Data Narrative Builder

When to use

• A presentation exists but feels like a data dump rather than a story
• The analysis has a clear finding but the team doesn't know how to make it compelling
• The audience is senior and will not read more than 5 slides
• A decision needs to be made and the data needs to make a clear, emotion-aware argument
• A recurring report needs to be restructured around a narrative rather than a metric list

Process

1. Identify the central message — write the single most important thing the audience should know and do after seeing this. Everything else is supporting material. If there's more than one message, there are multiple presentations.
2. Choose a narrative framework — select the structure that fits the context. Situation–Complication–Resolution works for most problem/solution stories. Before–After–Bridge is strong for demonstrating impact. See references/narrative_frameworks.md for all patterns with examples.
3. Assign an emotional arc — map each section to an intended emotional state: establish comfort (Situation), introduce tension (Complication), offer confidence (Resolution). Tone and data emphasis should match the intended emotion at each stage.
4. Draft each section with data woven in — apply the pyramid principle: lead with the conclusion, then support it with evidence. Numbers must serve the narrative, not interrupt it. Round large numbers for recall; humanise by expressing impact in terms the audience cares about.
5. Plan the visual sequence — assign one key chart or visual to each narrative beat. The visual should reinforce the spoken or written message, not repeat it. See references/data_writing_guide.md for annotation and emphasis techniques.
6. Write the opening hook and closing call to action — the hook must earn attention in under 10 seconds (a surprising stat, a question, or a contrast). The CTA must name a specific decision, person, and deadline. Complete assets/narrative_template.md.

Inputs the skill needs

• The central finding or insight to be communicated
• The audience (role, knowledge level, what they care about most)
• The desired action or decision at the end of the presentation
• All supporting data and charts already prepared
• Format and time constraints (slides, report, 5-minute presentation, 30-page doc)

Output

• references/narrative_frameworks.md — SCR, BAB, hero journey, and sparklines patterns with data storytelling examples
• references/data_writing_guide.md — pyramid principle, number formatting, annotation, plain-language techniques
• assets/narrative_template.md — fill-in-the-blank narrative structure: hook, situation, complication, resolution, call to action

Writing rules for Simon's deliverables

Any prose Simon will read or forward must also pass simon-skrivstil and stop-slop (required by his CLAUDE.md before every delivery). Where this skill's frameworks conflict with those two (dramatic hooks, emotional arcs, metaphors), the two writing skills win: a headline carries its number, no "not X but Y".

Merged modes (2026-10-12)

The skills insight-synthesis, executive-summary-generator, technical-to-business-translator and impact-quantification were merged into this one. Their files live under references/<old-name>/.

Findings to insights (was insight-synthesis)

Use when an analysis has many statistics but no clear "so what", or several analyses need one set of recommendations.

1. List every meaningful finding as a plain factual statement before interpreting.
2. For each: So what (why it matters), Why (likeliest explanation), Now what (specific action). See references/insight-synthesis/references/insight_framework.md.
3. Give each insight a magnitude (money, customers, hours). Without a number it is an observation.
4. Score impact, confidence and actionability 1–3 each and lead with the insights that score high on all three (references/insight-synthesis/references/prioritization_guide.md). Hold high-impact, low-confidence insights until validated.
5. Cluster related insights; where two point in opposite directions, state the tension and which data would resolve it.
6. Present the top 3–5 in references/insight-synthesis/assets/insight_brief_template.md.

Executive summary (was executive-summary-generator)

Use when a detailed analysis must fit 1–2 pages for a senior reader.

1. Keep only the 3–5 findings that change or confirm a decision.
2. Every finding carries a number (revenue at risk, cost saved, customers affected, payback time); see references/executive-summary-generator/references/executive_communication.md.
3. One paragraph of situation: why the analysis was done and why now.
4. Conclusion first, then evidence, then detail (references/executive-summary-generator/references/pyramid_principle_guide.md).
5. Each recommendation names what, who, expected outcome and by when.
6. Close with the decision block: the approval needed, resources, expected return, deadline. Template: references/executive-summary-generator/assets/executive_summary_template.md.

Technical to business language (was technical-to-business-translator)

Use when model output, test results or query findings go to a business reader, and to check your own draft for jargon.

1. Flag jargon: references/technical-to-business-translator/scripts/jargon_detector.py.
2. Score readability: references/technical-to-business-translator/scripts/readability_scorer.py (target grade 10 or lower for executives).
3. Pick the reader persona (references/technical-to-business-translator/references/stakeholder_personas.md).
4. Swap terms with references/technical-to-business-translator/references/translation_pattern_library.md; metaphors for hard concepts in metaphor_bank.md (check them against stop-slop first).
5. Produce the parallel version with references/technical-to-business-translator/assets/translation_template.md and keep the technical original in an appendix.

Sizing the impact (was impact-quantification)

Use when a finding suggests an action or opportunity and stakeholders need a defensible number, or to rank a backlog by impact.

1. Classify: revenue growth, cost reduction, risk reduction or efficiency gain (references/impact-quantification/references/impact_quantification_framework.md).
2. Gather baseline, affected volume, expected change, time horizon and input confidence.
3. Point estimate: references/impact-quantification/scripts/revenue_impact.py or cost_savings.py.
4. Always give a low/base/high range: references/impact-quantification/scripts/confidence_interval.py.
5. Document every estimated input and its sensitivity (references/impact-quantification/references/assumption_documentation.md).
6. Package with references/impact-quantification/assets/impact_estimate_template.md, or business_case_template.md for larger decisions.