Lesson 4 of 5
Decision criteria
9 min
Choosing a transition path is one of the largest decisions in a program, and it is hard to reverse. It should be made on explicit criteria, with evidence from analyses such as the SAP Readiness Check, rather than on habit or a partner's preference. Different systems in one landscape can even take different paths.
The main criteria
| Criterion | Points toward new implementation | Points toward conversion | Points toward selective data transition |
|---|---|---|---|
| Process fit | Processes need redesign | Processes still fit | Some areas redesigned, others kept |
| Custom code | Little worth keeping | Valuable and adaptable | Some worth keeping |
| History | Not needed in the system | Full history needed | History for some units or periods |
| Landscape | Single system or non-SAP source | Single SAP ERP system | Several systems to merge, or a carve-out |
| Target edition | Any, and the only path to Public Edition | Private Edition or on-premise | Private Edition or on-premise |
| Timeline pressure | Long enough for redesign | Tight; technical scope first | Depends on scope; often longer planning |
| Data quality | Poor data can be cleansed in migration | Data must be consistent enough to convert | Selection allows leaving poor data behind |
Decision criteria for the transition paths. Real decisions weigh several criteria together.
Effort and risk compared
| New implementation | System conversion | Selective data transition | |
|---|---|---|---|
| Business effort | Highest: every process designed, tested and trained | Lower: decisions on simplification items, validation and regression | Medium to high, depending on how much is redesigned |
| Technical effort | System build, extensions, interfaces, data migration | Technical conversion, custom code adaptation, rehearsals | Target build plus table-level transformation design and runs |
| Main risks | Scope creep, data migration, change fatigue | Hidden data and code issues, downtime, "IT only" mindset | Data consistency of selections, tool and provider dependency, complexity |
| Value | Maximum simplification and clean core | Fastest technical move; history and investments kept | Tailored balance; consolidation and phasing |
A worked decision: Contoso
Contoso's program weighed the criteria in its Discover and Prepare phases. The finance director needed ten years of history in the system for audit and warranty analysis. Operations were broadly happy with their processes. Custom code was mixed: some valuable fleet pricing logic, some reports that standard apps would replace. Contoso ran one SAP ERP system. The deadline was tight. And the target was Private Edition under RISE with SAP.
- Is the target Public Edition?No → all three paths remain open · Yes would have meant new implementation
- Is full history needed in the system?Yes, ten years → conversion or selective data transition
- Is there more than one source system, or a carve-out?No, one system → selective data transition loses its main advantage
- Do processes still fit, and is time short?Yes and yes → system conversion, with targeted improvements after go-live
The decision did not mean "no change". Contoso planned the mandatory simplification decisions as part of the conversion, retired custom reports in favor of standard apps, and scheduled process improvements in finance as follow-up projects. That combination, sometimes called a technical conversion with selective innovation, is common.
The decision is also made per system, not only per company. SAP's own examples show landscapes where one country's ERP is converted, another is reimplemented, and finance is consolidated separately. For each system, ask the same questions, and make sure the combined plan still fits the program's budget, its testing capacity and its cutover calendar, because two paths in one program mean two kinds of testing running side by side.
What would you do?
A retailer's SAP ERP system is twenty years old, heavily modified, and its processes no longer match a new business model. It needs only open items and two years of sales history for analytics, which can be kept in a data warehouse. Which path should the architects recommend evaluating first?
- Decide on explicit criteria: process fit, custom code, history, landscape, target edition, timeline, data quality.
- Compare effort and risk: new implementation is heaviest on business change; conversion on technical and data consistency; SDT on selection complexity.
- Contoso chose a system conversion: full history, fitting processes, one system, tight deadline, Private Edition.
- Record the decision with its evidence and trade-offs.
Check your understanding · not graded
Lesson check
1. A company needs full transaction history in the system and its processes still fit. Which path is a strong candidate?
2. Which criterion rules out conversion entirely?
3. Why did selective data transition offer little extra for Contoso?
4. What is a main risk profile of system conversion?