Variations in bullet length do not directly cause OAL variation. (FYI, neither does case length variation.)
To figure out what can cause OAL variation, you have to remember that bullets are seated by pushing them down using a contact point below the tip (so the meplat isn't crushed), somewhere along the curve of the bullet's front section (the ogive).
A deformed meplat (or just a piece of extra material on an edge of the hollow point) can force two otherwise absolutely identical bullets to have different OALs. Your caliper will measure one from base to tip, the other from base to deformed tip. Your seating die ignores the bad tip because it never touches it, but your caliper includes the damaged tip in its OAL measurement. For ALL practical purposes, these two otherwise identical bullets have actually been seated identically.
The second cause is a slight difference in the ogives of the two bullets. This will force your seating die to contact the bullet at slightly different points - think slightly different heights. So even though the die pushes down exactly the same amount, one OAL will be longer than the other. These bullets are NOT seated identically; however, for most folks small differences don't matter much if at all.
The reloader aiming for top precision will not use OAL measurements. Seating depth is controlled by measuring from the base to to a point on the ogive very close to where the bullet reaches full caliber size. This measurement is often called CBTO or cartridge base to ogive. The theory here is the fat part of the bullet will always be seated the same distance from your barrel's lands . . . which is considered by many to be the most important control measurement. IOW, the shape of the ogive, the condition of the meplat, and OAL are ignored.
TLDR Unless you control cartridge length by CBTO, you will always have OAL variations because bullets are not identical in shape. Controlling CBTO requires tools that allow your caliper to measure CBTO (duh). Those tools were mentioned in previous posts.