Solution Atlas
EverydayUser storyConsultative playbook

Board pack production takes a week and we want a day

The CFO's team produces the monthly board pack by hand from a tangle of Excel, Power BI, and ad-hoc data pulls. The board has asked for more frequent updates but the production cycle cannot compress further without breaking quality.

Trigger
Board pack pressure; finance team at manual-production limit.
Good outcome
Certified semantic model, automated board pack pipeline, Excel + Copilot in the finance workflow.
Diagnostic discovery

Signals this story fits

Observable cues that confirm the conversation belongs here.

  • ·Monthly board pack production takes 5+ business days
  • ·Finance team copy-pasting figures between Excel and PowerPoint
  • ·Numbers reconciled by hand against the GL each month
  • ·Board has asked for more frequent updates (mid-month, weekly)
  • ·No certified semantic model — every Power BI report queries differently

Questions to ask

Open-ended, SPIN-style — each one has a reason it matters.

  1. 1.Walk me through the last board pack — who produces what, and how long does each step take?

    WhySurfaces the manual choke points.

    Listen for: “three analysts for a week” · “we miss the deadline regularly” · “a different person owns each section”

  2. 2.When the board asks "why is this number different from last month's pack?" — who can answer that, and how long does it take?

    WhyTests semantic model maturity. Long answer = no certified definitions.

    Listen for: “nobody can answer that quickly” · “we have to go back to the GL” · “different reports give different numbers”

  3. 3.Where do the numbers in the pack originate — GL, CRM, operational systems, spreadsheets?

    WhySizes the data substrate scope. Pure-GL is simpler than multi-source.

  4. 4.Who is the audience — full board, audit committee, exec only? And what is the format expectation — PDF, deck, live dashboard?

    WhyShapes the delivery channel. Some boards still require PDF; others accept live links.

  5. 5.How sensitive is the data — must the pack stay inside finance until release?

    WhyDrives the row-level security and labelling posture.

  6. 6.Have you tried Copilot in Excel or Power BI for finance workflows already?

    WhySurfaces existing M365 Copilot exposure and the Excel-as-front-end opportunity.

Baseline → target architecture

TOGAF-style gap framing — what we typically see today, and what the proposed end state looks like. The gap between them is the engagement.

Baseline architecture

Board pack produced manually from a tangle of Excel workbooks, Power BI reports, and ad-hoc data pulls. Finance analysts spend the first week of each month reconciling numbers and assembling the deck. No certified semantic model — different reports give different answers to the same question. Board asks for higher frequency but the production cycle cannot compress without compromising quality.

Typical concerns

  • ·Production cycle takes 5+ business days
  • ·Numbers vary across reports for the same metric
  • ·Reconciliation against the GL is manual
  • ·Finance analysts are the bottleneck — no scale
  • ·Board is asking for frequency the cycle cannot deliver

Capability gaps

  • ·Certified semantic model with single source of truth
  • ·Automated pipeline from operational systems to model
  • ·Excel and PowerPoint integration with live data
  • ·Copilot for finance narrative generation
  • ·Row-level security for pre-release sensitivity
Target architecture

Microsoft Fabric ingests data from the GL and operational systems on a scheduled cadence. A certified Power BI semantic model defines every metric the board pack uses — one definition, used everywhere. Excel connects live to the model via the Analyze in Excel pattern. PowerPoint slides bind to Power BI visuals that refresh automatically. Copilot in Power BI generates the variance commentary draft. Finance shifts from production to review. Cycle time under one day.

Key capabilities

  • Certified semantic model
  • Automated Fabric pipeline from GL + operational systems
  • Excel live connection to the model
  • Live-bound PowerPoint visuals
  • Copilot variance commentary

Enabling SKUs

Resolved in the ‘Recommended cards’ section below.

Architecture decisions

Each decision is offered as explicit options with trade-offs — Hohpe's “selling options” principle. A safe default is noted where one exists.

  1. Decision 1.Semantic model location — Fabric vs Power BI Pro workspace vs Azure Analysis Services

    Fabric (Direct Lake)

    When it fitsStrategic data substrate on Fabric; Direct Lake performance for finance volumes.

    Trade-offsFabric capacity cost; newer pattern.

    Power BI Pro workspace

    When it fitsFinance volume modest; Pro workspace sufficient.

    Trade-offsCapacity limits as model grows; less integration with broader data substrate.

    Azure Analysis Services

    When it fitsLegacy AAS investment; large tabular models with mature governance.

    Trade-offsAAS roadmap deprioritised vs Fabric.

    Default recommendationFabric Direct Lake for new builds. Power BI Pro workspace only if Fabric is genuinely out of scope.

  2. Decision 2.Board pack format — live links vs PDF export vs PowerPoint with live binding

    Live links (Power BI app)

    When it fitsBoard comfortable with web access; iPad-based reading.

    Trade-offsBoard members offline or screen-sharing struggle with live links.

    PDF export

    When it fitsTraditional board; offline reading; mandatory for audit committee.

    Trade-offsLoses interactivity; point-in-time snapshot.

    PowerPoint with live binding

    When it fitsHybrid — narrative in PowerPoint, visuals refreshed automatically.

    Trade-offsBound visuals require Power BI access at open-time.

    Default recommendationPowerPoint with live binding as the primary; PDF export as the archive artefact for audit.

  3. Decision 3.Copilot scope — Excel only vs Power BI only vs full M365 Copilot

    Copilot for Excel only

    When it fitsFinance team works primarily in Excel; lower licensing footprint.

    Trade-offsNo narrative generation in Power BI or PowerPoint.

    Copilot in Power BI

    When it fitsVariance narrative is the value add; finance reviews in Power BI.

    Trade-offsExcel-driven workflows unchanged.

    Full M365 Copilot for finance team

    When it fitsFinance team broadly using M365; Copilot value across Excel + PowerPoint + Word + Teams.

    Trade-offs€28/user/month; cost vs benefit needs the trial to validate.

    Default recommendationFull M365 Copilot for the core finance team (5–10 users) during the trial; expand based on measured productivity gain.

Low-risk trial — proof of value

60-day certified board pack on Fabric + Copilot

8 weeks

Top 12 board pack metrics defined in a certified Power BI semantic model on Fabric. Fabric pipeline live for the GL and one operational system. PowerPoint board pack template with live-bound visuals for the certified metrics. Copilot in Excel + Power BI enabled for the core finance team (5–8 users). Variance commentary draft produced by Copilot for the trial month. Cycle time measured against the previous month baseline.

Success criteria

  • Top 12 metrics certified in semantic model
  • Board pack cycle time under 2 business days for the trial month
  • Variance commentary drafted by Copilot, reviewed by finance
  • CFO accepts the trial pack format for the next quarter

InvestmentFabric capacity F2/F4 for the trial scope + Power BI Pro for finance team + M365 Copilot per user. Estimated ~€4–8k/month for the trial scope. Existing board pack production continues in parallel as the control.

Proof metrics

  • ·Board pack cycle time reduced by 60%+ vs baseline
  • ·Single certified definition for top 12 metrics — no reconciliation needed
  • ·Copilot variance commentary accepted by CFO with light edit
  • ·Trial pack adopted as the standing format for the next quarter

Recommended cards

The SKUs and capabilities most likely to be part of the solution, with the editorial rationale for each in the context of this story. Add the ones that fit your situation.

Why for this story

Power BI is the presentation surface for the board pack — both the live visuals and the Excel-bound model. The certified semantic model is consumed through Pro licences across finance.

Why for this story

The certified semantic model lives on Fabric. Direct Lake gives the performance finance needs at month-end without copy-and-load. The substrate the whole solution rests on.

Why for this story

Copilot drafts the variance commentary in PowerPoint and answers finance questions in Excel against the certified model. Shifts finance from production to review — the heart of the cycle compression.

Back to Executive reporting & board packs