Travel speed will obviously be dependent upon several factors. Key of which are volts, amps/WFS, electrode diameter, and bead width and depth.
The only way to determine it is to set all of your other variables and then time it and try to keep all other variables constant. 'Try' as some are going to change with electrode stick out, arc length, slight variations in weld width, etc. You get your most desired for final weld quality and profile and measure your inches per minute. Then you can easily determine the 25% increase/decrease which is often the amount you need just to perform normal welding as you will slow down or speed up during the welding to keep the weld sized as you want because of changes in openings, variations from previous passes, etc.
He Is In Control, Have a Great Day, Brent
You can pretty easily calculate what your travel speed "should be" and then verify by trials.
Assuming GMAW, FCAW or SAW (semi-automatic or machine) you can be very scientific about this and there are several advantages to this approach if you are doing anything in volume. But you didn't ask about that :)
http://weldingdesign.com/archive/pre-calculating-wire-feed-speed-travel-speed-and-voltage?page=2Deposition rate and weld size are the variables that will be most important in factoring. The link above breaks it down really simply.
This information is also in the Lincoln Procedure Manual and several other good publications.
Travel speed is a by-product of the restrictions in table 3.7, but the range should not be wider than +/-25% of the average value. So 7.5-12.5 is an acceptable range, but 6-14ipm would be too wide (10 ipm average in both cases).
As others have stated, the initial/average value it could be determined by welding some samples within the electrical parameter and bead size limits of table 3.7.