Lesson 1 of 5
Coverage
10 min
Listen to this lesson
Controlling is mostly downstream. Very little of what a controller reports is entered by a controller: costs arrive from purchasing, production, sales, payroll and asset accounting, and controlling's job is to put them on the right objects and turn them into product costs, variances and margins. That makes controlling test coverage different from most functional areas. You cannot test it on its own. Each controlling scenario has to be linked to the logistics or finance process that feeds it, and the test has to follow the value from the source document to the controlling report.
Four families of scenarios
For Contoso's system conversion the controlling scope is roughly forty scenarios across those families. The number matters less than the link to source processes. Priya's team builds a coverage matrix with controlling scenarios as rows and feeding processes as columns. A cell is marked when the end-to-end test for that process checks a controlling result, not only that the logistics document posted.
| Controlling scenario | Feeding process | Controlling result to check |
|---|---|---|
| Cost center expense | Purchase-to-pay (non-stock PO, invoice) | Expense on cost center from the PO account assignment |
| Activity allocation to orders | Production confirmation | Detroit Assembly credited, order debited hours × $52 |
| Standard cost of CMC-ALT-100 | Costing run before the period | Released standard $150 in 1710, $156 in 1720 |
| Production variance | Make-to-stock production with extra scrap | Variance by category on the order, settled to finance |
| Margin by customer | Order-to-cash for Great Lakes Fleet Services | Revenue and COGS on the right market segment and profit center |
| Intercompany margin | 1710 sells to 1810 Toledo | Internal revenue and COGS in both company codes |
An extract of Contoso's controlling coverage matrix.
Where the scenarios come from
Start from SAP Best Practices content in SAP Signavio Process Navigator, where scope items describe standard processes and their test scripts, and from Contoso's own process design, which for a system conversion means the ECC processes that stay. Then add the controlling-specific cases standard scripts tend to skip: period-end steps, negative tests, reconciliations, and anything that differs because of the conversion, such as cost elements now being G/L accounts or margin analysis running next to costing-based CO-PA. SAP Activate expects formal, documented tests for end-to-end processes and user acceptance, so record each scenario as a test case in SAP Cloud ALM with its expected controlling result.
Test levels for controlling
Controlling cases appear at every test level, but they prove different things. In unit and string tests the configuration team checks single settings: a costing sheet, an allocation cycle, a derivation rule. In integration tests the end-to-end processes run with controlling checks added, and period-end runs for the first time on realistic volumes. In user acceptance controllers and key users run their own month: they post, close, report and reconcile, and they judge whether the numbers make sense to the business. For a system conversion there is one more angle: the comparison of results before and after the conversion on the same data, which the next course in this track covers in detail.
| Level | Who runs it | Controlling question it answers |
|---|---|---|
| Unit / string | Configuration consultants | Does each setting do what the design says? |
| Integration (SIT) | Process teams with Priya's test team | Do values flow from source documents to the right controlling objects? |
| UAT | Dan's controllers and the key users | Can the business run, close and report a month and trust the numbers? |
| Dress rehearsal | Cutover team with controllers | Does the close run in order and on time on production-like data? |
Controlling across test levels at Contoso.
Write entry and exit criteria for controlling at each level. A sensible entry criterion for UAT is that every derivation defect from SIT is fixed and retested, because UAT testers who meet wrong profit centers stop trusting every other result. A sensible exit criterion is two consecutive test periods where margin analysis reconciles with the trial balance and allocations balance, with no open high-severity controlling defect.
What would you do?
Midway through SIT, the order-to-cash team reports all its tests passed. Dan notices none of the test cases check the market segment or profit center on the billing document. What should Priya do?
- Controlling is downstream: every scenario links to the logistics or finance process that feeds it.
- Cover four families: overhead management, product costing, period-end and margin analysis.
- A coverage matrix marks where an end-to-end test checks a controlling result, not just a posting.
- Spend depth by risk: product costing, settlement, margin analysis and intercompany get the most.
Check your understanding · not graded
Lesson check
1. Why must controlling tests be linked to feeding processes?
2. What marks a cell in Contoso's coverage matrix?
3. Which scenario deserves the deepest testing after a conversion?
4. What is a sensible controlling entry criterion for UAT?