data-storytelling-and-business-translation

Translates the logic of data, an analysis, or an AI model into a story a decision-maker understands, using the Data-Information-Insight-Action ladder and the 'so what' test. Use when data or a model n

SKILL.md

GitHub

Data Storytelling & Business Translation

Purpose

A data analysis or an AI model doesn't produce value the moment it's technically finished — it produces value the moment someone makes a different decision because of it than they would have made without it. Many otherwise solid analyses end up as a report on a shelf because they were never translated into language the recipient can act on. This skill produces the transition from data to action: what the data shows, what it MEANS, and what the recipient should DO based on it.

Anchored in research

  • The "Engaging through data" area of data literacy (see ../data-literacy-competency-assessment/SKILL.md): synthesizing, visualizing, and narrating data is its own skill that has to be developed separately, not an automatic byproduct of good analysis.
  • The Data-Information-Knowledge-Wisdom hierarchy (Ackoff, 1989), applied here as a business Data → Information → Insight → Action ladder: raw data becomes information once it's structured, information becomes insight once it's interpreted in context, insight becomes action only once someone decides to do something based on it.
  • Minto, Barbara — the Pyramid Principle (The Minto Pyramid Principle, 1996): the answer/recommendation first, the reasoning and data after — the reverse of how analysis is usually built.

Method

  1. Place the finding on the Data → Information → Insight → Action ladder before presenting it, and identify what stage the presentation is currently at:
    • Data: raw numbers, tables — not yet structured.
    • Information: structured, visualized — tells WHAT happened.
    • Insight: interpreted in context — tells WHY it happened and what it means for the business.
    • Action: a concrete recommendation for what the recipient should do next. Most reports stay at the Information level. Don't present anything to a stakeholder until the finding has been taken at least to the Insight level, ideally all the way to Action.
  2. Test every figure/chart with the "so what" question before it goes into the presentation: "If the recipient sees this, what should they think or do differently?" If you can't answer, either drop the chart or add an interpretive sentence that answers the question — don't leave "so what" for the recipient to figure out themselves.
  3. Structure the presentation answer-first (Pyramid Principle): start with the recommendation/conclusion, not the chronology of how the data was gathered. Reasoning and data come afterward to support the claim already made, not as a long path leading up to it. Exception: if the audience is skeptical of the conclusion, it may be justified to build the evidence first — but that's a deliberate choice, not the default.
  4. Anchor the story to one critical business question, don't try to tell everything the data shows. If the analysis has several interesting findings, pick the one that answers the recipient's most important open question right now — other findings can be attached as an appendix.
  5. Show openly where the interpretation departs from the raw data. If the analysis contains assumptions or identified biases (see ../data-bias-and-quality-critical-reading/SKILL.md), mention them briefly as part of the story — don't hide them, because a hidden bias that surfaces later erodes trust in the whole analysis.
  6. Match tone and depth to the audience. For a technical audience, the model's logic and uncertainty ranges can be shown directly; for a leadership audience the same content needs to be condensed into business impact first, with technical depth left as backup material. For a mixed audience, build a layered presentation (summary first, depth available on request).

What this skill does NOT do

  • Doesn't do the data analysis or modeling itself — translates an already-existing finding into a message.
  • Doesn't replace a general presentation/storytelling skill for leadership more broadly — see ../../../change-and-communication/skills/executive-narrative-and-storyline/SKILL.md for the general storyline structure; this skill is its data-specific application.
  • Doesn't dress up or hide uncertainty to make the story more persuasive — transparency about uncertainty is part of credible data storytelling, not its opposite.
  • Doesn't confirm figures from memory — uses the inputs you provide, or marks an assumption clearly ([assumption — verify]).

Refinement notes

Areas to keep deepening with real practice:

  • your own examples of an analysis that stayed at the Information level and how it should have been taken to Insight/Action
  • a concrete presentation template for the Data→Information→Insight→Action structure (into ../../references/)
  • rules of thumb for which type of audience requires which depth of technical background to be shown

Once this section is filled in and validated in practice, update the maturity field in skills_index.json to draft, validated, or canonical (see ../../../meta/maturity_levels.md). Don't add new fields to the frontmatter — name and description are the only ones allowed (see ../../../meta/frontmatter_schema.md).

Continue from here

  • Before this: ../data-bias-and-quality-critical-reading/SKILL.md — confirm reliability before building the story on top of it.
  • Before this (if the audience's data literacy is unclear): ../data-literacy-competency-assessment/SKILL.md
  • Related skill in another pack: ../../../change-and-communication/skills/executive-narrative-and-storyline/SKILL.md — the general executive storyline structure that this skill specializes for data.
  • Related skill in another pack: ../../../prototyping-and-demonstration/skills/demo-to-business-case-bridge/SKILL.md — if the story concerns the results of a demo/PoC, use that bridge instead.
  • A ready-made skill chain for this situation: see ../../../playbooks/
  • This pack's shared guardrails: ../../CLAUDE.md

References

  • ../../references/ — the pack's shared background material
  • ../../CLAUDE.md — the pack's shared guardrails