Data Storytelling — Metrics to Meaning

Transforms raw metrics, scores, and quantitative findings into meaningful narratives that drive understanding and action. Owns insight extraction, comparison framing, magnitude communication, and the bridge between numbers and decisions. [EXPLICIT]

Grounding Guideline

A number without context is noise. A number with context, comparison, and consequence is an insight. 92% test coverage means nothing until we know that the uncovered 8% concentrates the payment modules — exactly where risk is highest. Data metodologia-storytelling turns metrics into comprehension.

Narrative Data Philosophy

1. Context before number. Not "coverage is 92%". Yes: "the team invested in quality (92% coverage), but the uncovered 8% concentrates the critical payment modules". [EXPLICIT]
2. Always compare. Every metric needs a reference: vs. baseline, vs. industry, vs. target, vs. prior quarter. [EXPLICIT]
3. Explicit consequence. So what? → "This means that..." → "Which implies that..." → "Therefore, we recommend..."
4. Tangible magnitude. FTE-months → "equivalent to a team of 5 people for 8 months". Abstract → concrete. [EXPLICIT]

Inputs

• $1 — Data context: metrics, scoring, financial, performance, coverage (default: metrics)
• $2 — Audience: executive, technical, mixed (default: mixed)

Parse from $ARGUMENTS. [EXPLICIT]

Core Patterns

Pattern 1: Metrics-to-Meaning

Raw metric → Context → Comparison → Insight → Implication → Action

Example:
  Raw: "Deployment frequency: 1/month"
  Context: "El equipo despliega una vez al mes"
  Comparison: "vs. benchmark DORA de equipos elite: múltiples por día"
  Insight: "La brecha de 30x indica proceso manual o miedo al cambio"
  Implication: "Cada feature espera en promedio 15 días de cola antes de llegar a producción"
  Action: "Pipeline CI/CD automatizado puede cerrar la brecha a 1/semana en 3 sprints"

Pattern 2: Insight Extraction

Data point → Pattern → Anomaly → Significance → Recommendation

Steps:
1. Observe the data point: "8 de 12 módulos tienen cobertura >90%"
2. Detect the pattern: "Los módulos con alta cobertura comparten equipo senior"
3. Identify the anomaly: "Los 4 módulos sin cobertura son todos del equipo junior"
4. Interpret the significance: "No es un problema de herramientas, es de capacitación"
5. Recommend: "Pair programming cross-team + coverage gates en CI"

Pattern 3: Comparison Framing

Frame Type	When	Example
Before/After	Projected improvement	"De 12 semanas a 4 semanas de time-to-market"
Peer Benchmark	Industry comparison	"vs. mediana del sector: 3 deploys/semana"
Industry Standard	Reference frameworks	"DORA elite: <1 hora lead time"
Internal Baseline	Historical comparison	"vs. Q1: incidentes reducidos 40%"
Target Gap	Distance to objective	"A 15 puntos del objetivo de disponibilidad 99.9%"
Cost Equivalence	Making FTE tangible	"Equivalente a 3 desarrolladores senior durante 6 meses"

Pattern 4: Magnitude Communication

Abstract → Concrete → Impactful

"40 FTE-meses"
  → "Equivalente a un equipo de 8 personas durante 5 meses"
  → "Es decir, todo el equipo backend dedicado exclusivamente
     desde enero hasta mayo, sin poder hacer nada más"

"99.5% disponibilidad"
  → "43 horas de downtime al año"
  → "Equivalente a casi 2 días completos sin servicio,
     probablemente concentrados en momentos de alta demanda"

"$2M de deuda técnica" → NEVER. Use FTE-month equivalents. [EXPLICIT]

Scoring Matrix Narratives

When presenting scoring tables:

1. Lead with the pattern, not individual scores:
   "De las 6 dimensiones evaluadas, 2 están en rojo y comparten causa raíz:
    acoplamiento entre el módulo de autenticación y el core de negocio."

2. Highlight the anomalies:
   "La dimensión de seguridad sorprende en verde dado que el equipo
    no tiene un rol dedicado — evidencia de buenas prácticas orgánicas."

3. Connect to action:
   "Los 2 rojos se resuelven con el escenario B en Fase 1 (Q2);
    los 3 amarillos mejoran orgánicamente con la nueva arquitectura."

Dashboard Narrative Sequences

For multi-chart metodologia-storytelling (presentations, executive summaries):

Chart 1: The headline
  "Aquí estamos" — current state summary metric

Chart 2: The context
  "Así llegamos aquí" — trend or historical view

Chart 3: The comparison
  "Así estamos vs. donde deberíamos estar" — benchmark gap

Chart 4: The path
  "Así cerramos la brecha" — roadmap or scenario projection

Each chart builds on the previous. No standalone charts. [EXPLICIT]

Semantic Density Rules

Type	Guideline
Table footnotes	Explain methodology, not data (data goes in cells)
Semaphore criteria	Define thresholds: >80%, 50-80%, <50%
Cross-references	"→ See 03_AS-IS § Cobertura for methodology"
Source attribution	Evidence tag inline: "92% cobertura [CÓDIGO]"

Output Configuration

• Language: Spanish (Latin American, business register — simple, clear, concise, direct)
• Attribution: Expert committee of the MetodologIA Discovery Framework
• Tagline: "Construido por profesionales, potenciado por la red agéntica de MetodologIA."

Validation Gate

Criterion	Check
Every metric has context	Not just the number — the story around it
Every metric has comparison	vs. baseline, benchmark, target, or prior period
Insights are actionable	"So what?" answered for every data point
Magnitudes are tangible	FTE-months translated to team-equivalents
Scoring patterns highlighted	Not just individual scores — the story across dimensions
No naked numbers	Zero metrics without interpretation

Assumptions and Limits

• Las metricas de input ya estan calculadas; esta skill interpreta y contextualiza, no calcula.
• NUNCA presentar metricas sin contexto y comparacion.
• NUNCA usar valores monetarios para costos. Solo FTE-meses.
• Esta skill posee interpretacion de metricas y framing narrativo. NO posee diseno de visualizacion (eso es data-viz-storytelling) ni arco narrativo general (eso es storytelling).

Edge Cases

Edge Case	Handling Strategy
No industry benchmarks available	Use internal baseline (prior quarter, other team, other project). Declare explicitly: "No industry benchmark available; internal Q1 baseline used as reference [SUPUESTO]". If no internal baseline either, use standard frameworks (DORA, SRE).
Contradictory metrics	Present the contradiction as a finding itself. "High coverage (92%) contradicts incident rate (8/month), suggesting tests that do not cover real scenarios [INFERENCIA]". The contradiction IS the story.
Scarce data (<10 data points)	Acknowledge limitation explicitly: "With [N] data points, the trend is indicative, not conclusive". Use confidence intervals. Recommend collection period before definitive conclusions.
Metrics that favor inaction (everything green)	Look for the story beneath the surface: trends, degradation velocity, opportunity cost. "Everything is green today, but the trend over the last 3 quarters shows..."

Decisions and Trade-offs

Decision	Justification	Discarded Alternative
Context before number as rule	A number without reference is noise. The reader cannot evaluate "92% coverage" without knowing the target, baseline, or benchmark.	Number first: the reader forms premature judgment before having a reference frame.
Mandatory comparison in every metric	Every metric needs at least one reference: vs baseline, vs industry, vs target, vs prior quarter. Without comparison there is no insight.	Isolated metric: informative but not actionable; the reader does not know if it is good or bad.
Tangible magnitudes over abstract	"40 FTE-months" means nothing to a CEO. "The entire backend team dedicated from January to May doing nothing else" generates visceral understanding.	Abstract magnitudes: precise but not communicative for executive audience.
Narrative sequence in dashboards (4 charts)	Each chart builds on the previous: status -> trend -> benchmark -> path. Without sequence, charts are isolated data.	Independent charts: flexible but do not build cumulative argument.

Knowledge Graph

graph TD
    subgraph Core["Core: Data Storytelling"]
        M2M[Metrics-to-Meaning]
        INSIGHT[Insight Extraction]
        COMPARE[Comparison Framing]
        MAGNITUDE[Magnitude Communication]
    end

    subgraph Inputs["Inputs"]
        METRICS[Metricas Crudas]
        SCORES[Scoring Matrices]
        BENCHMARKS[Benchmarks]
        CONTEXT[Contexto de Negocio]
    end

    subgraph Outputs["Outputs"]
        NARRATIVE[Metric Narratives]
        DASHBOARD[Dashboard Sequences]
        SCORING_N[Scoring Narratives]
        PROJECTION[Projections]
    end

    subgraph Related["Related Skills"]
        STORY[storytelling]
        COPY[copywriting]
        DATAVIZ[data-viz-storytelling]
        TECHWRITE[technical-writing]
    end

    METRICS --> M2M
    SCORES --> INSIGHT
    BENCHMARKS --> COMPARE
    CONTEXT --> MAGNITUDE
    M2M --> NARRATIVE
    INSIGHT --> SCORING_N
    COMPARE --> DASHBOARD
    MAGNITUDE --> PROJECTION
    STORY --> Core
    COPY --> NARRATIVE
    Core --> DATAVIZ
    TECHWRITE --> METRICS

Output Templates

Template 1: Metrics Narrative Report (Markdown)

Filename: Data_Narrative_{project}_{dimension}_{WIP|Aprobado}.md

# Narrativa de Datos: {project} - {dimension}

## Headline
{Una metrica clave con contexto y comparacion en una linea}

## Estado Actual
| Metrica | Valor | Baseline | Benchmark | Gap | Tendencia |
|---|---|---|---|---|---|

## Interpretacion
{Parrafo denso: patron detectado + anomalia + significancia}

## Implicacion
{So what? Que significa para el negocio en terminos tangibles}

## Recomendacion
{Accion concreta que cierra el gap, con timeline estimado}

## Fuentes
| Dato | Tag de Evidencia | Confianza |
|---|---|---|

Template 3: HTML (bajo demanda)

• Filename: Data_Narrative_{project}_{dimension}_{WIP}.html
• Estructura: HTML self-contained branded (Design System MetodologIA v5). Dark-First Executive. Incluye tarjetas de métricas con semáforo, tabla comparativa interactiva y callouts de insight. WCAG AA, responsive, print-ready.

Template 4: DOCX (formal circulation)

• Filename: {fase}_{entregable}_{cliente}_{WIP}.docx
• Generado via python-docx con MetodologIA Design System v5. Portada con metadata del engagement, TOC automático, encabezados/pies de página con marca. Tablas con zebra striping, tipografía Poppins en headings (navy), Trebuchet MS en cuerpo, acentos dorados. Para circulación formal y auditoría.

Template 5: XLSX (bajo demanda)

• Filename: {fase}_{entregable}_{cliente}_{WIP}.xlsx
• Via openpyxl con MetodologIA Design System v5. Headers con fondo navy y tipografía Poppins en blanco, conditional formatting por semáforo y tendencia de métrica, auto-filters en todas las columnas, valores directos sin fórmulas.

Template 6: PPTX (bajo demanda)

• Filename: {fase}_{entregable}_{cliente}_{WIP}.pptx
• Via python-pptx con MetodologIA Design System v5. Navy gradient slide master, Poppins titles, Trebuchet MS body, gold accents. Máx 20 slides ejecutivo / 30 técnico. Speaker notes con referencias de evidencia.

Template 2: Scoring Matrix Narrative (Markdown)

Filename: Scoring_Narrative_{project}_{WIP|Aprobado}.md

# Scoring Narrative: {project}

## Patron General
{De las N dimensiones evaluadas, X estan en rojo y comparten causa raiz: ...}

## Scoring Matrix
| Dimension | Score | Semaforo | Evidencia Clave | Causa Raiz |
|---|---|---|---|---|

## Anomalias
{Dimensiones que sorprenden -- positiva o negativamente -- con explicacion}

## Conexion a Accion
{Los rojos se resuelven con [escenario] en [fase]; los amarillos mejoran con...}

## Proyeccion
{Si no se actua: tendencia de scores en N trimestres}

Evaluacion

Dimension	Peso	Criterio
Trigger Accuracy	10%	Se activa ante metricas, scores, datos cuantitativos que requieren interpretacion y contexto
Completeness	25%	Toda metrica tiene contexto, comparacion, interpretacion, implicacion, y recomendacion
Clarity	20%	Magnitudes tangibles; secuencia logica dato -> insight -> accion; cero numeros desnudos
Robustness	20%	Produce narrativas utiles sin benchmarks, con datos escasos, con metricas contradictorias
Efficiency	10%	Genera narrativa completa por metrica sin requerir multiples iteraciones
Value Density	15%	Cada metrica interpretada genera insight accionable; ratio signal-to-noise alto

Umbral minimo: 7/10

Cross-References

• metodologia-storytelling — Arco narrativo general que consume las narrativas de datos
• metodologia-copywriting — Prosa persuasiva que envuelve los insights de datos
• metodologia-data-viz-storytelling — Visualizaciones que representan las narrativas de datos
• metodologia-technical-writing — Precision documental de las metricas fuente

Edge Cases

• No benchmarks available: Use internal baseline or state explicitly: "Sin benchmark sectorial disponible; se usa linea base interna Q1 como referencia [SUPUESTO]".
• Conflicting metrics: Present the contradiction as a finding: "La cobertura alta (92%) contradice la tasa de incidentes (8/mes), sugiriendo tests que no cubren escenarios reales [INFERENCIA]".
• Sparse data: Acknowledge gaps: "Con [N] datos, la tendencia es indicativa, no concluyente".

Limits

• This skill owns metric interpretation and narrative framing. It does NOT own visualization design (that's metodologia-data-viz-storytelling) or overall narrative arc (that's metodologia-storytelling).
• NEVER present metrics without context and comparison.
• NEVER use currency values for costs. FTE-months only.

Usage

Example invocations:

• "/data-storytelling" — Run the full data storytelling workflow
• "data storytelling on this project" — Apply to current context