Chart Big Idea
Every chart needs a claim: the Big Idea, the sentence it should prove at a glance.
When to use
Use before:
- creating a Flint chart
- writing a chart title or caption
- deciding between chart types
- turning a table into a figure
- reviewing whether a visual artifact communicates the intended point
Shared Knowledge Gate
Before selecting a consequential chart claim, stance, or recommendation, consult
scout-knowledge-base for
relevant prior decisions, failure modes, procedures, and gotchas. Read only
records whose signals or preconditions match the visual decision, and treat them
as context rather than authority. If unavailable, state that once and continue
with the current evidence; do not create a local fallback while framing.
Output
Produce a short chart brief:
Big Idea:
Audience:
Decision or use:
Measure and units:
Comparison:
Recommended chart type:
Risk to avoid:
Method
- Name the reader's task. What decision, scan, or explanation does the chart support?
- Find the comparison. Time trend, category ranking, distribution, relationship, composition, flow, or target progress.
- State the claim. Write one sentence that someone could argue with.
- Check the evidence. Confirm the data can actually support the claim.
- Choose the stance. Neutral, diagnostic, cautionary, comparative, or prescriptive.
- Compress. Remove words until another removal would damage the claim.
Title rules
Good chart titles state a finding:
| Weak |
Strong |
| Revenue by quarter |
Scout shows the fastest revenue climb by Q4 |
| Support tickets by hour |
Midday is the busiest support window |
| Implementation duration |
Most implementations finish inside four weeks |
Use subtitles for scope, population, time period, and units.
Tests
A chart Big Idea is ready when:
- it names what changed, differs, clusters, leads, lags, or matters
- the chart's encodings directly support the claim
- the title still makes sense if the reader sees only the visual for three seconds
- it avoids overclaiming beyond the data
Failure modes
- Topic title: names the metric but not the finding.
- Unsupported claim: title says "best" or "fastest" without data that proves it.
- Decorative surprise: chart type attracts attention but weakens the comparison.
- Missing stakes: the figure is true but irrelevant to the reader's decision.
1---2name: chart-big-idea-23description: Distill the central claim a chart should communicate before choosing the chart type, title, or encodings. Use when creating chart titles, figure captions, visual briefs, or deciding whether a chart earns its place.4---56# Chart Big Idea78Every chart needs a claim: the Big Idea, the sentence it should prove at a glance.910## When to use1112Use before:1314- creating a Flint chart15- writing a chart title or caption16- deciding between chart types17- turning a table into a figure18- reviewing whether a visual artifact communicates the intended point1920## Shared Knowledge Gate2122Before selecting a consequential chart claim, stance, or recommendation, consult23[`scout-knowledge-base`](../../skills/scout-knowledge-base/SKILL.md) for24relevant prior decisions, failure modes, procedures, and gotchas. Read only25records whose signals or preconditions match the visual decision, and treat them26as context rather than authority. If unavailable, state that once and continue27with the current evidence; do not create a local fallback while framing.2829## Output3031Produce a short chart brief:3233```text34Big Idea:35Audience:36Decision or use:37Measure and units:38Comparison:39Recommended chart type:40Risk to avoid:41```4243## Method44451. **Name the reader's task.** What decision, scan, or explanation does the chart support?462. **Find the comparison.** Time trend, category ranking, distribution, relationship, composition, flow, or target progress.473. **State the claim.** Write one sentence that someone could argue with.484. **Check the evidence.** Confirm the data can actually support the claim.495. **Choose the stance.** Neutral, diagnostic, cautionary, comparative, or prescriptive.506. **Compress.** Remove words until another removal would damage the claim.5152## Title rules5354Good chart titles state a finding:5556| Weak | Strong |57| --- | --- |58| Revenue by quarter | Scout shows the fastest revenue climb by Q4 |59| Support tickets by hour | Midday is the busiest support window |60| Implementation duration | Most implementations finish inside four weeks |6162Use subtitles for scope, population, time period, and units.6364## Tests6566A chart Big Idea is ready when:6768- it names what changed, differs, clusters, leads, lags, or matters69- the chart's encodings directly support the claim70- the title still makes sense if the reader sees only the visual for three seconds71- it avoids overclaiming beyond the data7273## Failure modes7475- **Topic title:** names the metric but not the finding.76- **Unsupported claim:** title says "best" or "fastest" without data that proves it.77- **Decorative surprise:** chart type attracts attention but weakens the comparison.78- **Missing stakes:** the figure is true but irrelevant to the reader's decision.
fabioc-aloha