Case 01
SAP Retail Transformation
Large African retail group
2,400+ stores. 11 go-lives. 9 months. 50bn+ rows.
- Stores migrated to SAP Retail
- 2,400+
- Stores live in a single cutover
- 450
- Go-lives in nine months
- 11
- Rows handled across the programme
- 50bn+
A major retail transformation moved more than 2,400 stores onto SAP Retail in aggressive deployment waves, with business-critical cutovers and stores trading throughout. Ixia delivered the migration workstream: transformation, validation, reconciliation and cutover support.
Challenge
- Move the store, article, pricing and merchandising data of a 2,400+ store estate onto SAP Retail without interrupting trade.
- Deployment ran in waves, so every cutover had to land while earlier waves were still stabilising and later waves were still being prepared.
Scale
- More than 2,400 stores across the group's brands and formats.
- 50bn+ rows of master and transactional data processed across the programme.
- 11 go-lives inside nine months, the largest taking 450 stores live in a single weekend.
Ixia's role
- Migration delivery: extraction, transformation and load design for each deployment wave.
- Business rule design and transformation logic, agreed with the business, not assumed.
- Validation and reconciliation reporting used as the go / no-go evidence at each cutover.
- Cutover planning and weekend execution support alongside the programme team.
Engineering approach
- Repeatable ETL built once and hardened wave after wave, rather than rebuilt per deployment.
- Iterative mock loads per wave, each producing a reconciliation pack the business could read and sign.
- Automated validation on the objects that break retail trading first: articles, pricing, site structures.
- Cutover runbooks rehearsed against real volumes before any store was committed.
What made it difficult
- Stores traded on Friday and had to trade again on Monday. The cutover window was a weekend, and rollback was not a plan anyone wanted to use.
- The 450-store wave compressed the full extract, transform, load and reconcile cycle into one weekend at maximum volume.
- Waves overlapped: stabilising the last go-live while rehearsing the next one, with the same data team across both.
Outcome
- 11 go-lives delivered in nine months.
- 450 stores taken live in a single weekend with full reconciliation and no critical issues.
- Every wave delivered with clean audit logs and business sign-off.
Delivery lessons
- Cutovers are won in rehearsal. The weekend is just the performance.
- Reconciliation only works as evidence if the business can read it without a data engineer in the room.
- Wave pace is set by data readiness, not by system readiness.
Technologies
- SAP Retail
- SAP Data Services
- SQL-based ETL
- Reconciliation dashboards in Power BI
“Ixia was the calm in the chaos. They knew what to do before we even had to ask.”
Related capabilities and insights
Facing something similar?
Bring us the programme. We will tell you honestly what it takes.