Turn Pasted Spreadsheet Data into Three Charts with Takeaway Titles

Paste your spreadsheet data and get three chart recommendations, each with a chart type, axis setup, a title that states the finding, and a confidence rating.

You getTakeaway title, chart type and why and X axis

Prompt

Prompt
You are a data visualization analyst who prepares charts for business presentations.

I will paste spreadsheet data right after this message. It may be copied cells, CSV or a table. Recommend three charts, each showing a different insight from the data.

Rules:
1. Each chart makes a distinct point. Do not show the same finding in two chart shapes.
2. Write each title as a full sentence that states the finding and includes a number, such as "Online sales grew 46% in six months." Do not use labels like "Sales by Month."
3. Match the chart type to the comparison:
   - line for change over time
   - bar for ranking or comparing categories
   - stacked or 100% bar for parts of a whole
   - scatter for relationships
   No 3D charts, and no pie charts with more than 4 slices.
4. Use only numbers that are in the data or calculated from it, and show each calculation.
5. Flag thin evidence. This includes fewer than 6 data points, a single period, one outlier driving the result, or missing values. When evidence is thin, say so and soften the title.

Output format:
For charts 1 to 3:
- Takeaway title
- Chart type and why
- X axis, Y axis and series, using the exact column names
- Supporting calculation
- Confidence: High, Medium or Low, with a one-line reason

Then add "Data gaps": up to 4 bullets on what missing data would make these charts more reliable.

If the data has no headers or mixed units, ask up to 3 numbered questions first and wait for my answers.

Example output

At a glanceOnline revenue by month, January to JuneOnline sales grew 46% from January to June
  • Month 1 (Jan): $12.4k
  • Month 2: $13.1k
  • Month 3: $14.0k
  • Month 4: $15.2k
  • Month 5: $16.8k
  • Month 6 (Jun): $18.1k

Example for: a small coffee roaster's monthly revenue (in $k) by channel, January to June, with online ad spend.

Chart 1

Takeaway title: Online sales grew 46% from January to June

Chart type: Line chart. It shows a steady trend over time, and the month-by-month rise is the story.

Axes: X = Month; Y = Revenue ($k); series = Online

Calculation: (18.1 − 12.4) / 12.4 = 46.0%. Online rose every month: 12.4, 13.1, 14.0, 15.2, 16.8, 18.1.

Confidence: Medium. The rise is consistent, but six months cannot rule out seasonality.

Chart 2

Takeaway title: Wholesale fell from 52% to 40% of total revenue

Chart type: 100% stacked bar, January vs June. It shows the shift in mix more clearly than three separate lines.

Axes: X = Month (Jan, Jun); Y = Share of total revenue (%); series = Online, Wholesale, Café

Calculation:

  • January total: 12.4 + 22.0 + 8.1 = 42.5. Wholesale share: 22.0 / 42.5 = 51.8%.
  • June total: 18.1 + 19.4 + 10.5 = 48.0. Wholesale share: 19.4 / 48.0 = 40.4%.
  • Online share moved from 29.2% to 37.7%.

Confidence: High. These are direct shares of reported totals with no gaps.

Chart 3

Takeaway title: Online sales rose alongside ad spend, but six months is too few to call it cause and effect

Chart type: Scatter plot. It tests whether two measures move together.

Axes: X = Ad Spend ($k); Y = Online ($k); one point per month

Calculation: Ad spend rose from 1.0 to 2.2 (+1.2k). Online rose from 12.4 to 18.1 (+5.7k). The points trend upward with no month breaking the pattern.

Confidence: Low. There are only 6 points, and ad spend and time rise together. Any channel growing over time would show the same pattern. The title is softened for this reason.

Data gaps

  • At least 12 to 24 months of history, to separate the trend from seasonality
  • Online order count and average order value, to show whether growth comes from more customers or bigger baskets
  • Wholesale revenue by account, to check whether one lost client explains the 11.8% drop
  • A month with ad spend paused or cut, to test the ad spend link fairly
Written by AI from this prompt. Your results will vary.

Tips and common edits

Why this works

  • Takeaway titles with a number force each chart to state a finding instead of naming a topic.
  • Fixed rules for matching chart type to comparison stop default pie charts and mismatched visuals.
  • The thin-evidence rule makes the model downgrade confidence and soften claims instead of overselling a weak pattern.

Use before building a slide deck or report. Have your table ready to paste: headers plus rows, copied from Excel or Google Sheets. Works in ChatGPT, Claude or Gemini.

Common edits

  • Ask for five charts instead of three for a longer report.
  • Add your audience, such as "for a board meeting," so titles focus on money and risk.

Prompt details

By
Prompts-GPT editors
Published

Did this prompt work?

Keep exploring

More in Productivity