Which criteria speak for and against taking a new kind of information into the IAM scope?
For: the information is cross-cutting (several applications use it), long-lived, general, and its benefit justifies the cost of providing it centrally. Against: it serves one application only, is short-lived, special, or the effort is disproportionate.
| Criterion | Speaks for the extension | Speaks against it |
|---|---|---|
| Reach | Cross-cutting: several applications can use it sensibly | Useful for one application only (local benefit) |
| Lifetime | Permanent, needed for years | Short-term, a project or a campaign |
| Generality | General (an organisational unit, a cost centre) | Special (a niche qualification only one team cares about) |
| Cost-benefit | The benefit justifies central provision | The effort is disproportionately high |
The reasoning behind the table: the IAM is a shared, central service. Everything inside it is maintained once and reused many times, which only pays off if the information is reused and stays relevant long enough to amortise the set-up (data model change, data-protection basis, source, delivery agreement, connector). Information that only one application needs is cheaper and cleaner to keep in that application.
Tip: Think "shared, stable, general, worth it". The more of these four hold, the stronger the case for the IAM.
Go deeper:
Single source of truth (Wikipedia) — the benefit of central data, and why it only pays off for data many consumers need.