I'm not quite sure what you're asking. If you don't maintain a constant velocity, retention times will vary. As the what that velocity is, for any reasonably "state of the art" column (say, 3.5 micron or smaller) and small molecule analytes, the efficiency vs velocity plots are relatively flat (for small molecules).
If you don't pack the column properly, you will get poor efficiency (but thst's a tautology!).
If you're familiar with the Van Deemter or the Knox equation: small-particle columns tend to have flat C terms and poorly packed columns tend to have large A terms.