Most chart types are never really chosen. The quarterly report used bars last time, so it uses bars again. The dashboard shipped with lines, so lines look normal. I inherited a format that way and ran with it for years before noticing I'd never once decided.
A bar chart compares values across categories with rectangular bars. A line chart connects data points to show change over time. Any publishing tool that supports charts will draw either one from the same dataset.
The textbook rule, bars for categories and lines for trends, settles the easy cases in seconds. The hard cases are the ones where the format changes what your reader walks away believing.
When to Use a Bar Chart vs a Line Chart
Reach for bars when:
- You're comparing discrete categories, like products, regions, or teams
- Each value stands on its own
- The x-axis has no natural sequence
- You want the reader to compare sizes at a glance
Reach for lines when:
- You're tracking change across a continuous stretch of time
- Direction matters more than any single value
- The gaps between points genuinely flow into each other
- You have 12+ sequential data points
Skip the line chart when categories have no inherent order (survey responses, product names) or the points land at irregular intervals.
Skip the bar chart when you're plotting 15 to 20 or more time periods, or when the trend matters more than any one bar.
Bar or Line Chart for Time Series
Say you've got monthly revenue, 12 points. Both formats render it fine, so the choice comes down to the question you want answered.
- Up to 12 points: either works. Bars make each month easy to weigh on its own. A line makes the trajectory the point.
- Past 12 points: lines win almost every time. 30 bars turn into a wall.
- Irregular intervals: leave the points unconnected. A line drawn across uneven gaps claims steady change that never happened.
Same Data Different Story
Here are six months of revenue, charted twice.
In the bars, March is a visible dip, a drop from $45K to $43K that stands alone and asks to be explained. In the line, your eye rides the overall rise and that same dip shrinks to a wobble on the way up.
The bars say revenue varied. The line says revenue grew. Both plot the same six numbers; the format picks which sentence your reader leaves with.
How Chart Format Steers Interpretation
This effect has a name: container bias. The format has already framed the data before anyone consciously reads a number.
When you embed a chart in a post, that framing locks in at publish time, and every reader who lands on the page absorbs it. If the numbers later shift, say revenue flattens and the growth story stops holding, the chart goes on asserting what it asserted on day one. That's how content debt starts.
Line Charts Imply More Than They Show
A line connects its points, and connection carries a message: something happened in between, and it happened smoothly. Monthly revenue is a snapshot. The segment between January 1 and February 1 corresponds to no recorded measurement; it's a narrative device your charting library drew for free. Bars present the same snapshots and leave the gaps blank.
You can measure the difference. Cleveland and McGill's research on graphical perception ranked visual encodings by how accurately people read them. Position along a common scale, which is how you read a bar, came out on top. Slope and angle, which is how you read a line, ranked lower. Readers judge position more precisely than implied motion, and the gap between the two is where distortion gets in.
It shows up in behavior, too. During COVID-19 reporting, researchers showed 596 participants identical case data in two formats: cumulative trend lines versus daily bars. Same 20 days of data. Participants' risk perception shifted with the container, whether cases were rising or falling at the time. The line read as control and direction; the bars read as volatility.
As Alberto Cairo argues in How Charts Lie, every chart says more than the data alone. Lines say the most by default.
Test Your Judgment
A company reports six months of revenue: $42K, $45K, $43K, $48K, $46K, $50K. Before you read on, pick the chart you'd actually publish.
The data alone can't settle it. Bars let each month stand for itself; a line sells the rise from $42K to $50K. Whichever you picked, you made a rhetorical choice, and most people make it without noticing they've made one.
Four Common Bar vs Line Chart Mistakes
Each of these redirects interpretation while passing review as a styling choice.
1. A line chart across unordered categories
Connecting "Marketing" to "Engineering" draws a slope between two things that have no order, so the slope means nothing. Use bars.
2. Bars across a long time series
Three years of monthly bars compress into noise, and the trend drowns in the density. Use a line.
3. A line across irregular intervals
January to March to September, connected, reads as uniform change across wildly unequal gaps. The slope fabricates continuity.
4. Switching formats without saying so
Bars in one section, a line in another, same metric. The framing shifts, so the perceived meaning shifts with it. I've done this one myself: swapped a bar chart that said revenue varied for a line that said revenue grew, and left the surrounding paragraphs written for the bars. The reader absorbs chart and prose as if they agree, and they no longer do. If the format has to change, change the words around it too.
When Your Data Changes After You Publish
Everything above assumes a fixed snapshot. Live data keeps arriving. Revenue posts a new month, benchmarks refresh each quarter, a survey keeps collecting responses, and a chart that worked at six data points can fail at 24. Embed six bars in a 2024 post and by now it's carrying 18, compressed to slivers.
At that point the choice stops being cosmetic and becomes architectural. Lines extend as points arrive; bars pile up clutter. Category lists grow too, and bars strain without grouping or filtering. When both dimensions keep growing, static charts break down entirely.
LiquiChart separates the data source from the visual format, so when a Google Sheet updates or a poll collects new responses, the chart reflects current data and switching from bars to a line is a configuration change.
The format question ages differently than the data question. A paragraph written to describe six quarterly bars references "each quarter" and "the gap between Q2 and Q3." When the dataset grows and the chart shifts to a line, those references do not rewrite themselves. The prose still narrates bars. The visual shows a trend. The reader processes both without noticing they describe different shapes.
So the real question widens from what your data looks like today to what it will look like in six months. A format that scales spares you that future cleanup; a frozen one accumulates it.
Six Questions Before You Commit
- Will this dataset update after publish?
- Am I comparing categories or showing direction?
- Is the continuity real, or am I drawing it in?
- How many data points now, and how many in a year?
- Does the x-axis have a natural order?
- What does this format tell the reader, and will that still be true next quarter?
If the continuity is invented, leave the gaps blank. If the point count will grow, pick the format that can absorb it. And if the data will keep changing, know that the format you commit to today is a maintenance commitment for as long as the post stays up.
Bars for categories, lines for trends answers the question you were asked. The question nobody asks is what you want the reader to believe, and for how long. A chart type is a claim. Choose the one that will still be true next quarter.