on what your experience is, no insult intended if this is too basic.
If'n it were me, I'd start by deciding what I wanted the round to do. Is it a hunting round where energy on target is most important, or is it long-range competition where energy is less important than tiny groups over a 20 round competitive string.
Regardless, look in a reloading handbook,they make a caliber-specific paperback compilation of quite a few manuals in a ~$6.95 paperback form) to see roughly what charges are used for a given projectile weight.
Back off the maximum load by a couple of grains of powder, and work your way up to a "reasonable" load in .3gr. increments. Typically I will run 5 round test groups of each powder load. For the powder/velocity test loads, hold the "jump" to the lands constant. For a LTB,length tolerant bullet i.e. sierra MK, most hunting bullets) start at .020 off the lands, for a VLD,very low drag, typically a long-range competition bullet) type bullet, set the length so that the bullet JUST touches the lands.
Fire your "powder test" loads over a chronograph at a target at 100 yards. Document both the group size and the velocity for each load. If you plot group size v.s. powder charge, the resulting curve should give you a very good idea what powder charge/velocity your rifle likes.
Take your best "velocity" load, load up a half-dozen more loads with the best powder charge. This time vary the seating depth between each load. For LTB, try about .010" on either side of .020" off lands, varying seating depth by ~.003-.004" per load.
Chronograph your range testing just as you did with the "powder" test, documenting group size and velocities for each load.
The low point on your curves should give you a reasonably close approximation of a "pet" load for that bullet type/weight. You can refine from there.
***NOTE***: For best results, do not do this sort of testing with virgin brass. Fire form around 30 cases in your rifle first so that the differences in performance between virgin and fire formed brass do not skew your results.
This what you were after?
Darrell