01
Name the reader’s question
Choose trend, rank, relationship, composition, sequence, or contribution.
Choose what the reader should understand, how your data is structured, and where the chart will be published. Get a chart recommendation, the data requirements, a useful alternative, and an editable starting point—no sign-in required.
01
Choose trend, rank, relationship, composition, sequence, or contribution.
02
Account for newsletter, social, slide, or report constraints.
03
Review the data shape, caution, and alternative before opening a template.
Chart selection cheat sheet
The message comes first, but the data still has to support the visual promise. Use this table to validate the recommendation before you design.
| What the reader should see | Best starting point | Data required |
|---|---|---|
| Change over time | Line or trend chart | Ordered dates and one or more measures |
| Rank categories | Sorted horizontal bars | Distinct categories with one comparable measure |
| Compare two moments | Slope chart | The same categories at two endpoints |
| Show a relationship | Scatter plot | Two numeric values per observation |
| Show composition | Stacked bar or proportion chart | Parts that add to one defined whole |
| Explain a total change | Waterfall chart | A start, signed contributions, and an end |
| Explain event sequence | Timeline | Dated milestones with concise labels |
Designed for repeat publishers
Find the visual that supports the issue’s headline, then use a reading-column layout with direct labels and a source that survives forwarding and screenshots.
Separate the analytical relationship from decoration, document the source and caveats, and make the chart understandable outside the article context.
Choose a format consistent with the variables and comparison, then preserve definitions, uncertainty, methodology, and exact lookup where readers need it.
Reuse one verified dataset across square, portrait, and landscape exports without changing the underlying claim or losing provenance.
Read the complete chart selection framework, including line vs. bar, specialized formats, data-shape checks, baselines, and a pre-publish workflow.
A chart type chooser matches the question you want the reader to answer with a visual format that fits the structure of your data. This tool also accounts for the publishing surface and explains the tradeoffs behind its recommendation.
Use a line or trend chart when dates are ordered and the path matters. For exactly two meaningful endpoints, consider a slope chart. For several series with different units or starting values, index them to a common baseline before comparison.
A sorted horizontal bar or ranking chart is the strongest default. It makes both order and distance visible and handles long category labels well.
Use a scatter plot when every observation has two numeric measures and the story concerns their relationship, clusters, or outliers. Do not imply that correlation proves causation.
Use a pie or donut only for a few clearly different, mutually exclusive parts of a meaningful whole. Choose a sorted bar when readers need precise comparisons or when there are many categories.
The message still determines the chart type. For layout, use a strong takeaway title, direct labels, readable mobile type, and a visible source. A 16:9 or reading-column landscape canvas is a useful starting point.
Use the chart that accurately expresses the comparison, then compose it for a square or 4:5 canvas. Put the takeaway, units, time period, and source on the graphic because it may circulate without the post text.
Yes. You can use the recommendation tool without an account. If a matching Charted template exists, you can open it as an editable starting point.