The count doesn't match. Before you write it off, ask why.
A cycle count variance is the difference between what a worker counted at a location and what Dynamics 365 believed was there. Dynamics 365 holds counts over the threshold for review instead of posting them, which gives a supervisor a decision to make. Most variances have a cause visible in that location's open work.
Written for warehouse supervisors. One of three seams covered under inventory reconciliation in Dynamics 365.
Five reasons a count doesn't match the system
Dynamics 365 holds a count over the threshold for review rather than posting it, so there is a real decision to make. Four of these five causes mean nothing is missing.
- 01
Picked but not confirmed
The stock has physically left the shelf and the work line has not been confirmed, so the system still counts it. The counter is right, the system is right, and the difference is a few minutes of lag.
- How it is found
- Open picking work against that location
- What to do
- No adjustment. Confirm the outstanding work and the variance disappears.
- 02
Put away in the wrong place
The most common real cause, and the most satisfying to catch, because the missing quantity is usually sitting one bay over. A variance that is mirrored by an opposite variance nearby is almost never loss.
- How it is found
- Short at the counted location, long at a nearby one
- What to do
- Move the stock to the correct location rather than adjusting either count.
- 03
Received but not put away
The goods are in the building but not yet at the location the system expects. They will be counted as missing anywhere they are looked for until the put-away completes.
- How it is found
- Open put-away work for that item
- What to do
- No adjustment. Complete the put-away work.
- 04
Counted boxes as pieces
A count entered in the wrong unit produces a variance that is a clean multiple of the case quantity. That multiple is the tell, and it is worth checking before anyone writes anything off.
- How it is found
- A unit of measure mismatch on the counting line
- What to do
- Re-enter the count in the correct unit.
- 05
Genuine loss
When the location has no outstanding work, no mirrored variance and no unit problem, the difference is real. Reaching that conclusion by elimination is what makes it credible to the person who has to approve the write-off.
- How it is found
- No open work and no offsetting variance anywhere nearby
- What to do
- Recount to confirm, then post the adjustment with approval.
What people ask about cycle count variances
What causes cycle count variances in Dynamics 365?
Most often, work that has happened physically but has not been confirmed in the system: stock picked from the shelf with the work line still open, goods received but not yet put away, or an item put away one bay from where the system expects it. Unit-of-measure errors produce variances that are a clean multiple of the case quantity. Genuine loss is what is left after those are ruled out, and it is the smallest category.
Should a supervisor accept a variance or recount?
It depends entirely on why the variance exists, which is the information the supervisor usually does not have at the moment of deciding. If there is open picking or put-away work on the location, accepting the count writes off stock that has not gone anywhere. If the quantity is short here and long one bay over, the answer is a movement, not an adjustment. If nothing explains it, a recount before the write-off is cheap.
Where does Dynamics 365 hold a count before it posts?
A counting journal whose variance exceeds the configured threshold goes to review rather than posting, so a person has to accept or reject it. That review step is a natural place for an agent to work: it already exists, a human decision is already required, and nothing changes in the system until that decision is made.
Can this run on the warehouse mobile app?
The count is taken in the warehouse app as it is today — nothing about that changes. What we are designing adds the explanation at the review step, delivered to the supervisor in Teams with the reason and the evidence, so the decision is made with the context rather than from the number alone.
How is this different from just counting more often?
Counting more often finds more variances. It does not make any one of them faster to explain, and the explaining is the work. A programme that doubles count frequency without addressing the investigation usually just doubles the queue waiting for a supervisor.