Narrative Frameworks for Data Analysis
SCR (Situation, Complication, Resolution)
The most versatile framework for analytical communication.
Situation: What is the current state? Establish shared context — the audience should nod along. Keep brief; this is not the point.
Complication: What has changed or what is the problem? This is where you introduce tension. The complication is the reason the audience should care.
Resolution: What do you recommend, and why? The answer to the complication. This is the primary content.
Example:
- Situation: Revenue grew 12% YoY in Q3, consistent with our 10–15% target.
- Complication: However, the growth is entirely driven by price increases. New customer volume declined 8% YoY for the second consecutive quarter.
- Resolution: We need to investigate acquisition funnel conversion, starting with the paid channel where the steepest volume decline appears.
Pyramid Principle (Answer First)
Lead with the conclusion, then support it with evidence. Opposite of academic writing.
Structure:
- Governing thought (the answer)
- Key lines (3 supporting arguments)
- Supporting data (evidence for each key line)
When to use: Executive presentations, written memos, any situation where the audience has limited time.
When not to use: When you need to build context before the conclusion will land — e.g., when the conclusion is counterintuitive and needs evidence before the audience will accept it.
SPB (Spark, Problem, Bridge)
Used for short narratives and insight summaries.
Spark: The hook — a surprising number, a counterintuitive finding, a trend the audience didn't expect.
Problem: Why the spark matters. What decision or risk does it create?
Bridge: The path forward. What do we do with this?
STAR (Situation, Task, Action, Result)
Better suited to retrospective narratives (post-mortems, case studies).
- Situation: Context
- Task: The objective or challenge
- Action: What was done
- Result: What happened, measured
Choosing a framework
| Situation | Framework |
|---|---|
| Executive briefing, time-constrained | Pyramid / Answer First |
| Complex problem requiring shared understanding first | SCR |
| Short insight callout (Slack, email, chart annotation) | SPB |
| Post-mortem or retrospective | STAR |
| Multiple findings, equal weight | Horizontal structure (side-by-side key points) |
Common narrative failures
Burying the lede: Spending three paragraphs on methodology before stating the finding. The audience disengages before they reach the point.
Findings without so-what: "Revenue was $12M in Q2" is a fact, not an insight. Add: compared to what? Is that good or bad? What does it imply?
Precision without certainty: Reporting a number to four decimal places when the underlying data has ±5% error implies false confidence.
Missing the action: Every analytical narrative should end with a clear next step, a question answered, or a decision enabled. "Interesting, isn't it?" is not a conclusion.