EPM 101: Reporting & Narrative Intelligence

1. Structure — the backbone

• Hierarchies (legal, managerial, segmental)
• Dimensionality (entity, product, cost center, channel)
• Standard chart of accounts
• Alternate rollups (geo, segment, function, P&L structure)

2. Logic — the brain

• Consolidation rules
• FX translation
• Variance formulas
• Scenario relationships
• Allocations
• Eliminations
• Bridges (price, volume, mix, FX, acquisition)

3. Workflow — the engine

• Submission deadlines
• Review & approvals
• Commentary inputs
• Task management
• Data refresh sequencing
• Audit trails

4. Presentation — the interface

• P&L formats
• Dashboards
• Executive scorecards
• Cashflow statements
• Operator-level reports
• Board decks

5. Narrative — the story

• What happened
• Why it happened
• Which drivers matter
• Forward guidance
• Risks & opportunities
• Strategic implications

BI = surface-level visibility

BI excels at:

• Visualizing data
• Slicing & dicing
• Exploring operational metrics
• Real-time dashboards
• Large datasets
• Interactive intelligence

BI is phenomenal for: ARR dashboards, CAC cohorts, sales pipelines, marketing attribution, product usage, operational KPIs, real-time scorecards.

What BI CANNOT do: Run consolidation, manage close workflows, enforce financial hierarchies, maintain dimensional consistency, provide scenario governance, maintain version control, apply FX rules, provide audit trails, generate financial packets, coordinate commentary inputs.

EPM = structured financial truth

EPM tools support:

• Accounting logic
• Consolidation
• Currency translation
• Intercompany elimination
• Allocations
• Financial statements
• Budget/forecast versions
• Scenario control
• Commentary workflows
• Data certification

EPM answers questions BI cannot: "Does this variance include FX?" "Did we eliminate intercompany?" "Are these numbers pre- or post-allocation?" "Was this forecast version the pre-board or post-board one?" "Has this cost center submitted commentary yet?"

3.1 Reporting is built on disconnected systems

The typical mid-market reporting stack is: ERP actuals, Excel budget, BI dashboards, PowerPoint management packs, Google Sheets for commentary, Email for approvals, Slack/Teams for 'does this number look right?'. Every component works in isolation. Result: No shared model, No shared truth, No structural consistency, Endless late rework. This is why reporting days are a disaster.

3.2 Excel becomes the accidental reporting system

Behind the scenes, analysts maintain: custom logic, pseudo-databases, month-over-month roll-forwards, hidden sheets, fragile macros, ungoverned assumptions. The entire company's reporting intelligence lives in a spreadsheet owned by one analyst. This is organizational risk.

3.3 Reporting cycles depend on heroic individuals

96% of companies have reporting dependent on 1–3 people who: know the files, know the logic, know the hierarchies, know the workarounds, understand the nuances. If those people quit, the reporting engine collapses. This is key-person fragility — a huge audit and continuity risk.

3.4 Commentary is unstructured (and often an afterthought)

Typical commentary workflows: Email your commentary, Add comments in Excel cells, Insert notes in PowerPoint, Slack message your narrative. This means: No coherence, No consistent driver analysis, No visibility into BU-level explanations, No version history, No audit trail, No roll-forward of last period's commentary. Narrative becomes chaotic — and executives notice.

3.5 No one has time to analyze, because they're too busy assembling

FP&A spends: 60–70% of reporting time collecting, 20–30% cleaning, 5–15% assembling, 0–10% analyzing. The value of reporting is in the 10%, but the effort goes into the 90%. Your entire finance team's ROI is underwater.

4.1 Reporting becomes governed and repeatable

With EPM, reporting is: controlled, structured, automated, versioned, audited, orchestrated. Not: improvised, hero-dependent, fragile, chaotic.

4.2 Variance analysis becomes standardized and explainable

EPM generates variances: automatically, consistently, with FX separated, with allocations clearly identified, with prior period adjustments explained, with drivers calculated (volume/price/mix). And with drill-down from P&L → entity → cost center → account → journal → transaction. Excel cannot do that.

4.3 Narrative becomes a FIRST-CLASS CITIZEN

Finance teams can capture: commentary by BU, commentary by entity, commentary by account, commentary by variance driver, commentary by period. With: approval workflows, visibility dashboards, late commentary tracking, narrative roll-forward, tagged commentary (risk, opportunity, root cause, trend). This is world-class narrative governance.

4.4 Reporting packets generate themselves

Imagine: Monthly packets, Quarterly packets, Board books, Regional decks, CFO summary packs...building themselves automatically when the close calendar completes. With: standardized formatting, live variance logic, up-to-date commentary, consistent structure. This is reporting at enterprise-grade maturity.

Gen 1 — Static Templates

Pre-written statements like: 'Revenue increased by X due to Y.' 'Expenses were unfavorable due to higher headcount.' Useful, but shallow.

Gen 2 — Rules-Based Narrative Engines

Driver-aware commentary: FX, Volume, Price, Mix, New business vs churn, One-time events, Restructuring, Seasonality. These are structured and explainable.

Gen 3 — AI-Assisted Narrative Intelligence (2024–2026 onward)

The new frontier. AI can: analyze the entire financial model, identify unexpected shifts, quantify drivers, compare against historical patterns, assess quality of earnings, generate 'executive-ready' narratives, produce summaries tailored to CFO/CEO/BU leader, highlight discrepancies between narrative and numbers, identify forecast risks automatically. This is transformational. It does not replace FP&A. It supercharges FP&A.

Level 1 — Reactive Reporting (Excel chaos)

• Email-driven workflows
• Manual consolidation
• No variance standards
• Commentary optional
• Data inconsistencies
• Analysts cleaning instead of analyzing

Level 2 — Visual Reporting (BI-driven)

• Dashboards
• Better visuals
• Faster access
• Still no structured logic
• Still no narrative system
• Still no workflow
• Finance numbers don't tie to BI exactly

Level 3 — Managed Reporting (EPM-driven)

• Unified dimensions
• Version-controlled reports
• Standard variance analysis
• Automated packets
• Commentary workflows
• Drill-down everywhere

Level 4 — Intelligent Reporting (NI-enabled)

• Automated narrative suggestions
• Anomaly detection
• AI-driven insights
• Proactive variance identification
• Real-time CFO summaries
• Publishing-grade packets

Step 1 — Create the reporting blueprint

• Define: reporting calendar, roles, ownership, data sources, hierarchies, CoA structure, mapping logic, reconciliation targets, narrative expectations

Step 2 — Design your reporting library

• You need: P&L (Mgmt + Legal versions), BS, Cash Flow, Regional P&L, Product/BU reporting, KPI dashboards, Forecast versions, Variance decks (M/M, Q/Q, Y/Y), Board packet template

Step 3 — Define narrative expectations

• For each report, define: who writes commentary, where, when, how much, on which drivers, with what approval. This is massively underrated.

Step 4 — Build financial logic inside EPM

• scenario definitions, FX/translation logic, consolidation rules, IC rules, allocation logic, mapping logic, variance formulas, driver logic, KPI calculations. This separates the professionals from the rest.

Step 5 — Automate report generation

• packets, decks, dashboards, PDF exports, scheduled distributions. All built on the same model, not copied spreadsheets.

Step 6 — Layer in Narrative Intelligence

• Start with: anomaly detection, variance highlights, driver breakdown, commentary prompts. Then scale to: AI-generated draft commentary, automated executive summaries, risk & opportunity analysis, automated insights. This is the future.