Now accepting guest editorial from SAP, Oracle, Workday and ServiceNow practitioners.Pitch a guide
The Sign-Off Table

Lesson 5 of 5

Testing derivation

9 min

Listen to this lesson

0:00 / 7:34

Profit center testing is a coverage problem. A posting with the wrong profit center usually raises no error, especially when a default or dummy profit center is in place. The only way to prove derivation is to run every process that creates journal entries and check the profit center on every line, not just whether the document posted.

Build a derivation coverage matrix

List the processes in scope and, for each, the line types they create. Then write the expected profit center for each line before you test. The table below is an extract of Contoso's matrix for plant 1710.

ProcessLineExpected profit centerSource
Purchase to payExpense to Detroit assembly cost centerCMC-FLEETCost center
Purchase to payGoods receipt of CMC-RTR-200CMC-FLEETMaterial, plant 1710
Plan to produceGoods receipt of CMC-ALT-100 from productionCMC-FLEETProduction order
Order to cash, fleet channelRevenue and COGS for Great Lakes Fleet ServicesCMC-FLEETSales order item
Order to cash, aftermarket channelRevenue and COGSCMC-AFTMKTSubstitution on sales order item
Asset accountingDepreciation of test benchCMC-AFTMKTAsset via cost center
PaymentsBank and payable linesSplit from invoiceDocument splitting
Manual journal entryBank charges accountDefault from FAGL3KEHDefault assignment

Derivation coverage matrix (extract).

Rare, complexAsset transfers, reorganizations: test each case once, carefully
Frequent, complexSales with substitution, intercompany: test every variant, highest priority
Rare, simpleOne-off manual postings: a sample is enough
Frequent, simpleCost center expenses, goods receipts: test the pattern, then monitor
Where to put derivation test effort.
The Sign-Off Table · thesignofftable.com/academy/sap-co-203

Check every line

After each test, open the journal entry and read the profit center and segment on every line, including the lines the system generated: cost of goods sold, price differences, GR/IR clearing, tax and receivables. For volume testing, query the Universal Journal for the test period and company code and list lines with a blank profit center or the dummy profit center. That list is the defect list.

  1. ExpectWrite the profit center for each line in the matrix
  2. RunExecute the process in the test client
  3. ReadCheck every journal entry line, including generated lines
  4. SweepList lines with blank or dummy profit center for the period
  5. ExplainEach hit becomes a defect with a root cause
A derivation test cycle.
The Sign-Off Table · thesignofftable.com/academy/sap-co-203

Pay special attention to processes that cross company codes. When Contoso Parts Distribution LLC (company code 1810) sells an alternator that plant 1710 produced, the selling side should land on CMC-PARTS and the supplying side on the profit center of the producing plant, with the partner profit center filled on both. Intercompany lines are generated by the system, rarely looked at by key users, and a frequent home for the dummy profit center. Add at least one intercompany sale and one intercompany stock transfer to the matrix, and read both company codes' journal entries line by line.

In an intercompany sale, each company code posts to its own profit center with the other side as partner, and both sides belong in the derivation test.
The Sign-Off Table · thesignofftable.com/academy/sap-co-203

Negative and change tests

  • Post an expense to a cost center that has no profit center in a test copy: the system should warn or block as designed.
  • Enter an aftermarket sales order in the fleet channel by mistake and confirm the result is predictable and visible in the drill-down.
  • Change a time-dependent profit center field from a future date and confirm earlier postings are unaffected.
  • Run one reorganization scenario in the test system and check the balance transfers.

Evidence for sign-off

The evidence package has three parts: the coverage matrix with document numbers and pass/fail per line; the sweep of the Universal Journal showing no postings to the dummy profit center or blank profit center other than planned, documented cases; and a profit center and segment report for the test period whose totals the process owner recognizes. For Contoso, the sweep runs after every mock conversion as well, because converted open items must also carry the right profit center.

What would you do?

Priya's derivation test shows every document posted, and a sweep finds 37 lines on CMC-DUMMY, all of them price differences from production. What should she do?

Key points
  • Test derivation by process and by line, including system-generated lines.
  • Write the expected profit center for every line before executing.
  • Sweep the Universal Journal for blank and dummy profit centers after every cycle.
  • Evidence: coverage matrix with document numbers, a clean sweep and balanced profit center reports.

Check your understanding · not graded

Lesson check

  1. 1. Which evidence best shows derivation is complete for a test period?

  2. 2. Which lines are most often missed in derivation tests?

  3. 3. Where should derivation test effort be highest?

  4. 4. A Chicago (plant 1720) sale of CMC-ALT-100 posts revenue to CMC-AFTMKT without any substitution. How is the test rated?