SAP Cost center

A cost center answers "where was this cost incurred?" It is a responsibility area to which expenses are posted so managers can see and control their spending. Examples: the maintenance department, the finance team, a specific machine group.

• Costs are posted or allocated to cost centers.
• Managers monitor actual versus planned costs per cost center.
• Overhead is allocated from cost centers to products or projects.
• Cost-center reporting supports internal management accounting.

A cost center captures where costs occur; a profit center captures responsibility for profit (revenues minus costs) for a business area. They are complementary CO objects, cost centers for cost control, profit centers for profitability analysis.

The clearest way to hold it: a cost centre answers who spent it, a profit centre answers which part of the business it belongs to. A cost centre only ever receives costs. A profit centre receives costs and revenues, so it can produce something resembling a profit and loss for a segment of the business that may cut across legal entities entirely.

Most postings carry both: the cost centre is entered or derived, and the profit centre is derived from the cost centre. That is why a wrong cost centre is also a wrong profit centre, and why the divisional numbers stop reconciling.

Version note: in S/4HANA both live in the Universal Journal, so a posting carries its cost centre and profit centre on the same line item rather than in separate FI and CO tables. This is the change people notice most, because reconciliation between FI and CO stops being a task. The concepts, the allocation methods and the master data are unchanged.

Collecting costs is the easy part. Moving them to where they were actually consumed is the discipline, and there are three mechanisms with real differences.

Distribution moves primary costs and keeps the original cost element. The receiver sees that they were charged for electricity, specifically. Transparent, and it produces more line items.

Assessment moves costs under a secondary cost element, so the receiver sees a single charge for facilities rather than its components. Simpler to read, and the receiving manager cannot see what makes it up, which is either a feature or a complaint depending on who you ask.

Activity allocation is different in kind. A cost centre provides an activity at a rate, machine hours or labour hours, and a consumer is charged for what it used. That is how a production order picks up the cost of the time it consumed, and it is the mechanism behind manufacturing absorption.

All three run at period end, in a defined order, and the order matters: allocating before all costs have arrived allocates the wrong number, and rerunning is not always clean.

The choice between distribution and assessment is a reporting decision rather than a technical one, and it is worth asking the business rather than choosing by habit.

Half an hour, and it covers the whole idea.

1. Post an invoice with FB60 to an expense account, entering a cost centre.
2. Run KSB1 for that cost centre. The cost appears as a CO line item referencing the FI document.
3. Try posting the same expense account without a cost centre. It is refused, because a cost element needs a CO object to carry it.
4. Create and run a distribution cycle moving that cost to two receiving cost centres.
5. Run KSB1 on all three. The sender is credited, the receivers debited, and the original cost element is preserved.
6. Check FB03 for an FI document from the allocation. There is none, because no money left the company.

Step six is the distinction people find surprising and it is the heart of controlling: internal movements are real in CO and invisible in the ledger.

A cost centre that only collects actual costs answers one question: what was spent. That is bookkeeping. The reason controlling exists is the second question: what should have been spent, and why the two differ.

So each cost centre carries a plan as well as actuals, entered by cost element and by period. The comparison between them is the standard cost centre report, and reading one is a skill worth practising, because most of what it shows is not an error.

Three shapes of variance, and what each usually means.

Price variance. The activity or the input cost more than planned. Often outside the manager's control, which matters when the report is used to judge them.

Quantity variance. More hours or more consumption than planned for the output achieved. This is usually the one worth investigating.

Timing. The cost landed in a different period than the plan assumed. Not a variance at all, and the most common cause of an alarming looking report in month one.

The consultant's job here is less about configuration than about making the report answerable. A cost centre receiving allocations from four other cost centres shows a total its manager cannot explain, and the fix is either fewer allocations or a report that shows received costs separately from own costs. The second is usually right, because the allocations exist for a reason.

Cost centers are master data, organised in a standard hierarchy that mirrors the organisation. Good cost-center design makes management reporting meaningful; poor design makes it noise.

Two fields on the master record do more than they appear to. The category controls what may post to it, which is how a statistical cost centre is prevented from receiving real costs. And the validity period means cost centres are date-dependent, so a reorganisation creates new ones from a date rather than editing old ones, which is what preserves last year's history.

Two fields on the master record decide more than they look like they should. The validity period means a cost centre exists only between two dates, so postings outside that window fail with an error that names the cost centre and not the real cause. And the lock indicators allow a cost centre to accept actual postings but not plan, or neither, which is how you retire one without deleting history.

Deletion is generally not available once anything has posted, and that is deliberate. A cost centre with history is part of the audit trail. Retiring it means locking it and setting the validity end date, and the cleanest landscapes do this every year rather than accumulating a list nobody dares touch.

• Confusing cost centers with profit centers.
• Too granular or too coarse a cost-center structure.
• Postings to wrong cost centers distorting management reports.

• A hierarchy that mirrors the org chart exactly. Org charts change annually and cost history does not move with them.
• Allocation cycles nobody reviews. A cycle built three years ago still allocates to cost centres that closed.
• Assessment where the receiver needed detail. See profit centres, transaction data and the business partner for the surrounding objects.

The habit that prevents most of these is asking, before creating anything, who will be shown this cost centre's report and what decision they are expected to make from it. A cost centre nobody is accountable for is a bucket, and buckets grow.

Posting to a cost centre is trivial, and designing the hierarchy and the allocation cycles so managers see costs they recognise is the part you do in the course.