Lesson 6 of 6
Testing sourcing outcomes
8 min
Sourcing tests prove that the right demand reaches the right supplier, at the right price, under the right agreement. The defects they find are expensive and quiet: a contract that is never used, a quota that sends everything to one supplier, a blocked supplier that still gets orders, a price that falls back to the old info record. None of these produce error messages; they just cost money.
Contoso's sourcing test set
| Test | Setup | Expected result and evidence |
|---|---|---|
| SR-01 RFQ to award | RFQ with three bidders; award Harbor | Comparison shows landed costs; follow-on contract created with $40.80 |
| SR-02 release order | PO for 600 referencing contract 4600000021 | Price $40.80 copied; released quantity 600 in release documentation |
| SR-03 contract target | Release orders beyond 7,200 | Message per design (warning) and buyer decision |
| SR-04 expired contract | Requirement dated after validity | Contract not proposed |
| SR-05 scheduling agreement via MRP | Rotor demand in plant 1710 | Schedule lines in 5500000310 for Lakeside's share |
| SR-06 releases | ME84 or scheduled job after MRP | Forecast and JIT releases created and output; history visible |
| SR-07 quota split | Series of requirements over a month | Allocated quantities approach 70/30 |
| SR-08 blocked source | Manual PO to Northfield | Refused |
| SR-09 source list requirement | PO to a supplier not on the list | Refused |
| SR-10 ambiguous sources | Two valid info records, none fixed, no list | Requisition without automatic assignment; buyer chooses |
Tests that prove sourcing master data and agreements work together.
Evidence that the expected source was used
For every requisition or PO in a sourcing test, record the supplier, the agreement number and item it references, and the price and its origin (the condition record from the contract or scheduling agreement, not the info record). The PO's item shows the outline agreement reference; the contract's release documentation shows the PO; the quota arrangement shows the updated allocated quantity. Three views that agree are strong evidence.
Testing quotas and dates
Quota tests need a series, not a single requirement. Run MRP on a set of requirements in client 100 (or create requisitions with automatic sourcing) and check the allocated quantities after each step against a hand calculation. Reset or record the quota base quantity first, because existing allocations change the result. Also test the year boundary: create requirements on December 31 and January 1 with the old and new source lists and agreements, and confirm that each day uses the source valid on its date.
- Dec 31Old sourcesRequirement uses 2026 Lakeside info record at $42.00
- Jan 1New sourcesRequirement uses scheduling agreement 5500000310 at $41.50 or Harbor contract at $40.80 per quota
- Jan 2CheckNo requisitions without source; no POs at the old price
- Jan 31ReviewSpend with contract reference versus without; quota split for the month
Plan the test data so that each case has exactly one right answer. If client 100 still contains last year's rotor contracts, a leftover Lakeside info record flagged for automatic sourcing, or quota allocations from earlier test cycles, results become hard to read and harder to repeat. Before the sourcing cycle, Priya's team lists every info record, source list entry, quota and agreement for the test products and plants, removes or blocks the leftovers, and records the starting allocations. That list becomes part of the evidence: it shows the conditions under which the expected sources were proven.
What would you do?
SR-02 creates a PO to Harbor for 600 rotors, but the price is $42.00, not $40.80, and the PO item shows no contract reference. What is the most likely cause?
- Test RFQ award, release orders, contract limits and expiry, scheduling agreements and releases.
- Prove the agreement reference and price origin, not just the supplier.
- Test quotas as a series against a hand calculation; test blocked and unlisted sources.
- Test the validity boundary when sourcing data changes, especially at year end.
Check your understanding · not graded
Lesson check
1. What is the strongest evidence that a release order used the contract?
2. How should a quota arrangement be tested?
3. Which test proves a blocked source list entry works?
4. Why test requirements on December 31 and January 1?