Stacked & 100% stacked bar charts from Google Sheets, composition over time

Stacked & 100% stacked bar charts from Google Sheets, composition over time

Stacked bar charts are the chart you use when "what's the breakdown" and "how is the breakdown changing" are the same question. They're underused, partly because plain bar charts are easier, and partly because the choice between absolute-stacked and 100%-stacked trips people up.

Both render correctly through Chartcastr. The harder question is which one to use.

The two flavours

Absolute stacked. Each bar's total height is the absolute value. Each segment is one category's contribution to that total. Use this when the total matters as much as the mix.

Example: monthly revenue split across product lines. The total height shows monthly revenue. The segments show what made it up. If revenue grows, the bar gets taller. If a product line shrinks while others grow, you see both at once.

100% stacked. Every bar is the same height, 100%, and segments show share. Use this when only the proportions matter and you want to compare them across time without the totals getting in the way.

Example: customer support tickets by category, by week. You don't care that you got 312 tickets one week and 280 the next. You care that "billing" went from 18% of tickets to 41%.

How to choose

The rule we use:

• If a stakeholder reading the chart is going to ask "is the total going up or down?", absolute stacked.
• If they're going to ask "is the mix changing?", 100% stacked.
• If they'll ask both, ship two charts.

Don't try to get one chart to answer both questions. It can't.

Shapes of data that work

Stacked bars want long-shaped data: one row per (category, time period) combination, or a wide table with categories as columns and time periods as rows. Either way, Sheets handles the chart configuration. Chartcastr reads it as you've set it up.

Common cases:

• Revenue by product line, by month. Absolute. "Total revenue is growing, and the new product line is now 30% of it."
• Traffic by source, by week. Often 100%. "Direct used to be 10% of traffic. It's now 22%."
• Spend by campaign, by week. Absolute. "We spent $80k last week, mostly Meta."
• Active users by tier, by month. 100%. "Free is shrinking as a share. Paid tiers are growing."

A worked example

Marketing keeps a sheet with weekly spend across Meta, Google, LinkedIn, and TikTok. Stacked bars, one bar per week, segments in fixed colours.

Send to #marketing every Monday. With AI summary on, the message reads:

Total spend: $61k last week, up from $54k. Meta now represents 68% of spend, up from 60%, driven by the new Lookalike push. LinkedIn flat. TikTok dropped to 4% as planned.

That summary covers both the absolute (totals) and the relative (mix). The chart shows you the trajectory. The summary names what changed.

Setting it up

1. Build the stacked bar chart in Google Sheets. Pick stacked or 100% stacked at chart creation time, Sheets calls this "Stacking" in the chart editor.
2. In Chartcastr connect the source and pick the chart.
3. Pick a Slack channel.
4. Set the cadence. Weekly is the most common. Monthly works for slower-moving data.

If you change stacking type in Sheets later, the next pulse uses the new setting, no re-import. Full reference: Premade Charts docs.

When stacked bars stop working

Two failure modes to watch for:

• Too many segments. Past 5 or so, segments become hard to distinguish. Group small ones into "Other".
• Tiny segments at the top of tall bars. They become invisible. If the smallest segment is under 3% of the total, consider showing it separately.

The companion chart when the mix dominates the story is the pie or donut, but only for a single point in time. Stacked bars are pies that learned to track time.