What is a 100% stacked bar chart?
Mechanically it is a stacked bar chart with each bar scaled to 100%. Each segment's width (or height) represents that segment's share of the category total, not its absolute value. The trade-off is clean: you lose the totals, but you gain easy proportion comparisons across categories.
When to use it
- When totals don't matter — you only want to compare shares.
- When the underlying totals vary so much that absolute comparisons would dwarf small categories.
- For survey results, where the question is "what fraction of each group answered each way?" rather than "how many total responses?"
How to read it
Each bar is the same length, so read by segment width: a wide segment means that sub-category dominates the bar; a narrow one means it's a small share. Comparing the same segment across bars tells you whether that share is larger or smaller in different categories.
Common mistakes
- Using it when totals matter. If readers need to know that one category is far bigger than another, a regular stacked chart shows that and 100% stacks hide it.
- Showing too many segments. Past five, segments are too thin to compare.
- Forgetting to mention that totals are hidden. A small footnote ("counts vary; this chart shows shares") prevents confusion.
Examples
Likert-scale survey breakdown by team. Browser market share by region. Spend mix by department. Vote share by district. Anywhere you'd describe the data with the word "share" or "proportion", a 100% stacked is a natural fit.
Frequently asked
When should I use 100% stacked instead of regular stacked?
When you care about composition (percentages) and not totals. Regular stacked shows both at the cost of harder comparisons; 100% stacked optimizes for the proportional comparison.
Should I label every segment with its percentage?
Only if you have three or four big segments. Past that, labels crowd each other and the chart becomes a wall of numbers. A legend plus tooltips is usually enough.
Can I sort categories by one segment's share?
Yes, and it's often the most useful sort: rank rows by the share of "Strongly agree" or whatever your headline segment is. The tool currently sorts by overall row order; sort manually in your data.
What's the difference from a pie chart?
A pie chart shows one composition (one whole, sliced). A 100% stacked bar shows multiple compositions side by side, which is much easier to compare than a row of pies.
Pasted your data? Open the full bar chart maker for the same tool plus more options.
Open the maker