Physical inventory question

So I have not used the Epicor physical inventory before. We want to do a count coming up in about a month. I have no confidence that our inventory procedures are that close to reality. I think overall, the on hand quantities are probably close, but the locations I’m pretty sure are hosed. I’ve tried messing with the cycle count system that epicor has out of the box, and it was overly complicated and too hard to use.

Is physical inventory similar to that?

I’m thinking about simply having everyone use count tags, turn them in then make an entry screen for a UD table (train some monkeys to do the data entry). Two and a half years ago we went live with E-10 and actually had live inventory. Before that, it was walk around and see if you had it. So when they did counts, it was just to get a dollar amount. However, the tag system I am talking about would be what they are used to, so no retraining people there. They had an access database for keying in the tags. I would just replace that with a UD table entry screen.

Once I had the data into the database, I would attack the reconciliation on the parts with the most off on the on hand and just doing adjustments with inventory adjust. I’ll get as much as I can done until we have to start again, prioritizing with part numbers that have planned usage first, then fix the ones with no usage after they start up the shop again, and just get as much done as we can.

Does this sound like a better plan then trying to use epicors physical inventory? At least until we get a better handle on inventory? I’m worried that the epicor process is too rigid and to transaction heavy to work for a poorly organized facility.

1 Like

A “Full Physical” is just a special cycle count but functions the same, using the same screens.

We use the Epicor functionality and it works very well. A slightly customized count ticket is used, and I have this really cool Cycle Count tracking Dashboard to show things in ‘real time’. We have a solid process with solid instructions. Willing to share it all since it’s all pretty simple once you get your head around it… Let me know if you want to talk more.


I’d love to see those instructions. I can’t understand epicors documentation well enough to make understand what it needs.

Let me ask you this, say we have well over 200k+ part numbers in the system. A large majority of them are non-stock and we do not have them. Some we do. Some are on the shop floor but not in the database. How hard is it to be able to make that work? If I include all the parts, we have a tons of tags we don’t need. If I only include stock parts, that probably only 1/4 of out inventory. If I only include parts On Hand, how do parts with 0 OH but actually on the shelf get captured? Also, if the bin locations aren’t right, do the tags have specific bin locations on them? How do we pick up things not put away correctly?

Sorry, this is a flood of questions. Answer whatever you feel like you have time for.

I’m interested too. We don’t have that many parts, but we have lot tracked UOMs, which makes a lot of parts. Printing Tags is not really practical for us, and to be honest I have not yet been able to spend enough time on this yet to get a good solution for our Ops team, so any real live experience would certainly help.

Personally, I did not enjoy the epicor implementation of an inventory count. To me (opinion ahead), it had way too many steps and it wasnt very intuitive. I ended up writing a completely custom solution for it that was well received by my employer at the time… in fact they recently emailed me about it since their yearly is coming up. My solution, was an external app that interfaced epicor and employed many laptops and mobile devices to do the scanning. It essentially stored our scans in a UD, then provided a nice interface to monitor the process. As discrepancies appeared, teams were sent to recount simultaneously while other counting was still going on. I kept a record of up to 3 counts (the oldest was overwritten as needed). When done, a report was generated to show the differences (and costs). When admin pulled the trigger, it automatically went and updated all of the locations (And generated a report to show the changes).

Here is the actual scanning process


In a word WOW! That is fantastic!

Great contribution Chris.

What about Multi Site and warehouse? How would it behave if you had all sites running a count at the same time?


Number one silly thing that will trip you up, is that when trying to locate the count group for the Phys Inv, is you have to select the checkbox to include physical inventory

  1. I find the Full Physical to be much simpler than cycle counting.

  2. Don’t print tags for zero QOH parts. Just a print a dozen blank tags. And you can print more blank tags if you need to


@ckrusen, I hate that checkbox.

Physical inventory is a lot less work to setup, the initialize physical inventory process does most of the work.

Compared the Vantage, the posting process allows you to select the adjustment posting date.

1 Like

It doesnt matter about sites, since each location is a WH and bin, and UD records track the company

1 Like

What happens if for some reason you had the same warehouse code in multiple sites? Thankfully we don’t.

One more very important part… Make sure the count date is correct.

We had one scheduled for 11/30/17, but the person mistakenly set the count date to 11/3/17.

So all the adjustments created were to get the QOH to match our counts. But it was trying to get the QOH as it was on the 3rd, to match the counts done on the 30th.

For example,

  1. 11/3, the system QOH for a part was 100, actual qty was 99
  2. 11/15, 50 pieces issued, QOH -> 50, actual qty ->49
  3. 11/30, 49 are counted and entered into Count Entry

This creates a Qty-Adj of -51. Because it was expecting 100 (on the 3rd), and you told it that it was only 49. And it sets the date of this Qty-Adj to 11/3.

If you then check QOH after posting the count, it will be -1. Because those 50 issued on 11/15 are subtracted from the 49 that was the QOH after posting.


"What happens if for some reason you had the same warehouse code in multiple sites? "

I don’t think you can have the same warehouse namecode even if it is two different sites.
The PartBin table doesn’t have a field for the Plant (aka site). So the only way it could keep the inventory straight, is by having entirely unique warehse codes.

That’s good to know. I have inherited this implementation so a lot of this was setup prior. I have a lot to learn.

Appreciating everyone’s contributions.

Here’s a doc I recently sent our new controller. It’s the basic step-by-step instructions (and some notes) of doing a Phys Inv in E10

There are some specifics to our company (multi-site, and the references to a custom report styles we used for the tags and the Count Variance Calculation Report)

Hope it helps

E10 Phys Inv.pdf (338.3 KB)

Edit: That doc is just a cut and paste from an email I sen, and not intended to be a descriptive step-by-step for all variations. It’s just what works for us.


The doc provided the other day is really good. I might replace mine with that one! Here are two docs that we use as reference. The first one is our cheat sheet (technically it’s from E9, but it’s the same process). The second is one I picked up along the journey a while back. Not mine, so credit goes elsewhere. I’ve also attached our dashboard for you to look at and tweak as needed of course.

Regarding your questions starting with “200k” parts… Let me start with, I am not in accounting or material handling. Comments below are “to the best of my knowledge” and should be consumed appropriately.

The key will be how they are set up and which ABC code is applied. You can set up ABC codes and exclude codes from cycle counting. We exclude all of our “C” parts, many of which we do not stock because we purchase as needed. Non-stock and non-quantity bearing would also help you weed out a number of them. Even if they are stock/qty bearing, if there is zero on hand, you do not have to print those count tickets - but you should if you believe there might actually be qty on the floor. I’m not directly in the count/verification/audit end of the cycle counts, so I’m not exactly sure how they guys deal with all of ‘those’ parts, but they’ve never mentioned any problems.

Parts that are on the floor but not in the database are a problem. Accounting rules see those as assets and therefore must be included in the value of inventory unless they’ve been outright expensed upon receipt. We treat a number of items this way - mostly small stuff, but a number of consumables are done this way as well. If they have been expensed, then no worries as they do not qualify for a the physical count.

In order to properly count anything, you need a tag. Might be a combination of ABC groups with C = no count, B = small group of tags that you have qty, but the system shows none on hand (or some other strange situation), and A being the stuff you regularly count.

Tags do include bin locations, and will have multiple tags if the same part# has qty across bins (if I recall correctly). As for ‘finding’ inventory in other bins - we just mark the tags with the new bin location and the folks who enter the count tags can adjust during data entry, or make an inventory move transaction right after it’s all done.

The key is to minimize your variance report in the end. Depending on a lot of factors, your variance report and it’s $$ value could equal the “change in inventory” charge that will appear in the GL. A big one of those is never something the auditors want to see. It begs the question of overall inventory control/procedures and might equate to increased auditor requests for spot checking inventory during the physical count audit and possibly during the year. You didn’t mention auditors, but we have them so I added that bit just in case.

If this isn’t clear or you just want to shoot the breeze some more, PM me and we can swap digits for a chat.

AshCycleCountMonitor_wBAQs.dbd (782.9 KB)
Steps for cycle counting and full physical inventory in 9.doc (145.5 KB)
Cycle count process(EUG).docx (3.5 MB)


:laughing: a dozen? That’s hilarious. We would probably need hundreds if it’s for parts not where they are supposed to be. Our location control is horrible, mostly due the fact that every single shelf, nook, and cranny has it’s own bin location. That’s what I’m most worried about, if we are way they hell off, is this process going to be manageable.

The instruction really help, and it makes it look a lot simpler than I was thinking. I’ll run a test in pilot and see if the process makes sense.

One question. There is the start date mentioned a bunch. If we were to take the count, can we start up again before the counts are entered and posted? It sounds like it will be able to make the change based on what it should have been at the time of the count, even if you have been running. Is this a correct assumption? That would be pretty useful if it is the case.

1 Like

As for running while counting - we do not. We take the day and have the guys on the floor do the counts since they know where everything is/supposed to be. Only takes about 4 hours, but we have like 8 crews counting and 6 auditors spot checking/etc behind them.

Now, I believe you could start running again, but I’m not sure how any sort of audit would work. We do the audit to make sure the counts are correct - you do not NEED to do this, it’s just part of OUR process. The process of marking it as a ‘full physical’ creates a table of frozen qty counts and then the variance reports uses this, plus the new counts, plus any PartTran entries to create the true variance - as far as I know. I would really like someone else to confirm this though.


I believe so. But don’t dare think about recounting once you’ve let people start taking inventory.

In a nutshell, the Count date is used to capture the expected QOH. And the posting process (after Count Entry), calculates the difference, and creates a Qty-Adj based on that diff.

So in Qty Adjust screen, you enter the change in Qty. Ex: entering -10, for part/bin that has a QOH of 100, reduces the QOH to 90.

In a Phys Inventory, you enter 90 as the count for a part/bin the expects to be 100, and it does the -10 Qty Adj to get the QOH to 90.

1 Like

Currently it takes us 2 days with our current system. Normally occurs on a Friday so we only lose 1 day of production/shipping then on a Saturday to fix mistakes, manual reads where the scanners have flaked out, rechecks because you can’t read someones handwriting, or the hunch where the stock controller knows like heck we have more that that was reported, they go an do a double check.

Finally, thanks for posting that information it is very useful.

1 Like

This was really helpful. The “include physical inventory” was particularly helpful, since I couldn’t find that anywhere else. Unfortunately, printing tags resulted in a HUGE number because we serialize a lot of parts. Not sure how to deal with that.

1 Like