Modify configuration breaks on second try

When we are deploying new computers we push the image and run a modify configuration job on the computer to add it to the domain to a specified OU using the serial number for the computer name. Then after we push the software and know which prefix we are going to add to the computer we rename the computer in the console adding the prefix and then run another modify configuration job that changes the OU to the default OU and changes the computer name to the new name with the prefix using the token %NAME%. The reason we do it this way is we have too many prefixes to create jobs for each one. The first modify configuration job works correctly without any problem, but the second time we try to run the job to change the name to the token to %NAME% it changes the name in the console to %NAME% and the job fails because %NAME% is not a valid computer name. To make this odder, if you again rename the computer in the DS to the name you want, ie locNAB-Serial, and then just tell the failed job to restart, it works correctly.
I may have temporarily fixed this by changing the first modify configuration not to change the name as it uses the serial number anyways (previously the first modify also used the token %NAME%, now i just removed the name change). Now the second works fine the first time, but if it is run again, it actually trys to change the name to the actual token.
Has anyone seen this or have any idea how to fix this. It is quite a pain to have to run the same job twice.

I am running DS 6.9 with all the latest hotfixes, build 179.