Lesson 6 of 6
Testing physical inventory
8 min
Physical inventory tests are easy to get half right. A tester creates a document, types a count, posts, and sees "document posted." That proves the transaction runs. It does not prove that the posting block blocks, that the difference is valued at the right plant price, that the right accounts are hit, or that cycle counting selects the right items. Contoso's tests are built to prove each of those.
Prepare test stock you control
Differences are only readable if you know the starting point. In client 100, Priya's team uses a dedicated storage location for inventory tests, receives known quantities with distinctive numbers (for example 137 rotors), records the book quantity and the inventory G/L balance, and only then creates the documents.
| Test | Setup | Expected result and evidence |
|---|---|---|
| PI-01 surplus | Book 137 rotors, count 140 | 701 for +3; inventory debit $120; inventory difference credit |
| PI-02 shortage | Book 137, count 130 | 702 for −7; inventory difference debit $280 |
| PI-03 zero count | Item not found, zero-count indicator set | 702 for full book quantity; reason required |
| PI-04 posting block | Document with block; try a 201 issue | Issue refused until the count is entered |
| PI-05 recount | Difference above threshold; MI11 | New item; first count kept; only the recount posts |
| PI-06 plant valuation | Alternator surplus in 1720 | Valued at $156, not $150 |
| PI-07 quality and blocked stock | Differences in QI and blocked stock | 703/704 and 707/708 as appropriate |
| PI-08 tolerance refusal | Clerk posts a difference above their group limit | Posting refused |
| PI-09 reason code | Post 702 without reason | Refused if reason is mandatory |
| PI-10 cycle count selection | Products with known last count dates and indicators | MICN control list shows exactly the items due and overdue |
A physical inventory test set that covers documents, counts, differences and cycle counting.
- BeforeBook quantity per stock type; inventory and difference G/L balances
- ExpectedDifference quantity, movement type, value at plant price, accounts
- ExecuteCreate, count, recount, post with the warehouse and controller roles
- AfterMaterial document, journal entry, stock report, PI document status
- CompareAfter minus before equals expected; document completed; count date updated
Testing the cycle counting schedule
Cycle count selection is date logic, and date logic needs prepared data. Pick a handful of products, set their indicators, and make sure their last count dates are known (post a count for each with a chosen date). Then run the MICN control list for a window and compare it with what you calculated by hand: which items are due, which are within float time, which are overdue. Repeat after posting one count to see the planned date roll forward.
| Product | Indicator | Last count | Window | Expected |
|---|---|---|---|---|
| CMC-RTR-200 | A | Jan 6 | Feb 3–7 | Due |
| CMC-ALT-100 | A | Dec 15 | Feb 3–7 | Overdue |
| Packaging carton | C | Jan 6 | Feb 3–7 | Not due |
| Safety-critical bearing (fixed A) | A, fixed | Jan 31 | Feb 3–7 | Not due; indicator unchanged after MIBC |
Reconcile at the end
After the cycle, reconcile three things: the stock value of the test storage location against the inventory G/L account movements, the sum of difference values against the inventory difference account, and the number of physical inventory documents created against those completed or deliberately deleted. Open documents left behind by testing are themselves a finding, because the same thing will happen in production.
Finally, rehearse the real event. Before the first year-end count after go-live, Luis and the Chicago warehouse lead run a dress rehearsal in client 100 with the actual count sheets, the actual number of documents and the actual users. Time each step: creating 40 documents, printing, entering counts, recounting and posting. The rehearsal finds the practical problems (printer queues, missing authorizations for temporary counters, unit-of-measure confusion) that a functional test of three items never will.
What would you do?
PI-02 posts correctly, but the journal entry hits a general expense account instead of the agreed inventory difference account. All quantities are right. How should Priya classify it?
- Use controlled test stock with distinctive quantities and record the before state.
- Test surplus, shortage, zero count, recount, plant valuation and every control.
- Test cycle count selection against a hand calculation of due and overdue items.
- Reconcile difference values to the inventory difference account and close all test documents.
Check your understanding · not graded
Lesson check
1. Why receive a distinctive quantity like 137 before a count test?
2. How do you prove that a posting block works?
3. What should a cycle count selection test compare the MICN control list with?
4. Quantities are right but the journal entry uses the wrong account. Result?