Scenario Lab 19–24

19 — Popular report is slow

Stem: report runs thousands of times daily and performs poorly.

Primary: high-impact performance problem.
Supporting: Storage/Operations, physical design, usage metrics, BI product management.
Roles: DBA/DW operations, BI developer, product owner.
Best: use usage/performance evidence to prioritize tuning; consider indexing, aggregates/precomputation, query redesign or other justified optimization while preserving needed detail.
Weaker: tune all reports equally or aggregate away required atomic data.
Changed: nobody uses it → retirement/rationalization may beat optimization.
Source: pp. 364, 391–393.

Stem: no planned increments; enhancements are reactive.

Primary: missing product/release lifecycle.
Supporting: Program/Product Management, Operations, Governance, Architecture.
Roles: product owner, business partners, operations, developers.
Best: backlog prioritization, roadmap, planned releases, monitoring and proactive engagement.
Weaker: incident response as complete maintenance model.
Changed: production instability still needs incidents, but alongside—not instead of—lifecycle management.
Source: pp. 380–392.

Stem: sandbox model is useful and team wants immediate publication.

Primary: production-readiness gate.
Supporting: Security, DQ, Metadata, Operations, Governance, UAT.
Roles: business owner, IT ops, security/governance, DW/BI team, Stewards.
Best: require business+IT readiness, DQ/lineage/security/governance, support/monitoring, configuration/release control, acceptance.
Weaker: promote solely because pilot looks useful.
Changed: pilot fails objectives/controls → return to development/reject.
Source: pp. 386–392.

Stem: 10,000 licenses exist, only 400 people connect.

Primary: misleading adoption evidence.
Supporting: product metrics, satisfaction, support.
Roles: product owner, sponsor, CoE/support.
Best: actual connected/concurrent users, query/activity frequency and trends.
Weaker: licenses as proof of value; entitlement is not behavior.
Changed: entitlement/audit question → registered-user count remains useful for that purpose.
Source: pp. 391–393.

Stem: Marketing never uses Product subject data while Sales and Finance do.

Primary: subject-area/business coverage.
Supporting: lineage, mappings, adoption, roadmap.
Roles: product owner, DW architect, business sponsors.
Best: coverage metrics to expose domain/department gaps and prioritize roadmap.
Weaker: query-response metric; speed does not measure breadth.
Changed: issue becomes slow response → performance metrics become primary.
Source: pp. 392–393.

Stem: expected four-hour load now takes six.

Primary: load/service performance against expected window.
Supporting: Operations, SLA, capacity/performance, Integration.
Roles: DW operations, integration engineer, product owner, consumers.
Best: measure load duration/success against agreed window, diagnose bottleneck, tune or renegotiate justified expectation.
Weaker: focus on license/adoption metrics; irrelevant to load window.
Changed: data arrives on time but values wrong → DQ metrics/remediation dominate.
Source: pp. 391–393.