As Ledd Slinger noted, many reloaders use the relationship between seated bullet ogive and lands as a guide to how much freebore the chamber will need to keep the seated bullet within a certain region within the neck. Really long bullets need more freebore, short bullets need less. The amount of freebore needed for any given bullet or class of bullets can be readily estimated by making a few dummy rounds and taking some measurements. Alternatively, talking with someone from a reamer manufacturer such as JGS or Manson will likely be beneficial as they may have many reamer designs on file for any given cartridge that have already been optimized for a given weight class (length) of bullets. Also, a conversation with knowledgable gunsmith can achieve the same goal.
With a commercial rifle, one takes whatever freebore they get from the factory and tries to reload using a bullet and/or seating depth that will generate acceptable results within those parameters, meaning one may have either limited bullet length/weight choices, or may need to cover a much wider range during seating depth testing if the factory freebore is fairly long, meaning they can seat the bullet of choice almost as far in or out of the neck they want.
Over the years, I have commonly heard the value tossed about at shooting forums and such that one would ideally like at least half a caliber of bullet bearing surface seated in the neck. Many different cartridge necks are close to one caliber in length, so that notion provides some reference as to how much neck length there is to "play with" when determining optimal freebore for a given bullet. In other words, the ideas would be to keep approximately [at least] half the neck length occupied with bullet bearing surface while maintaining a specified distance relationship between bullet ogive and the rifling. In this scenario, one would thus have about half the neck length to play with in terms of choosing a freebore length for a given bullet that is aimed at maintaining a certain distance between seated bullet ogive and the lands.
For this reason, choosing the correct freebore length for a given bullet or class of bullets should not be guesswork. For example, the neck length for .308 Win is very close to one caliber. In making freebore estimates for a specified .30 cal bullet or class of bullets, if one wanted maintain at least half a caliber of bullet bearing surface in the neck, they would have approximately 0.150" to play with in terms of keeping the bullet boattail at or above the case neck/shoulder junction, but not more than halfway out the neck. This is a fairly generous margin for error, but certain bullet-specific factors can still make a difference. Does the bullet (or bullets) of choice want to be seated out to hard jam for optimal performance? That might use up as much as .020" to .025" of the 0.150" total available to play with. Does the bullet of choice want to be jumped .030" or more? That would mean a difference in bullet seating depth of as much as .050" to .055" from the jammed bullet example. So even though 0.150" might seem like a lot of room to play, it can change pretty quickly.
Finally, another critical question that needs to be addressed is the bullet design itself. Will the bullet (or bullets) of choice actually shoot well with only a certain amount of bearing surface seated in the neck? It is possible to seat bullets with only a third, or even a quarter caliber (or less) bearing surface in the neck. Although it may appear that bullets seated with so little bearing surface in the neck are "about to fall out", with sufficient neck tension/interference fit, you probably cannot pull one out solely using your fingers. They can be that tightly seated, even if it looks odd. So it's not necessarily an issue of bullet grip in the neck. Nonetheless, some bullets do not appear to shoot well seated way out in the neck. Alternatively, some may shoot quite well. Unless one has information regarding seating depth optima from other users of a given bullet, testing is the only way to know with certainty whether a given bullet will tolerate (or even prefer) being seated with only a small amount of bearing surface seated in the neck.
I have always tried to select freebore length such that when seated at "touching", the bullet boattail/bearing surface junction is at or above the case neck/shoulder junction, but not more than halfway out the neck. In other words, the bullet boattail/bearing surface junction will be located in the bottom half of the case neck. That means a seated bullet will have somewhere between one-half and a full caliber worth of bearing surface in the neck. Preferably, I target the middle of that range (i.e. the bullet boattail/bearing surface junction is approximately one quarter of the neck length above the shoulder). I should note that none of the various competition/target bullets I shoot require being seated into the lands; they are all jumped in tuned loads. In general, this has worked pretty well for me. Only once have I ever ordered a reamer with a freebore optimized for a much longer bullet and then tried to load a markedly shorter bullet that I knew needed to be seated within about .010" of the lands or so, meaning it seated with only about 0.100" bearing surface in the neck (0.30 cal bullet). Unfortunately, that particular bullet did not shoot so well seated so far out in the neck, so I simply stuck with using the much longer bullet in that particular rifle. Nonetheless, there are bullets that work just fine seated with only a small amount of bearing surface in the neck. If one has certain bullets in mind for a given application, it is sometimes possible to make inquiries at various shooting forums about the seating depth preferences for a specific bullet. Someone may have the exact information needed, especially if it is a commonly-used bullet. Other resources including gunsmiths and reamer manufacturers will also likely have good information about commonly-used bullets and optimized freebore lengths. It is a good idea to find and obtain as much of this info as possible before making a final decision that is irrevocable once the chamber has been cut. Even so, sometimes one may have to make their best guess and then deal with whatever consequences arise once they begin load development.
