Heatmap Maker: Data Heatmaps from CSV & Spreadsheets
Paste a grid of numbers from Excel, Google Sheets or a CSV file, choose a sequential or diverging color scale and download the chart as SVG or PNG. It charts your data, not website clicks.
← All free toolsYour heatmap
Loading the correlation example…
Downloads are static images: SVG stays sharp at any size and PNG suits email or chat. Hover details and the controls stay on this page, so regenerate here to change colors or limits.
Data table
How this heatmap was built
What a data heatmap maker does
A data heatmap maker is an online tool that turns a data matrix (row labels, column labels and numeric values) into a color-coded grid, where each cell's color stands for the size of its value. Use it for correlation matrices, activity patterns, performance grids and regional comparisons. On a sequential scale, darker cells mean larger values; a diverging scale splits colors at a midpoint such as zero.
Key takeaways
Summarize raw records before you paste them.
Sequential for counts. Diverging when a midpoint such as 0 matters.
Blank and NA cells stay grey.
On Viridis, the highest values are yellow, not dark.
What is a data heatmap?
The word heatmap covers two unrelated things. Here it means a chart: a table of numbers where color sits behind, or replaces, each value. You scan it like a spreadsheet with conditional formatting, looking for the dark corner or the row that breaks the pattern.
Website heatmaps are a separate product category. Plenty of people searching for a heat map maker want that kind, so here is the difference.
| Data heatmap (this tool) | UX / website heatmap | |
|---|---|---|
| What it visualizes | A numeric data matrix | Where users click, scroll or hover on a webpage |
| Input | Your own dataset (CSV, spreadsheet) | Tracking code on a live website |
| Output | A downloadable color-coded chart | A behavior overlay on page screenshots |
| Example tools | Amridata Heatmap Maker, Datawrapper, Flourish | Hotjar, Crazy Egg, Microsoft Clarity |
This tool creates data heatmaps. If you need to track user behavior on your website, search for a website heatmap tool instead.
Types of data heatmaps
Most heatmaps you meet at work take one of four shapes, and each expects its data arranged a little differently.
| Heatmap type | What it shows | Typical data format |
|---|---|---|
| Correlation matrix | Pairwise correlation between variables (−1 to +1) | Square numeric matrix |
| Calendar / activity heatmap | Activity by day, week or month | Date column + value; here, weekdays as rows and weeks as columns |
| Pivot table heatmap | Values across two categorical axes | Row × column × value |
| Geographic heatmap | Values by region | Region name + value |
The first three draw directly in this tool. A geographic heatmap on a real map needs mapping software, though you can still compare regions as rows in a grid. To build a correlation matrix heatmap, start with the Correlation Calculator to compute pairwise r values, then paste the results here.
How to use the heatmap maker
- Format your data. A heatmap needs a matrix: items in rows, categories or periods in columns and a number in each cell. Raw transaction lists need summarizing first. The Pivot Table Maker does that, or choose the long list layout above and let the tool sum or average repeated pairs.
- Paste or upload. Copy cells from Excel or Google Sheets, or load a CSV file up to 1 MiB. Untick the label boxes if your sheet has no label row or column.
- Choose your color scale. Sequential palettes suit values that only grow from zero, such as orders. Diverging palettes suit values around a meaningful midpoint: r around 0, or attainment around 100% of target. Viridis and Cividis are safer than red-green pairs for readers with color vision deficiency because their lightness changes steadily. No palette suits every reader, so keep values in the cells when exact numbers matter.
- Export. SVG for slides and print, PNG for email or chat, CSV to reuse the matrix. Image files are static; hover details and controls stay on this page.
When to use a heatmap vs. other chart types
Use a heatmap when two categorical axes meet and there are too many cells to read one by one.
| Use case | Best chart | Why |
|---|---|---|
| Comparing values across two categorical axes | Heatmap | Color shows patterns across the whole grid at once |
| Correlation between two variables (raw data pairs) | Scatter plot | Better for raw x,y data; a heatmap suits a pre-computed correlation matrix |
| Tracking one metric continuously over time | Line chart | Continuous trends are clearer as a line |
| Comparing a few categories on one metric | Bar chart | Bar lengths are easier to compare than shades |
Calendar heatmaps work because days and weeks are buckets; a continuous trend still belongs in a line chart. If your data is raw x,y pairs rather than a pre-computed matrix, the Scatter Plot Maker is a better fit. For other cases, see how to choose a chart type.
Heatmap use cases
Marketing analytics
Put channels in rows and weeks in columns, and a stalled channel shows up as a pale streak. Swap the counts for cost per conversion and the same grid answers a new question. Customer segment by purchase frequency works the same way. See how heatmaps fit into a data analysis dashboard alongside other visualization types, or how marketing teams use Amridata.
Business performance monitoring
KPI attainment by department reads well on a diverging scale centered on 100% of target, as in the KPI example. Product teams map user cohorts against feature usage. Sales managers line up reps against deal stages to see where deals stall; related reading: sales data analysis and KPI examples.
Research and statistics
Survey response matrices, confusion matrices and gene expression grids share this layout. Learn how to interpret correlation matrices in our Correlation Matrix guide, and remember that a dark cell shows association, not cause.
Common heatmap mistakes
- Sequential colors for signed data. A correlation of −0.6 then looks like a weak positive one.
- Blank cells read as zero. A skipped survey question is not a score of 0.
- Different limits on charts you compare. Set a custom minimum and maximum so one color means one value.
- Pasting raw rows. Summarize first so each cell holds one number.
Methodology and limitations
The parser follows CSV quoting rules and detects tabs, semicolons or commas from the first line. In semicolon mode, 3,5 reads as 3.5; elsewhere, thousands separators are rejected rather than guessed. Blank, NA, N/A, NaN, null and a lone hyphen count as missing.
With limits taken from the data, a diverging scale is made symmetric around the midpoint, so equal distances above and below get equal color strength. Values outside custom limits take the end color and are reported. Limits: 1 MiB of text, 120 rows, 120 columns and 50,000 long-list records. Palette sources: Matplotlib colormaps (Viridis), Nuñez et al., PLOS ONE 2018 (Cividis) and ColorBrewer (Blues, RdBu, PuOr).
Use it when two categories meet a number.
Campaign results by week, ratings by customer segment, correlations between metrics: if each value has a row label and a column label, it fits. Check the color limits before you share the image.
Heatmap Maker questions
How the free browser tool behaves.
What data format does the heatmap maker accept?
Comma-, tab- or semicolon-separated text with column labels in the first row, row labels in the first column and numbers in every other cell. That is the layout of an Excel or Google Sheets pivot table. A long list of row, column and value records also works with the long list layout.
What is the difference between a data heatmap and a website heatmap?
A data heatmap colors the cells of a numeric table you supply. A website heatmap, from tools such as Hotjar or Crazy Egg, records where visitors click or scroll on a live page. This page makes the first kind.
Can I make a correlation matrix heatmap with this tool?
Yes. Compute the pairwise r values first, for example with the Correlation Calculator, then paste the square matrix. Set the limits to Correlation: −1 to +1 and use a diverging palette so 0 gets the neutral color.
Which color scale should I use for my heatmap?
A diverging scale, such as blue-white-red, for values that run from negative to positive or sit around a target. A sequential scale for values that only grow from zero, such as counts. Viridis and Cividis are safer than red-green pairs for readers with color vision deficiency.
Can I export the heatmap as an image?
Yes, as SVG (sharp at any size) or PNG (easy to paste into email or chat). Exported images are static, so hover details and the controls only work on this page.
Is the heatmap maker free to use?
Yes. There is no account, subscription, trial or watermark. Paste your data, adjust the chart and download it.
How are blank or missing cells handled?
They appear as grey hatched cells marked No data. They are never filled with zero and are left out of the color limits.
Is my data uploaded or saved?
No. The matrix is processed in your browser tab, and nothing is written to browser storage.