The poll's question potentially confuses two issues - working loads up from scratch without experience of that cartridge and/or powder and fine tuning for precision. It also makes no distinction between an experienced precision handloader and Joe or Jo Tyro starting out with a bullet maker's reloading manual and puzzled by the sometimes vast discrepancy between highest and lowest listed charges (>15% in the Hornady manual last time I bothered looking) and how many increments should be used between them. The answers so far tend towards the latter interpretation, as one would expect on the
Accurate Shooter forum, albeit taking as a given that there are many inexperienced members or visitors looking for basic advice here too. As
@ShootDots says, you can short-circuit a lot of work up on a cartridge you know well.
On the working loads up from scratch interpretation of the question, I bear very old advice from two excellent sources in mind. Sierra Bullets which used to say, use ~1% of max charge weight size increments, so 0.2-03gn in a 222/223 Rem size cartridge; 0.5/0.6gn in the SAUMs/WSMs etc. Also, for years and years, Hodgdon never listed starting loads, only maximum weights with the advice to reduce those by 6% to get a starting load with the exception of H110 where it was 2%. So, taking Hornady's 7th ed manual and the popular 165gn / H4350 30-06 hunting load, max 58.9gn, minus 6% = 55.3gn start, not the manual's 48.0gn (19% reduction!) Cross checking with current Hodgdon practice, its starting loads generally run 8-10% below max these days.
On fine tuning, having done the basic coarse work-up and satisfied with velocity / absence of pressure signs, I tend to fall in the 0.2gn camp for mid size cartridges, but often go down to 0.1gn for smaller numbers such as 223 Rem with heavy bullets, 6mmBR with 105-108s. Of course, you've got to be weighing to this level of accuracy which I've found I fail to do with beam scales .... so it's the tedious business of not only weighing each charge but doing so on lab quality electronic scales and adjusting charges down to a single powder kernel or two at most of most powder grades. On some cartridges, I've found that 0.1gn makes a significant distance - when I shot 223 in international level F/TR with 90s, I wonder just how many hours of my life were spent trying to achieve the single kernel of Re15 level of precision (1 kernel = 0.02gn). A far cry from loading 223 in the hundreds on a progressive press!
........... and let's not get into clever use of QuickLOAD and all these these three-alpha abbreviated techniques (OCW, OBT, OCD ??

) whose names / abbreviations alone confuse me.