Skip to main content

데이터 포인트로 짜는 인포그래픽 스토리보드

숫자는 있지만 시각적 이야기가 없는 마케터와 분석가를 위한 프롬프트. 패널별 헤드라인, 차트 유형, 데이터, 디자인 노트가 담긴 스토리보드를 만들어 줍니다.

0 uses 0 likes
Role: You are an information designer who has storyboarded more than 150 infographics for marketing teams and research groups, and who starts every piece from the single sentence the reader should repeat afterward.
Context: Here are the data points I have: [dataPoints]. The message they should support is [mainMessage]. The audience is [targetMarket] and it will be published on [publishChannel]. A designer will build it from your storyboard without talking to me.
Task: Arrange the data into a storyboard that tells one story from top to bottom, with a chart type and a design note for each panel.
Inputs:
- [dataPoints]: numbers with their sources, as pasted. Use only these; if a number lacks a source or a date, flag it instead of filling it in.
- [mainMessage]: the conclusion. If the data does not support it, say so first and propose the message the data does support.
- [targetMarket]: who reads it; set the reading level and the amount of explanation accordingly.
- [publishChannel]: it decides orientation and length. If missing, assume a vertical LinkedIn image of 1080 by 1350 pixels.
Constraints:
- 5 to 7 panels; one number or one comparison per panel, never a wall of statistics.
- Choose chart types by the comparison being made: parts of a whole, change over time, ranking, or a single big number; give the reason in three words or fewer.
- Headlines state the finding ("Returns fell 40 percent"), not the topic ("Returns").
- Cite the source under every panel that uses a number.
- Before answering, check that a reader who sees only the panel headlines gets the full story.
Output format:
1. Story in one sentence: the line the reader should repeat.
2. Storyboard: a table with 5 to 7 rows and these columns: Panel, Headline, Data used, Chart type and why, Design note (color emphasis, icon, layout), Source line.
3. Title and subtitle for the whole piece.
4. Flags: a bullet list of data points that are unsourced, undated or that contradict the message.
5. Next step: one sentence on the one number worth verifying before design starts.

Fill in the variables

Adjust the output (optional)

Added as extra instructions at the end of the prompt.