Stock or Not to Stock, That is the question - HMLV

Hi Everyone,

So we are currently a high-mix, low-volume shop that builds complex assemblies. Such as it can take up to 134 sub jobs to make the final top level job.

Right now we are a mix of buy/make to stock and to the next job.

We are in a constant discussion of if one method is better than another and we wanted to reach out to others who might be in the same situation and learn from them what is working / what their struggles are to see if we can help each other.

We have a similar environment with some manufactured components built to stock and other we cannot.
Instead of Job To Job, we us Pull as Assembly within the method (BOM) and create sub-assemblies for the manufactured components. Less jobs to maintain for us and less planning.


We moved to all make to stock about a year ago. Make direct and job to job do not consider other sources except what they are linked to and if we are building subs on several different jobs that sourcing can change. We use multi-level pegging to align all of the moving demands and sources.

1 Like

Highly recommend - make to order.
Go into Company config to set the job number to be set to
Order-Line-Release - this will make it easier to track.

Setting the 124 subs into Subassemblies allow you to contain all of your work into one job. Depending on the subassembly - some may be phantoms boms to flatten the job tree some.

If you must/want to make direct to the job- do so by using planning workbench and create related job numbers which will pull in the order-line-release and you can append -01, -02 or -A, -B, etc to keep all jobs in a family.

1 Like

Hi Bryan,

Thanks for the response.
We have tried doing this is the past (before I worked here) but the feedback I had got was that they lost visibility into the subassemblies and their demand and there was also a lack of a requirement to issue materials to them as there was nothing stopping them from moving it up to the next level.

Have you experienced anything like that? Have you created a way to have more visibility into those subassemblies?

Hi Bryan,

I am curious. Are your jobs long running? Do you setup the method with all materials required at the start of the job?


If you build everything to stock, you do lose visibility to the specific supply for a demand. Although if you are planning and executing by dates that should not matter (unless the same part number can be different configurations).

If you use MRP and auto firm I would not suggest subassemblies having separate jobs. If a child job is firmed before a parent there is a great risk of corrupt data. Epicor will give you a data fix if this happens. We tried to prevent it by having a separate planner for the child jobs and not using auto firm for that planner. Jobs still fell through the cracks here and there - so we got extra jobs to make the same parts because the unfirm parent number changed, or worse, operators got locked into child jobs and couldn’t clock into anything else because of corruption on the job there were on.

From experience, I’d rather support subassemblies within the job if they are specific to a parent part. If the subassemblies can be swapped out as needed, then make them to stock as a separate job.


@Nancy our jobs are short; from one shift up to a week long. We try to backflush materials and have that system working fairly well for us. we just need to make sure we have materials in the right place.

1 Like

@awissmann we solved for the material issue by backflushing. Less transactions for us and less mistakes to fix. Backflushing doesn’t work for everyone, but it does here.
To your question about losing visibility to the sub assemblies, we actually prefer the sub assemblies because it’s easier to track than separate jobs. gets a little tricky when the traveler needs to be in two separate areas in the warehouse. it’s rare and we deal with it.
We have plans to move to better job operation scheduling and the MES work queue to solve for most of that.