An add-on is worth testing when the core estimate is already dependable and the missing task has a clean input/output boundary. Replacement is worth evaluating when fragmented data structures or unsupported software make the core workflow itself the problem.
| Decision question | Favours a bounded addition | Favours evaluating replacement |
|---|---|---|
| Is the approved estimate reliable? | Yes; preparation is the main burden | No; core assumptions and controls are fragmented |
| Can data leave and return with identity preserved? | Supported exports or verified API | No usable integration route |
| Is the gap narrow? | One schedule, revision or approval task | Multiple coupled failures |
| Can the team support another component? | Named owner and support | Unmanageable duplication |
| Is migration feasible? | Better to avoid unnecessary migration | Migration has an accountable plan and tested mapping |
These are decision prompts, not rules that force a purchase. A failing base process may first need standardisation, with no new system.
Sources cited on this page
No external sources are cited on this page. Every source for this section is listed on its Sources page.