A scatter plot with 3 variables in Excel is a 2-axis chart where a third numeric value is encoded through point size, point color, or a category label, because the chart canvas only has X and Y dimensions. In Excel the data layout is straightforward: one column holds the X values, a second column holds the Y values, and a third column holds the numeric or categorical value that drives the bubble size or color group. The browser-based Scatter Plot Maker handles the first two columns by accepting one strict x,y pair per non-empty line, drawing one blue circle for each valid pair, and reporting invalid, valid, and over-limit rows as three separate numbers so nothing is silently dropped. The third column lives in Excel as a bubble-size field, a color rule, or a category label that you reattach after plotting. This guide walks through the Excel column prep, the paste-and-plot step, the rules that govern what counts as a valid pair, and the honest limits of a 2-axis SVG output.
What "3 Variables" Actually Means in a Scatter Plot
A standard scatter chart has only two axes, so any chart advertised as "3 variables" must borrow a third visual channel from the marker itself. Excel's Bubble chart uses the size of each marker; a colored scatter chart uses hue or a categorical group; some workflows add a text label per point. The numeric value that lives in the third column never becomes a third axis on a 2D canvas, it becomes a property of each plotted circle.
This is the first thing to clarify before opening any tool: the underlying relationship is still between X and Y, and the third column is a derived visual layer. Plotting the Excel columns the standard way keeps that structure intact, and a 2-axis scatter plot exported from those two columns will look identical for every marker regardless of the third value until you reattach the bubble size or color rule.
Excel Data Layout and the Browser Tool's Input Rules
Before any chart can be drawn, the three columns in Excel need to be split into a 2-axis input and a third-value side channel. The Scatter Plot Maker reads only the X and Y columns; the third column is held back for Excel's Bubble chart or color rule and is not fed into the parser.
| Row classification | What it looks like | What the tool does |
|---|---|---|
| Valid x,y pair | "3.5, 12" or "-2, 1.2e3" | Plotted as one blue circle at the linear coordinate |
| Invalid row | "3.5, 12 kg", ", 12", "0xFF", "NaN", "5e15" | Skipped and counted in the invalid total; never placed on an axis |
| Over-limit row | Any non-empty row past the first 200 checked rows | Counted as an over-limit skip, kept separate from invalid |
| Blank line | Empty whitespace between data rows | Ignored entirely; does not count toward any total |
The parser also enforces a hard ceiling on input size of 50,000 UTF-16 code units, accepts signed values, decimals, zero, and scientific notation, and normalizes negative zero to plain zero. Hexadecimal, Infinity, NaN, magnitudes above 1e12, unit text, and rows with the wrong number of commas are rejected outright. Because the three counters (valid, invalid, over-limit) are always shown after generation, including when one of them is zero, the user can audit exactly how much of the pasted block was actually used.
Build a 2-Axis Scatter Plot from Your Excel Columns
- In Excel, copy the X column and the Y column only, leaving the third-value column behind. Each row should read as a single line of text in the form x,y, for example 3.5,12.
- Open the Scatter Plot Maker and type an optional title in the title field; the title is XML-escaped before it is added to the SVG, so plain text with apostrophes or ampersands stays as visible text rather than breaking the markup.
- Paste the x,y rows into the input box, one strict pair per non-empty line. Do not add column headers, currency symbols, or units; remove any blank rows you want ignored or leave them blank and the parser will skip them.
- Click generate. The tool checks the first 200 non-empty rows, places each valid pair as a blue circle, and shows three counters: valid points drawn, invalid rows skipped, and over-limit rows past the boundary.
- Inspect the axis labels and the five reference lines per axis. Both axes are independent linear scales; the smallest x maps to the left edge, the largest x to the right, the smallest y to the bottom, the largest y to the top. Zero is not added automatically, so read the labels before interpreting a cluster's apparent position.
- Download the standalone 800 by 500 SVG. The on-page preview and the downloaded file come from the same percent-encoded SVG string, so they are pixel-identical. Editing the title or the data clears that string and disables the old download.
Add the Third Variable Through Size, Color, or Group
The Scatter Plot Maker treats every valid pair as a single blue circle of the same size, so it cannot encode the third Excel column directly. Reattach the third value in Excel after exporting the SVG. In Excel, a Bubble chart uses the third column to scale each marker; a colored scatter chart uses it through a conditional formatting rule, a series-by-category structure, or a manual color mapping. The exported SVG remains the clean 2-axis foundation for either of those Excel-side embellishments.
A useful workflow is to plot the X and Y columns through the browser tool to confirm the cluster shape, the axis ranges, and the per-row titles that the SVG embeds, then return to Excel to layer bubble size or category color on top. Because each circle in the SVG carries a text title with its exact formatted x and y values, hovering or inspecting the SVG in an SVG-aware viewer shows the original pair, which is useful for sanity-checking which row produced which marker.
Read the Row Counts, Axis Labels, and SVG Output
Every generation step shows three counters: valid points drawn as circles, invalid rows that the parser rejected, and over-limit rows past the first 200 checked lines. Zero values for any of those three are still displayed, so the summary never hides what was skipped. If no valid pair survives, the tool reports an error and clears any older chart, instead of silently producing a blank canvas.
When every x value is identical, the smallest x and the largest x would otherwise be equal and the coordinate calculation would divide by zero. The tool handles this by expanding the displayed x domain symmetrically around the constant. The expansion is ten percent of the absolute constant, or one unit, whichever is larger, so a column of x = 5 produces a domain of [5 − 1, 5 + 1] = [4, 6] because 10% of 5 is 0.5 and the rule takes the larger value of 1. The same logic applies independently to a constant y column. The result is a single visible point or vertical line of points centered on the chart instead of a generation failure.
Both axes use independent linear scales, so two points whose x values differ by 1 occupy more horizontal space when the x range is small than when the x range is large. This makes the chart honest about unequal numeric gaps but also means that small datasets can look stretched out. Coordinates in the SVG are rounded to three decimal places, axis labels use practical significant digits or scientific notation for extreme values, and the title is XML-escaped so a closing tag in the user text stays as visible text and cannot inject SVG elements.
When a 3-Variable Plot Needs a Different Tool
The Scatter Plot Maker is a 2-axis visualization, not a statistical engine. It does not calculate correlation, regression, a trend line, confidence intervals, error bars, density, clusters, or significance, and the on-page FAQ confirms that explicitly. It also does not assign colors by group, draw logarithmic axes, label overlapping points, or scale bubbles by a third column. All parsing, layout, XML creation, preview, and download happen locally in the current browser tab, so nothing is uploaded, but the chart itself is a single SVG file that should be paired with the original spreadsheet for any quantitative claim.
For dense datasets, overlapping identical pairs, categorical color, accessible labeling, log scales, publication styling, or formal statistical analysis, use a dedicated plotting package and keep the source rows. The browser tool is the right fit for small experiments, paired observations, calibration samples, classroom demonstrations, rough correlation checks, and quick visual inspection of an X and Y column before the third variable is layered on in Excel. Verifying the axis labels, the three row counters, and a few embedded point titles in the exported SVG before sharing the file takes only a moment and catches the most common mistakes.
For a deeper look, see Create Yes or No on Excel Without a Spreadsheet Formula.