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

Lesson 1 of 5

Types of SAP data

9 min

Listen to this lesson

0:00 / 10:07

Ask a tester why a test failed and the most common honest answer is "the data". The customer was blocked, the product had no sales view, the stock was used up by someone else, or the price condition expired last month. Before you can plan data, you need a clear picture of what kinds of data an SAP system holds, who creates each kind, and how each kind reaches a test system.

SAP practitioners usually sort data into four types. They behave very differently in a test landscape, so the first job of a test data plan is to put every data item a scenario needs into one of these boxes.

Configuration (customizing)Rules the system follows: document types, pricing procedures, account determination, number range intervals. Reaches QAS by transport.
Organizational dataThe skeleton: company code 1710, plants 1710 and 1720, sales area 1710 / 10 / 00, purchasing org 1710. Defined in customizing and transported.
Master dataLong-lived business objects: business partners, products, prices, bills of material, routings, G/L accounts, cost centers. Created, migrated or copied.
Transactional dataShort-lived documents that record events: sales orders, deliveries, purchase orders, goods movements, journal entries. Created by running processes.
The four types of SAP data, from most stable (top) to most volatile (bottom).
The Sign-Off Table · thesignofftable.com/academy/sap-tst-206

Configuration and organizational data

Configuration (SAP calls it customizing) is maintained in the development or customizing client and moved to quality assurance (QAS) and production through the transport system. Organizational units such as Contoso's company code 1710 and plant 1720 Chicago are defined in customizing, so they travel the same way. Testers do not create either of these in QAS. If a test needs a new shipping point or a new pricing condition type, that is a configuration change with its own transport, not a data task.

This matters because many "data defects" are really configuration gaps. If QAS has no number range interval for a new document type, no amount of master data will make the test pass. Learn to ask: is the thing missing a record someone enters, or a rule someone configures?

  1. Does the error point to a rule, such as a number range, document type or determination?Yes → configuration gap: raise a change and transport it · No → next question
  2. Is a master data record or view missing, such as a sales area or plant view?Yes → master data: fix it at the source, often the migration object · No → next question
  3. Is a balance too low, such as stock in 171A or a credit limit?Yes → transactional: post or reset it in a setup step · No → next question
  4. Does the test need a document that no longer exists or is already used up?Yes → create it by running the preceding process step · No → investigate as a possible system defect
Routing a "data problem" to the right owner by asking which type of data is actually missing.
The Sign-Off Table · thesignofftable.com/academy/sap-tst-206

Master data

Master data describes the things a business deals with repeatedly. In SAP S/4HANA, customers and suppliers are both business partners, with roles that add the customer or supplier views. Products (still often called materials) carry views for sales, purchasing, MRP, accounting and more, at client, plant and sales organization levels. Conditions hold prices and discounts. Finance has G/L accounts, cost centers and profit centers.

Master data is the data testers depend on most and the data most likely to be wrong. A product can exist at plant 1710 but not at plant 1720. A business partner can have the customer role but no sales area data for 1710 / 10 / 00. Each of those gaps fails a scenario in a different, confusing way.

On a migration program, most QAS master data arrives through migration loads, so a data defect in test is often a migration defect in disguise. When Contoso's test of a Chicago order fails because CMC-ALT-100 has no sales view for plant 1720, the right owner is the migration object or its mapping, not the tester who hit the error. Record the defect against the source so the next load, and the production load, are fixed too.

Transactional data

Transactional data is created by running processes: Contoso's sales order for 50 alternators CMC-ALT-100, the outbound delivery, the goods issue and the billing document. Two properties make it tricky for testing. First, it is consumed state: once an order is fully delivered, it cannot be used again for a delivery test. Second, it changes balances such as stock and open items that other tests rely on.

  1. Stock on hand50 pieces of CMC-ALT-100 in storage location 171A, posted in setup.
  2. Sales orderGreat Lakes Fleet Services orders 50; credit exposure rises.
  3. Delivery and goods issueStock in 171A drops to zero; the order is now fully delivered.
  4. BillingAn open customer item is created and the journal entry posts.
  5. Consumed stateThe order cannot be delivered again and the stock is gone: the next run needs fresh data.
One sell-from-stock run uses up its transactional data and changes balances that other tests rely on.
The Sign-Off Table · thesignofftable.com/academy/sap-tst-206
Data typeContoso exampleHow it reaches QASCan testers create it?
ConfigurationSales document type, pricing procedureTransport from developmentNo: raise a change
OrganizationalPlant 1720 Chicago, storage location 172ATransport (defined in customizing)No
Master dataCustomer Great Lakes Fleet Services, product CMC-RTR-200Created, migrated, or copied with the client or systemUsually yes, with the right role
TransactionalPurchase order for 500 rotors, opening stockCreated by running the process, migrated, or copiedYes, by executing transactions

Who owns each type of data in a test landscape.

What would you do?

Maria, Contoso's order-to-cash key user, reports that her test of a new rush-order type fails in QAS with a message that no number range is assigned. She asks Priya to "create the missing data". What should Priya do?

Key points
  • SAP data falls into configuration, organizational, master and transactional data.
  • Configuration and organizational units reach QAS by transport; testers do not create them there.
  • Master data is created, migrated or copied, and gaps in it cause most data-related test failures.
  • Transactional data is consumed by tests and changes balances other tests depend on.

Check your understanding · not graded

Lesson check

  1. 1. Which of these normally reaches a QAS system by transport?

  2. 2. In SAP S/4HANA, customers and suppliers are maintained as what?

  3. 3. Why is transactional data harder to reuse across tests than master data?

  4. 4. Contoso's plant 1720 Chicago is which type of data?