A Tibetan syllable is a stack, and the stack has slots. Reading the writing means knowing which slot each letter is sitting in, and the slot after the root is the suffix, the rjes jug.
Ten of the thirty letters may take that position: ga, nga, da, na, ba, ma, the small a, ra, la and sa. Two of those ten, da in the classical language and sa still, may be followed by a further letter in a second suffix slot. Everything else is barred from the end of a syllable, so a reader who sees a letter after the root already knows it belongs to a set of ten and can rule out the other twenty at once.
The suffix does far more than close the syllable. In the modern language it reshapes the vowel in front of it, so that the same written vowel is heard differently depending on what follows, and a syllable therefore cannot be read letter by letter from left to right and assembled at the end. It also carries grammar: many of the case particles and verb endings of written Tibetan are suffixes or begin with one.
Between one syllable and the next stands the tsheg, a small raised dot. It is not a space and it is not a hyphen; it marks where one stack ends and the next begins, and without it the string of letters would be unreadable, because Tibetan does not otherwise separate its words.