Generating Board Packs with AI Assistance
Deliberate Academy Editorial Team
Reviewed for accuracy and professional relevance
You're 6 lessons in — don't lose your progress.
Sign up free to save where you are and earn a verified certificate when you pass.
- Apply AI to accelerate the production layer of board pack preparation — chart narration, executive summary drafting, and appendix formatting — while keeping message selection and framing with the FP&A team
- Structure an effective AI brief for board-level narrative that produces commentary a CFO can stand behind rather than generic restatement
- Explain the certainty-overstatement failure mode specific to board pack narrative, and the review discipline that catches it before distribution
- Assess which sections of a board pack are appropriate for AI-assisted drafting and which require the FP&A leader's direct authorship
The board pack is the highest-stakes recurring deliverable most FP&A teams produce: it is read by people with limited time and significant decision authority, and every number and claim in it is implicitly attributed to the CFO who presents it. AI can meaningfully compress the production time for a board pack — chart narration, first-draft executive summaries, appendix formatting — but the stakes of this specific deliverable mean the review discipline matters more here than almost anywhere else in the FP&A cycle.
For public companies, that attribution is not only reputational. The CFO's periodic certification under Section 302 of the Sarbanes-Oxley Act (SOX) that the company's financial disclosures and controls are effective sits on top of a stack of monthly and quarterly board reporting that increasingly starts life as an AI-assisted first draft, and the board pack's GAAP-basis actuals — or IFRS-basis actuals, for organizations reporting under IFRS — are the same figures that internal controls over financial reporting (ICFR) under SOX Section 404 exist to protect. Organizations without a codified SOX-equivalent certification regime still rely on board and audit-committee oversight of internal controls to play an analogous role, which is exactly why the review discipline in this lesson applies regardless of jurisdiction.
The Board Pack Production Pipeline
A typical board pack moves through several stages: consolidating the underlying financial and operational data, building the charts and tables that will carry the numbers, drafting narrative commentary that explains what the data shows, and assembling everything into a final reviewed document. AI's contribution is concentrated in the narrative and formatting stages — turning structured data into clear prose and consistent chart labeling — rather than in the underlying data consolidation, which still needs to be accurate before any narrative can be trusted.
Vena, built around an Excel-native planning and reporting workflow, and Workday Adaptive Planning's Office Connect feature, which links planning data directly into board-ready PowerPoint and Word output, are both used by FP&A teams specifically to compress this production pipeline — reducing the manual copy-and-format work that historically consumed a disproportionate share of board pack preparation time.
AI-Assisted Narrative and Chart Commentary
Given a chart's underlying data and a short note on what matters about it, AI can draft the one or two sentences of narration that typically accompany a board pack chart — explaining the trend, calling out the notable movement, and setting it in context. This is a strong AI use case at the sentence level: the narration task is genuinely mechanical once the analyst has identified what the chart needs to communicate.
Brief AI on board pack narration the way you would brief a new analyst writing for the board for the first time: tell it who the reader is (a non-executive director, typically not a finance specialist), what tone is expected (formal, direct, no unnecessary hedging but no false certainty either), and what the chart is meant to prove. "Revenue grew 12% this quarter" is a caption. "Revenue growth accelerated to 12% this quarter, driven primarily by the enterprise segment following the Q1 pricing change" is board-ready narration — the difference is entirely in what you told the model mattered.
Board pack chart narration
Before
Operating expenses were $2.1M this quarter compared to $1.9M budget.
Restates the chart without adding any interpretation — a board member can read the chart itself and gets nothing additional from this sentence.
After
Operating expenses came in $200K above budget this quarter, driven primarily by the accelerated hiring plan approved in February; excluding that planned investment, expenses were in line with budget.
Explains the variance, attributes it to a known and approved driver, and gives the board the context needed to interpret the number correctly rather than assume an unplanned overspend.
The Certainty-Overstatement Failure Mode
Board pack narrative carries a specific risk that is different from ordinary management reporting: because the audience has limited time and typically will not interrogate the underlying model, AI-drafted commentary that reads with more confidence than the data supports is far less likely to be caught before it shapes a board-level decision. A forecast described as "on track to reach $12M" when the underlying model shows a wide plausible range, or a risk described as "contained" when the mitigation plan is still unproven, are both examples of language that quietly upgrades uncertainty into certainty — exactly the kind of statement a board relies on and rarely questions in the room.
Before a board pack goes to print, review every AI-drafted narrative statement specifically for language that implies more certainty than the underlying model supports — words like "on track," "confirmed," "secured," or "contained" should only appear if the underlying data genuinely supports that level of confidence. This is not a proofreading pass; it is a distinct verification step that checks meaning against data, not grammar against style. Financial communications to the board carry fiduciary weight, and sign-off from the CFO or FP&A leader is required regardless of how the first draft was produced. Where a board pack presents an adjusted or non-GAAP figure — adjusted EBITDA, normalized margin — alongside the GAAP or IFRS number, the same rigor extends to labeling: SEC Regulation G requires any non-GAAP measure a public company discloses to be reconciled to the nearest GAAP measure, and an AI-drafted narrative that quotes an adjusted figure without carrying that reconciliation forward creates exactly the kind of gap this review step exists to catch.
A board pack draft, produced with AI assistance, includes the sentence: 'The cost reduction program is on track to deliver the full $2M in annual savings.' The underlying model shows the program has delivered $600K in savings to date, with the remaining $1.4M dependent on a vendor renegotiation that has not yet been finalized. What is the problem with the sentence as drafted?
Select one answer.
Rewriting an overconfident board narrative before a fundraising-adjacent board meeting
Context
An FP&A director at a growth-stage technology company used Vena's reporting workflow to consolidate quarterly board pack materials, including AI-assisted narration for the pack's fifteen core charts, ahead of a board meeting that would also touch on a planned Series C raise. The AI-assisted narration reduced chart commentary drafting time from roughly a full day to under two hours.
Action
Reviewing the draft before distribution, the director specifically checked every chart narration for certainty-overstating language, per her team's standing pre-distribution checklist. She found that the customer retention chart's narration described churn as 'stabilized' based on a single strong quarter, when the underlying twelve-month trend still showed meaningful volatility and the most recent quarter's improvement had a plausible one-time explanation — a large customer's renewal that happened to land favorably.
Outcome
She revised the narration to describe the recent quarter's improvement accurately while noting the twelve-month volatility and the specific factor behind the recent uptick, rather than declaring the metric stabilized. The board's discussion of retention during the meeting focused on the actual underlying drivers rather than treating churn as a solved problem, which the director considered directly relevant given that investor due diligence in the upcoming raise would scrutinize the retention trend closely regardless of how the board pack had characterized it.
Why does the lesson describe the review discipline for board pack narrative as more important than for ordinary management reporting?
Select one answer.
Exercise
Your Task
Take three charts or figures from your most recent board pack, or a representative set if you do not produce board packs directly. Draft AI-assisted narration for each, briefing the tool on audience, tone, and what the chart needs to communicate. Then review each narration specifically for certainty-overstating language — words like 'on track,' 'secured,' 'stabilized,' or 'contained' — and check whether the underlying data genuinely supports that level of confidence. Revise any sentence that does not pass this check.
Success looks like
- Each narration is grounded in a specific brief about audience and what the chart needs to prove, not a generic restatement request
- You have specifically checked each narration for certainty-overstating language as a distinct review step from a grammar or style check
- You can identify at least one instance where a sentence needed revision, or can explain clearly why all three passed the check
Watch out for
- Conflating a proofreading pass (grammar, tone, length) with the certainty-verification pass — they are different checks and both are necessary
- Reviewing only the headline figures and skipping narration for charts that seem lower-stakes, where overstated certainty is just as likely to slip through unnoticed
Hint
For each narration sentence, ask: if a board member asked 'how confident are we in that, really?' would the answer match the tone of the sentence as written?
- AI compresses board pack production time most effectively at the narrative and formatting layer — chart narration, executive summary drafting, appendix formatting — using platforms like Vena and Workday Adaptive Planning's Office Connect built around this workflow.
- Effective AI briefing for board narrative specifies the audience, the expected tone, and what the chart or figure needs to prove — a generic restatement prompt produces generic restatement output.
- The certainty-overstatement failure mode is the central risk: AI-drafted narrative can use confident language — "on track," "secured," "contained" — that implies more certainty than the underlying model actually supports, and a board audience is structurally less likely to catch this than a management report reader.
- Reviewing board pack narrative for certainty-overstating language is a distinct verification step from proofreading — it checks meaning against data, not grammar against style, and it must happen before every distribution regardless of how the first draft was produced.
- Sign-off from the CFO or FP&A leader remains required for board pack content regardless of AI involvement in drafting — financial communications to the board carry fiduciary weight that does not transfer to the tool.