Skip to content This library is free, offered as dana. If it helps your work, donate here →
penguin tree ai
← back to libraryDonate here

This is one of 3,622 free prompts in the penguin tree ai library. Sign up once and the whole library opens, with new prompts by email.

open the library →

KPI review synthesizer

Transforms raw KPI data into an executive-ready narrative summary with performance highlights, lowlights, trend analysis, root-cause hypotheses, and ranked recommended actions. Use when preparing a weekly, monthly, or…

download .md

Purpose

Convert a set of KPI figures into a structured, executive-ready narrative that separates signal from noise — surfacing what improved, what slipped, why it likely happened, and what leadership should decide or do next. Eliminates the manual grind of writing the "story" behind the numbers every review cycle.

When to Use

Invoke this skill when the user asks to:

  • Summarize weekly, monthly, or quarterly KPI results for executives
  • Turn a metrics dashboard or spreadsheet into a written performance narrative
  • Prepare the commentary section of an operating review or board deck
  • Identify highlights, lowlights, and off-track metrics from a set of numbers
  • Explain why key metrics moved and recommend corrective actions
  • Draft a standing performance update for a leadership standup or all-hands

Inputs Required

Before starting, collect:

  • KPI data: The metrics with current-period values — pasted table, list, or dashboard export
  • Review period: The period being reported (e.g., "March 2025," "Q2," "week ending 5/12")
  • Comparison baseline: Prior period and/or target for each KPI so movement can be assessed
  • Metric ownership (optional): Which team or leader owns each KPI, for accountability tagging
  • Known context: Any events that affected results (launches, outages, seasonality, hires, price changes)
  • Audience: Who reads this (CEO, exec team, board) and their tolerance for detail
  • Threshold rules (optional): What counts as on-track vs. at-risk (e.g., "within 5% of target = green")

If targets or prior-period figures are missing for some metrics, flag them as "no baseline available" rather than inventing comparisons. If threshold rules are absent, default to: within 5% of target = on-track, 5–15% off = watch, >15% off = at-risk.

Instructions

Step 1: Classify Every Metric

For each KPI, compute movement and assign a status:

  • Delta vs. prior period: absolute and percentage change
  • Delta vs. target: absolute and percentage gap
  • Status: On-track (green) / Watch (yellow) / At-risk (red) — using the provided or default thresholds
  • Direction of good: Confirm whether up or down is favorable for each metric before judging (e.g., churn down is good; a rising delta is bad)

Do not narrate anything yet — build the classification table first so the narrative is grounded in the same numbers.

Step 2: Identify Highlights (Top 3–5)

Select the strongest results — biggest favorable moves, targets beaten, or trends that reversed positively. For each highlight write:

  • The metric, the number, and the movement (e.g., "Net revenue retention rose to 118%, up 6 pts vs. February and 3 pts above the 115% target")
  • One sentence on why it matters to the business

Prioritize by strategic weight, not just size of the percentage swing — a 2-point NRR gain often outranks a 40% jump in a minor vanity metric.

Step 3: Identify Lowlights (Top 3–5)

Select the metrics that slipped, missed target, or show a worrying trend. For each lowlight write:

  • The metric, the number, and the movement — stated plainly, no softening
  • Root-cause hypothesis: The most likely driver, labeled as a hypothesis when unconfirmed
  • Severity: Is this a one-period blip, an emerging trend, or a structural problem?

If a metric is at-risk but explained by known context from the inputs, name the context explicitly.

Step 4: Trend & Interaction Analysis

Look across metrics for the story numbers tell together:

  • Leading vs. lagging: Note where a leading indicator (e.g., pipeline, activation) predicts movement in a lagging one (e.g., revenue, retention)
  • Tradeoffs: Flag cases where one metric improved at another's expense (e.g., CAC fell but so did lead quality)
  • Consistency check: Call out any metric that appears to contradict the overall narrative

Step 5: Recommended Actions (Ranked)

Produce 3–6 actions ordered by expected impact. Each action includes:

  • What to do — specific and decision-ready, not "monitor closely"
  • Which metric(s) it addresses
  • Suggested owner (if ownership was provided)
  • Type: Decision needed from leadership / Action already underway / Investigation required

Step 6: Executive Summary (Written Last)

Write a 3–5 sentence top-line summary that leads with the single most important takeaway of the period, states overall trajectory, and previews the one decision leadership most needs to weigh in on. This goes at the top of the output but is written last so it reflects the full analysis.

Output Format

  • Structure: Executive summary first, then a status table, then narrative sections
  • Length: 500–800 words plus the classification table
  • Format: Executive summary as short prose; metric classification as a table; highlights, lowlights, and actions as labeled bullet blocks
  • Sections: Executive Summary → KPI Status Table → Highlights → Lowlights (with root-cause) → Trend & Interaction Notes → Recommended Actions (ranked)
  • Tone: Chief-of-staff register — concise, candid, non-defensive; states misses without spin and wins without hype

Quality Checklist

Before delivering output, verify:

  • [ ] Every KPI provided appears in the status table with a status label and both deltas (vs. prior, vs. target)
  • [ ] Direction-of-good is correctly applied — no metric is called a win or miss in the wrong direction
  • [ ] Each lowlight includes a root-cause hypothesis, clearly labeled as a hypothesis when unconfirmed
  • [ ] Recommended actions are ranked and specific enough to act on — none say only "monitor" or "keep an eye on"
  • [ ] Metrics missing a baseline are flagged as such rather than given fabricated comparisons
  • [ ] The executive summary leads with the single most important takeaway, not a generic recap

Notes

  • The value of this summary is judgment, not restatement — a good synthesis tells leadership what to worry about and what to ignore, not just what the numbers are
  • When context explains a miss, name it, but avoid using context to excuse structural problems; distinguish blips from trends explicitly
  • If the data spans three or more periods, prioritize trend direction over single-period snapshots — a metric that's green this month but declining for three straight months belongs in lowlights
  • Keep the same metric order across recurring reviews so leadership can scan consistently period over period

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:

  1. Drag this file into a new chat, or copy and paste its contents
  2. 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