Purpose
Turn a set of financial numbers into a clear, decision-driving narrative that a non-finance audience can follow and act on — structured around a context → insight → implication → recommendation arc with a single headline finding, plain-language framing, and pre-empted objections. Solves the recurring problem of analysts presenting accurate data that fails to land because the story, the "so what," and the ask are buried.
When to Use
Invoke this skill when the user asks to:
- Present monthly, quarterly, or annual financials to executives or a board
- Turn a variance report or budget-vs-actuals into a story that drives a decision
- Explain a financial trend or anomaly to non-finance stakeholders
- Build the narrative for a business review, forecast update, or investment ask
- Translate a spreadsheet of results into a slide narrative or talking points
- Make a recommendation (cut spend, reallocate budget, approve investment) backed by numbers
Inputs Required
Before starting, collect:
- The financial data: Key figures, metrics, or variances (e.g., revenue, margin, spend, forecast vs. actual) — pasted values or a summary
- The audience: Who is in the room and their finance fluency (e.g., CEO, product leads, board, department heads)
- The decision or ask: What you want the audience to do or approve after seeing this
- Time period and comparison basis: What period this covers and what it's measured against (prior period, budget, forecast, YoY)
- The business context: What was happening operationally that explains the numbers
- Known sensitivities: Any figure that will be questioned, resisted, or is politically charged
If the decision/ask is missing, default to framing the strongest actionable recommendation the data supports and flag it as an assumption.
Instructions
Step 1: Identify the Single Headline Finding
Distill the entire dataset into one sentence a non-finance person would repeat to a colleague.
- State it in plain language with one anchor number and its direction (up/down/on-track/off-track)
- Avoid finance jargon in the headline — no "unfavorable variance," say "we spent more than planned"
- The headline is the spine; everything downstream supports it
- Test: if the audience remembers only this sentence, do they know what happened and why it matters?
Step 2: Build the Context Layer
Set the stage so the numbers have meaning before they appear.
- Baseline: What was expected, budgeted, or normal for this period
- The comparison frame: Against what are we measuring (budget, prior year, forecast)
- The operational backdrop: The 1-2 business events that shape how to read the numbers
- Keep to 2-3 sentences — context orients, it does not lecture
Step 3: Deliver the Insight
Present what the data actually shows, translated for the audience.
- Lead with the number that matters most, then the two or three supporting figures — not a full table dump
- Convert finance terms to business language: "gross margin fell 4 points" → "we kept 4 cents less of every dollar we sold"
- Quantify the gap: how big, versus what, and whether it's a trend or a one-time event
- Separate signal from noise — explicitly name what is meaningful and what is timing/noise
Step 4: Explain the Implication
Answer the "so what" — connect the number to business consequences the audience cares about.
- Translate the financial movement into operational or strategic impact (runway, hiring capacity, margin cushion, growth pace)
- Project it forward: if this continues, what happens by quarter-end or year-end
- Tie it to something the audience owns or is measured on
- Name the risk and the opportunity — implications run both directions
Step 5: Make the Recommendation
Close with a specific, decidable ask.
- State the recommended action in one sentence, then the rationale in two
- Make it decidable in the room: what to approve, stop, continue, or investigate
- Include the expected financial outcome of taking the action (and of not taking it)
- Offer one alternative or fallback if the primary ask meets resistance
Step 6: Pre-Empt Objections
Anticipate the 2-3 questions or pushbacks the audience will raise.
- For each: state the likely objection, then the one-line factual response
- Address the known sensitivities from intake directly rather than hoping they don't come up
- Distinguish "we don't know yet" from "we know and here's the answer"
Step 7: Add Visuals Guidance
Recommend how to show, not just tell, the key numbers.
- One chart per key point (e.g., waterfall for variance drivers, line for trend, bar for period comparison)
- Name the single takeaway each visual should make obvious
- Flag any number better shown as a callout stat than a chart
Output Format
- Structure: Headline finding up top, then five labeled narrative sections following the arc, then Objections and Visuals guidance
- Length: 700–1,000 words
- Format: Structured document with named sections and short scannable paragraphs; talking points as bullets where useful
- Sections: Headline Finding → Context → Insight → Implication → Recommendation → Anticipated Objections → Visuals Guidance
- Tone: Confident FP&A partner speaking to smart non-finance people — plain-language, direct, no hedging, no jargon left untranslated
Quality Checklist
Before delivering output, verify:
- [ ] The headline finding is one sentence with a direction and an anchor number, and contains no finance jargon
- [ ] Every finance term used in the body is translated into business language at least once
- [ ] The Implication section projects the number forward and ties it to something the audience owns
- [ ] The Recommendation is a single decidable ask with the financial outcome of acting and not acting
- [ ] At least 2 specific anticipated objections are listed with factual one-line responses
- [ ] Each visual recommendation names the one takeaway it should make obvious
Notes
- The most common failure is a strong insight with a weak or missing implication — the "so what" is where non-finance audiences disengage; over-invest there
- Resist including every number; a narrative with three well-chosen figures beats a slide of twelve
- If the recommendation depends on a figure the audience will resist, address that sensitivity in Context before it becomes an objection later
- For recurring reports (monthly reviews), the Context layer can be templated and only the operational backdrop refreshed each cycle
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills