Have a client who will do a setup and run say 10 jobs for 10 different parts. However, the setup is the same for the 10 jobs. The combinations can be different all the time.
How do you handle spreading the setup time cost over the other jobs so the first job doesn’t get over costed and the others under costed?
Keep in mind, currently the setup process and time is set on each part. They never know which will be first or used on a given day or time sequence before a new setup
Do you care about estimating cost or just job profitability?
If the answer is profitability, your folks on the floor can clock the ‘setup’ for each of the 10 jobs at a reduced time. Say each job takes 60 minutes but you are doing 10 jobs, so clock each at 6 mins - the math will vary, but it’s still simple math. The system will calc the actual setup cost at the reduced value based on actual setup time clocked for each job, and you can vary the time they clock for each job as you see fit.
If the answer is costing (or both), then that is a whole different problem. 
There’s job batching, which will let you merge your setup op before dovetailing out into the N jobs you’ll be running. That requires the advanced production module, which is an extra purchase which pretty much does job batching and co-parts. We’re a bigger operation, so that investment would be paid back in reduced labor and clarity in short time. Your mileage and operating budget my vary.
If you can’t swing $4K-ish, my second inclination is to break set the setup op as a dummy MOM that only covers the setup time. Add the dummy part to the real MOMs (none which will have setups themselves), link the dummy job directly to the real jobs. End result is 11 jobs (10 real, 1 dummy) with one block of setup.
Third option is skipping all that stuff and amortizing the typical setup time across all the jobs. If it usually takes an hour, and you’re usually batching 10 jobs, give each job an estimate of 6mins. Sometimes that will be right, sometimes it will be wrong, over the course of the year it will be about right.
Which option is right for you depends on what level of accuracy you need and what kind of workload or investment you are willing to add to achieve it.
Thanks all. Very helpful