Skip to main content

15.4 Lineage, Impact & Trust

1. Lineage answers "where did it come from?"​

Lineage describes the path from source to consumer.

Example:

S/4HANA
  |
  v
CDS / extraction
  |
  v
Remote / local acquisition
  |
  v
Transformation Flow
  |
  v
Curated Table
  |
  v
Semantic Model
  |
  v
Analytic Model
  |
  v
SAC Story

A consumer can use lineage to evaluate provenance.

2. Impact answers "what breaks if I change it?"​

Impact analysis reverses the perspective.

                         Customer Master
                               |
                         +-----+-----+
                         |           |
                     Sales Model   Margin Model
                         |           |
                       SAC A       SAC B
                         |
                      Planning

If Customer Master changes, impact analysis identifies downstream dependencies.

3. Why lineage matters​

Lineage supports:

  • audit,
  • incident response,
  • root-cause analysis,
  • migration,
  • change management,
  • compliance,
  • consumer trust.

4. Technical lineage versus business lineage​

Technical lineage​

Shows object-to-object movement and transformation.

Business lineage​

Explains the business journey:

Source revenue transaction -> Finance curated revenue -> Revenue KPI -> Executive dashboard.

Architects should make both available where possible.

5. Lineage is not proof of quality​

A perfectly documented pipeline can still contain bad data.

Therefore:

Lineage
+
Ownership
+
Quality
+
Freshness
+
Definition
=
Trust context

Do not label an asset "trusted" merely because lineage exists.

6. Change impact workflow​

Before modifying a Tier 1 asset:

  1. Identify object.
  2. Inspect lineage.
  3. Inspect downstream impact.
  4. classify consumers by criticality.
  5. identify semantic dependencies.
  6. identify security implications.
  7. assess data contract changes.
  8. notify owners.
  9. test.
  10. deploy.
  11. verify consumers.

7. Breaking versus non-breaking changes​

Usually breaking​

  • removing a column,
  • changing meaning,
  • changing grain,
  • changing key semantics,
  • changing unit,
  • changing time zone semantics.

Potentially non-breaking​

  • adding a nullable attribute,
  • improving metadata,
  • optimizing implementation without semantic change.

Do not assume a schema-compatible change is business-compatible.

8. Trust score​

An enterprise can create a composite trust score:

Trust =
  ownership score
+ definition score
+ lineage score
+ quality score
+ freshness score
+ certification score

Do not treat this as an SAP product metric unless explicitly supported. It is an enterprise governance pattern.

9. Architect exercise​

A Finance table changes its fiscal-period logic.

Design the impact process.

Your answer must identify:

  • technical lineage,
  • semantic consumers,
  • KPI consumers,
  • security effects,
  • notification path,
  • rollback strategy,
  • post-change validation.