dashboard-narrator

name	dashboard-narrator
description	Translate dashboard data and charts into plain-English narrative: key insights, anomalies, recommendations — written for non-technical stakeholders who don't read charts

Dashboard Narrator Skill

When to activate

• A stakeholder needs a written summary of what the dashboard shows — not just a link to it
• Preparing a weekly or monthly business review and need narrative to accompany charts
• Your leadership team isn't reading the dashboards and you need to bring the data to them
• You have multiple metrics to synthesize into a coherent story, not a list of numbers
• Translating a complex multi-metric dashboard into an executive briefing

When NOT to use

• Root cause analysis requiring SQL queries against raw data — use /sql for that
• Building the dashboard itself — use your BI tool (Looker, Tableau, Metabase)
• Statistical analysis or hypothesis testing — use /pandas-polars or /sql
• Real-time data alerts — set those up in your BI tool's alerting system

Instructions

Core dashboard narration prompt

Translate this dashboard data into a plain-English narrative for a non-technical audience.

DASHBOARD: [name and what it tracks — e.g., "Weekly Business Review", "Growth Metrics", "Product Health"]
AUDIENCE: [who will read this — executive team / board / department head / investors]
REPORTING PERIOD: [this week / this month / Q? 202?]
COMPARISON PERIOD: [vs. last week / last month / same period last year]

METRICS (paste your data):
[For each metric provide: name, current value, prior period value, target/plan if applicable]

Example format:
- Weekly Active Users: 48,200 (↑ 3.1% vs. last week, target: 50,000, -3.6% vs. target)
- Revenue: $1.24M (↑ 8.4% vs. last week, ↑ 22% vs. same week last year)
- Conversion rate: 3.2% (↓ 0.4pp vs. last week — previously 3.6%)
- Customer churn: 1.8% monthly (↑ 0.3pp vs. last month — highest in 6 months)
- CAC: $142 (↓ 12% vs. last month — improving)
- LTV/CAC: 4.1x (stable)
- NPS: 42 (down from 48 last quarter)

CONTEXT I KNOW:
[Any business events that explain the data — product launch, marketing campaign, pricing change, seasonality, incident]

Write:
1. HEADLINE (1 sentence): What is the overall state of the business this period?
2. WINS (2-3 bullets): What improved and why it matters
3. CONCERNS (2-3 bullets): What deteriorated, the magnitude, and whether it's a trend or one-off
4. ANOMALIES (if any): Anything that doesn't fit the pattern — investigate flags
5. RECOMMENDATION: 1-2 actions the team should take based on this data
6. WATCH LIST: Metrics to monitor closely next period

Keep it under 400 words. Write for a CEO who reads it in 90 seconds.

---

Anomaly detection and explanation

I have an anomaly in my dashboard data. Help me describe it clearly and investigate possible causes.

METRIC: [metric name]
EXPECTED VALUE: [what it typically is or what the plan says]
ACTUAL VALUE: [what it is this period]
MAGNITUDE: [X% above/below expected, X standard deviations from 30-day average]
DURATION: [when it started — one-off spike or sustained change?]

SURROUNDING CONTEXT (paste nearby metrics that might be correlated):
[Other metrics from the same time period]

Possible causes to investigate:
1. [business event — did something change operationally?]
2. [data quality — could this be a tracking or logging issue?]
3. [seasonal or external — is there a known pattern or external factor?]
4. [upstream dependency — did a data source or pipeline change?]

Write:
1. A plain-English description of the anomaly (1-2 sentences) that a non-technical stakeholder can understand
2. The 3 most likely explanations, ranked by probability
3. How to determine which explanation is correct (what to check)
4. Whether this requires urgent action or monitoring

---

Multi-chart synthesis (weekly business review)

I have multiple dashboards to synthesize into a single weekly business review narrative.

BUSINESS CONTEXT:
- Company: [brief description]
- Stage: [seed / Series A / growth / established]
- Primary business model: [SaaS / marketplace / ecommerce / etc.]
- Current strategic priority: [growth / profitability / retention / expansion]

DASHBOARD 1 — GROWTH:
[Paste metrics: new users, signups, MQLs, trials, demo requests]

DASHBOARD 2 — REVENUE:
[Paste metrics: MRR/ARR, expansion, contraction, churn, NRR]

DASHBOARD 3 — PRODUCT:
[Paste metrics: DAU/WAU/MAU, activation rate, feature usage, NPS]

DASHBOARD 4 — UNIT ECONOMICS (if applicable):
[Paste metrics: CAC, LTV, payback period, gross margin]

EVENTS THIS WEEK:
[Product releases, campaigns, incidents, external news]

Write a single coherent business review narrative that:
1. Opens with the overall business health (1 sentence verdict)
2. Tells the story across growth → revenue → product → efficiency in logical sequence
3. Highlights the 2-3 most important things happening across all dashboards
4. Flags any contradictions (e.g., "activation improved but NPS dropped — worth investigating")
5. Ends with what to watch next week

Target: 500 words max. Readable in 3 minutes. No bullet points for bullets' sake — narrative prose with embedded data points.

---

Stakeholder-specific framing

Adjust the output based on who's reading:

For the CEO:

Frame the dashboard narrative for a CEO.
Focus on: Is the business on track? Where should we focus? Any urgent decisions?
Skip: Technical metric definitions, methodology notes.
Lead with the verdict, support with 3 data points, end with recommended action.

For the board:

Frame the dashboard narrative for a board update.
Focus on: Progress vs. plan, key risks, capital efficiency.
Format: 3 bullets — what went well, what didn't, what we're doing about it.
Include: Comparison to plan/target, not just period-over-period.
Avoid: Operational detail they don't need to approve or decide on.

For a functional team (marketing, product, sales):

Frame the dashboard narrative for the [marketing / product / sales] team.
Focus on: Metrics they own and can act on.
Include: Specific actions they should take based on what the data shows.
Tone: Direct, action-oriented. They want to know what to do, not just what happened.

---

Chart-to-words translation patterns

Use these patterns when describing specific chart types:

Describe a trend line chart:
"[Metric] [direction: rose / fell / remained stable] from [X] to [X] over [period], 
a [magnitude: sharp / gradual / modest] [increase/decrease] of [X%]. 
[If trend): The [upward/downward] trend began in [month] and has [continued / reversed / plateaued]."

Describe a bar chart comparison:
"[Category A] outperformed [Category B] by [X%] ([A: X] vs. [B: X]). 
[Category C] showed the largest [increase/decrease], [up/down] [X%] vs. [prior period]."

Describe a funnel:
"Of [X] [top of funnel], [X]% ([N]) reached [Stage 2], and [X]% ([N]) converted to [final stage]. 
The largest drop-off occurs at [stage], where [X]% [of those who reached this stage] did not proceed."

Describe a distribution / histogram:
"The median [metric] is [X], with [X]% of [entities] falling between [X] and [X]. 
The [right/left] tail indicates [X]% of [entities] with values above/below [threshold]."

---

Insight quality checklist

Before sending a dashboard narrative, verify:

Review this dashboard narrative for quality.

Does it:
[ ] Lead with the most important finding, not the most obvious one?
[ ] Quantify every claim (not "revenue grew" — "revenue grew 14%")?
[ ] Distinguish between correlation and causation?
[ ] Separate facts from interpretations (facts: "churn rose 0.3pp"; interpretation: "likely due to...")?
[ ] Flag what we don't know or can't explain from the data?
[ ] End with a specific action, not a vague "we should monitor this"?
[ ] Avoid jargon the audience won't understand?

If any item fails, rewrite those sections.

Example

User: Weekly dashboard — WAUs: 48,200 (+3.1% WoW, target was 50K). Conversion rate: 3.2% (was 3.6% last week). Revenue: $1.24M (+8.4% WoW). Churn: 1.8% (was 1.5% last month). NPS: 42 (was 48 last quarter). New product feature launched Tuesday. Audience: CEO and executive team.

Expected output: Opening verdict sentence ("Growth is accelerating on revenue but conversion and retention signals warrant attention"). Wins: revenue growth strong, WAU trend positive. Concerns: conversion rate drop coincides with Tuesday launch — new feature may be disrupting the signup flow; churn uptick is early-stage but 3-month trend to watch. Recommendation: A/B test the new onboarding flow against the prior version; schedule churn cohort analysis to identify if a specific segment is driving the 1.8% rate. Watch list: conversion rate and churn for next 2 weeks.

---