Stacked vs grouped bar charts: which one to use

You have quarterly revenue for three product lines, or survey answers from four regions, or headcount by department and year. A bar chart can show all of it. The question is whether the reader should compare the individual series, the totals, or the shares.

The difference in one sentence

• Grouped bars put each series side by side so you can compare one series to the same series in the next category.
• Stacked bars put the series on top of each other so you can see the total and its composition at the same time.

In both charts the categories are along one axis. What changes is whether the bars sit next to each other or on top of each other.

A worked example

Three products, four quarters:

If the question is “which product is growing fastest?” grouped bars put A, B and C side by side in every quarter, and the answer is visible: A grows steadily, B jumps in Q2 then slows, C climbs slowly.

If the question is “how big is the business, and what is it made of?” stacked bars show the total per quarter and the share of each product inside it. The total grows, and the mix barely changes.

If the question is “is the mix changing?” use a 100% stacked bar. Every bar is normalised to 100%, so the shares line up directly and the total is deliberately removed from the picture.

The problem with stacked bars

In a grouped chart, every bar starts at zero, so every bar length can be compared. In a stacked chart, only the bottom segment starts at zero. The segment above it starts at the top of the segment below — a floating baseline. That means:

• Middle segments are hard to compare. A segment that looks shorter in Q3 than in Q2 might be a different value entirely, because the segment underneath it changed.
• Small segments disappear. A series worth 2% of the total becomes a thin sliver — invisible in the chart, even though it may be the thing the report is about.
• Too many series make it unreadable. Three or four series is usually the limit; beyond that the eye cannot track which colour is which across bars.
• Negative values do not stack meaningfully. A decrease cannot be stacked above an increase. If the data contains negatives, use a waterfall chart instead.

When 100% stacked is the better answer

Use a 100% stacked bar when the question is about share, not size. “What percentage of revenue came from each product?” “What share of tickets were resolved in each category?” Normalising removes the column total from the comparison and makes the shares directly readable across categories.

The trade-off: the total is no longer visible. A 100% stacked chart can make a tiny business and a large one look identical. If the total matters, show it separately — in a table, or in a second chart.

The total-column trap

This is the most common stacked-bar mistake in real spreadsheets. The table has a “Total” column at the end, copied along with the parts:

Stack Online, Retail and Total, and the bar for Q1 is 400 instead of 200. If you paste a table like this into a stacked chart, remove the total column first — the parts already add up. ChartPaste detects this layout and offers to drop the total column for stacked charts, but the cleaner habit is to keep totals out of the charting range entirely.

A quick decision checklist

1. Compare the series to each other → grouped bar.
2. Show the total and its composition → stacked bar (three to four series, no total column).
3. Compare shares → 100% stacked bar.
4. Show increases and decreases leading to a final value → waterfall.
5. Show change over many periods → line or stacked area, not bars.

If two of these sound like the same question, the chart is probably trying to do too much. Splitting one chart into two often makes both answers clearer — and makes the report easier to read.