How to Read Your X Analytics: 5 Charts That Actually Matter
Exported your X Analytics CSV? Make these five charts first.
Welcome to DataVisuals.XYZ
Pick the right chart. Then make it clear.
Notes from a working chart notebook: which chart answers which question, how to label and color it, and how to turn a pile of numbers into a point people remember. Written in plain language, with examples you can copy.
Entry 001 · same data, two charts
Chart guide
All chart types →Bar chart
Compare amounts across categories. The default chart for a reason.
Line chart
Show how a number changes over time, and compare a few trends at once.
Scatter plot
Plot two numbers against each other to see whether, and how, they move together.
Histogram
Show how values of one measure are spread out, where they cluster and how far they reach.
Heatmap
A grid of colored cells that shows patterns across two dimensions, like day and hour.
Stacked bar chart
Show totals and their parts side by side, with honest limits on what readers can compare.
Pie and donut charts
Fine for two or three parts of a clear whole. Beyond that, reach for bars.
Treemap
Nested rectangles sized by value, for parts within parts of one total.
Visualizing X (Twitter) data
X data guides →X Analytics hands you a CSV full of impressions, engagements and follower counts. These guides show which chart fits each of those numbers, how to read the patterns, and the traps that make an account look better or worse than it is.
Charting X follower growth
Total followers, daily net change and fair comparisons between accounts.
Engagement heatmaps for X posts
Day-by-hour grids that show when your posts do well, and when the data is too thin.
Network graphs of replies, reposts and mentions
Map who talks to whom on X, and keep the result readable.
Comparing tweet performance
Fair comparisons between posts, formats and months, with one viral post in the mix.
Fundamentals
All guides →- 01How to choose the right chart
Start from the question, not the chart menu. A table maps each question to a chart.
- 02Using color in charts
Pick the palette by data type, then use one highlight to point at the story.
- 03Labels, titles and annotations
Titles that say the point, labels next to the data, and notes that explain.
- 04Making charts accessible
Contrast, more than one cue, alt text with the finding, and the data as a table.
- 05Data storytelling: from numbers to a point
Find the one point, give it context, and cut everything that doesn't support it.
Tools we keep coming back to
All tools →
TableauDrag-and-drop analysis and dashboards, with a free public edition.
Power BIMicrosoft's dashboard tool, with a free desktop app for Windows.
DatawrapperClean, responsive charts, maps and tables for the web.
FlourishTemplate-based interactive and animated charts and data stories.
D3.jsThe JavaScript library for fully custom, interactive web charts.
From the blog
All posts →Common questions
How do I choose the right chart for my data?
Start with the question the chart has to answer. Comparing amounts between categories points to a bar chart, change over time to a line chart, a relationship between two measures to a scatter plot, and parts of a whole to a stacked bar or, with very few parts, a pie chart. The chart guide on this site walks through each type with examples.
What software do I need to make good charts?
Any tool works if the design decisions are right. Spreadsheets like Excel or Google Sheets handle most everyday charts. Datawrapper and Flourish are quick for publishing, Tableau and Power BI suit dashboards, and D3.js gives full control if you write code.
Can I use these guides to chart my X (Twitter) analytics?
Yes. The X data guides cover follower growth, engagement heatmaps, network graphs of replies and reposts, and comparing tweet performance, using the numbers X Analytics exports.