Lesson 3 of 6
Process redesign opportunities
6 min
A conversion keeps your processes by default. That is its strength: lower risk and less change for users. But every conversion also creates moments when changing a process is cheaper than it will ever be again. Custom code has to be checked anyway. Users will be retrained on changed apps anyway. Data is being cleaned anyway. The process owner's task is to spot the redesigns worth doing now, and to say no to the rest.
Where conversion invites simplification
- Custom code with a standard successor. If a custom report rebuilds what a standard Fiori app now does, retiring it saves remediation, testing and future upgrades.
- Workarounds for ECC limits. Spreadsheets and manual reconciliations that existed because FI and CO were separate can often go, because the Universal Journal puts them in one place.
- Simplification items that force change anyway. Credit management is changing regardless; that is a natural moment to fix credit rules that the business disliked.
- Dead process variants. Order types, document types and approval paths nobody has used in years can be retired instead of converted and tested.
- Manual controls that can become system controls. A check done by email may become a workflow or a configured block.
Scope against risk
The bottom-right box matters most. Conversions fail on their timeline when "while we are at it" requests pile up. Each redesign adds design, configuration, testing, training and data work, and it also makes comparison testing harder, because the converted process no longer should give the same result as ECC. A common, sensible rule is convert first, innovate after: keep the conversion scope tight and put larger improvements on the post-go-live roadmap, where the continuous change process from the previous course can handle them.
A light business case
Every redesign request should answer five questions on one page: what pain it removes, which process steps and users change, what custom code or data it touches, how it will be tested, and what happens if it slips. If nobody can answer them, the request goes to the post-go-live list.
- RequestBusiness raises a redesign idea with a one-page case
- SortProcess owner places it on the benefit and disruption grid
- Design authorityProgram reviews impact on scope, timeline, testing and comparison
- DecideIn conversion scope, or on the post-go-live roadmap with a target release
- PlanIf in scope: design, configuration, test scenarios and training added to the plan
| Candidate at Contoso | Benefit | Disruption | Decision |
|---|---|---|---|
| Retire 14 custom finance reports replaced by standard line item and balance apps | Less remediation and testing | Low: users retrained on apps already planned | Do it in the conversion |
| Redesign credit risk classes with the move to SAP Credit Management | Fewer false credit blocks | Medium: rules and training | Decide deliberately: approved, with tests |
| Introduce a new intercompany process for Chicago | Faster month end | High: new flows and controls | Defer to after go-live |
| Rename document types for consistency | Cosmetic | Low but touches every test | Skip |
Contoso's redesign decisions, one line each.
What would you do?
Midway through the conversion, the sales director asks Maria to add a new rebate process for fleet customers "since we're changing everything anyway." There are eight weeks to the first user acceptance test. What should the order-to-cash owner do?
- Conversion creates cheap moments to simplify: custom code, workarounds, forced changes and dead variants.
- Sort candidates by benefit and disruption; defer high-disruption, low-benefit changes.
- Convert first, innovate after keeps the timeline and comparison testing intact.
- Every redesign needs a one-page case answering pain, change, code and data, test, and slip impact.
Check your understanding · not graded
Lesson check
1. Which redesign is usually worth doing during a conversion?
2. What does "convert first, innovate after" mean?
3. How does redesigning a process during conversion affect comparison testing?
4. Where does a high-disruption, low-benefit idea belong?