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

Lesson 3 of 5

Data and timing

12 min

Integration storylines fail more often from data and timing than from defects. A step cannot start because the product's costing view is missing; MRP run by one team overwrites another team's planned orders; the period closes before production settles. This lesson is about planning the data and the calendar so the story can actually be told.

Master data in dependency order

For the Contoso storyline, master data built by five teams must exist and agree before day one. Build and check it in dependency order: what a later object refers to must exist first.

  1. Organization and financeCompany code 1710, plants, G/L accounts, profit centers, cost centers
  2. Business partnersCustomer Great Lakes Fleet Services (sales, credit); supplier Lakeside Components (purchasing, bank)
  3. ProductsCMC-ALT-100 and CMC-RTR-200 with sales, purchasing, MRP, work scheduling, accounting and costing data
  4. Production master dataBill of material, work centers and routing for the alternator
  5. Purchasing and pricingInfo record or contract for rotors from Lakeside; fleet price for the customer
  6. Cost estimateStandard cost estimate released: $150 in plant 1710, $156 in plant 1720
Storyline master data in dependency order.
The Sign-Off Table · thesignofftable.com/academy/sap-xpr-302

Avoiding collisions

CollisionWhat happensPrevention
Two teams run MRP for the same plant and productsThe second run replaces planned orders and requisitions the first team was usingOne owner runs MRP, scoped to the storyline's products, at agreed times
Two storylines consume the same stockOne storyline's goods issue uses stock built for anotherSeparate products or storage locations per storyline; Detroit and Chicago storylines use different plants
Batch jobs run at random timesBilling due list or payment run picks up documents from other storylinesRestrict job selections to storyline data; run jobs at scheduled checkpoints
Someone changes shared master data mid-cycleResults change partway through the storylineFreeze storyline master data; route changes through the test lead
Common collisions in integration testing and how to prevent them.
The Sign-Off Table · thesignofftable.com/academy/sap-xpr-302

Dates and periods

A storyline that ends in a period close needs all its postings in the period being closed, and the close itself after the last posting. Plan the calendar backwards from the close. Agree with finance which period the cycle uses, when it opens and when the period-end steps run. Remember that materials management and finance have separate period controls, and that validity dates on prices and cost estimates must cover the posting dates.

  1. Week 0Data readyMaster data checked; cost estimates released; period open; test users assigned
  2. Week 1, day 1Demand and MRPFleet order entered; MRP run by the planning owner at 14:00
  3. Week 1, days 2 to 3Buy and receivePO to Lakeside; goods receipt; supplier invoice
  4. Week 1, days 3 to 5MakeProduction orders released, confirmed and received
  5. Week 2, days 1 to 2Ship, bill, collectDelivery, goods issue, billing, incoming payment; payment run
  6. Week 2, days 4 to 5CloseSettlement, variance analysis, period-end checks by finance
A two-week calendar for Contoso's main storyline, planned backwards from the close.
The Sign-Off Table · thesignofftable.com/academy/sap-xpr-302

Entry criteria

Before day one, Priya runs a short readiness check with each team lead. Is every master data object for the storyline present and reviewed? Are the test users created with their target roles? Are the periods open, the cost estimates released and the interfaces pointing at test endpoints? Are the jobs scheduled or ready to run on demand? A cycle that starts with a red item usually loses more time than the day it takes to fix it first. Record the check, because when a storyline fails later, the first question will be whether its data was ready.

Simulating time

Real processes take weeks; test cycles take days. You compress time by agreeing that "delivery date = today + 2" and by running jobs on demand at checkpoints rather than overnight. Be careful with anything that depends on real time passing: payment terms and due dates, dunning, aging. Either post with dates that make items due, agreed with finance, or schedule those branches for a later cycle.

What would you do?

On day 3, the production team reports that their planned orders for alternators have disappeared. The procurement team reran MRP for plant 1710 that morning to test a different requisition scenario. What should Priya do?

Key points
  • Build and check master data in dependency order; the standard cost estimate is a dependency.
  • Prevent collisions: one MRP owner, separate stock per storyline, scoped batch jobs, frozen master data.
  • Plan the calendar backwards from the period close; check both financial and materials management periods.
  • Compress time deliberately and handle due-date-driven steps with agreed dates or later cycles.

Check your understanding · not graded

Lesson check

  1. 1. Why build storyline master data in dependency order?

  2. 2. Which entry criterion protects the finance checks in Contoso's storyline?

  3. 3. Two teams run MRP for the same plant. What is the risk?

  4. 4. How should the storyline calendar be planned?