Pie & donut charts from Google Sheets to Slack, the only place we still recommend them
Pie charts get a bad rap, mostly deserved. But there's still one job they do better than anything else. How to send Google Sheets pies and donuts to Slack with Chartcastr, and the specific shape of data they're actually built for.
Pie & donut charts from Google Sheets to Slack, the only place we still recommend them
The conventional wisdom on pie charts is that they're bad. The conventional wisdom is mostly right. Most pie charts in the wild should be bar charts. People are bad at comparing angles. They're good at comparing lengths.
But there's a small slice of cases, pun reluctantly accepted, where a pie or donut is the right call. Chartcastr renders both. Here's the rule we use internally for when to ship one.
The case for the pie
Pies do exactly one thing well: they communicate that a few categories make up a whole, and roughly what proportion each contributes. That's it.
Concretely, a pie works when:
- You have 2 to 5 categories. Not more.
- The categories sum to a meaningful 100%. Revenue by channel, traffic by source, support tickets by category.
- You want the reader to grasp the share, not the absolute number.
- One slice dominates and you want that dominance to be the message. "60% of our revenue comes from one channel" reads loud and clear in a pie. The same fact in a bar chart looks like "this bar is taller than the others".
If your data fails any of those, use bars.
The case for the donut
Donuts are pies with the centre removed. The hole is sometimes used for a total, "$1.4M total spend", which is genuinely useful in a Slack report because it answers "of how much" in the same glance as the breakdown.
If you're going to ship a pie, a donut with the total in the middle is usually the better version.
The Slack pattern
The shape that works in Slack is small, occasional, and paired with a number:
- A weekly
#financepost with a donut showing revenue by product line and total revenue in the middle. - A monthly
#supportpost with a donut showing tickets by category and total tickets in the middle. - A campaign-end summary with a pie of conversions by channel.
It's not a chart you check daily. It's a chart you check when you need a sanity check on the mix.
What we render
Pie and donut charts both come through Chartcastr cleanly. We preserve the colour palette you set in Sheets, the legend position, and donut hole sizing. If you've labelled slices with values or percentages in Sheets, those labels come through.
Where it goes wrong
The places we see pies break down:
- Too many slices. Anything past 5 becomes a colour wheel. Group small categories into "Other".
- Slices of similar size. If three slices are all "about a third", a pie chart is no faster to read than a sentence saying so.
- No clear story. A pie that exists to be a pie is just decoration.
Setting it up
- Build the pie or donut in Google Sheets, sum-to-100% data, 2 to 5 categories.
- In Chartcastr add the source and pick the chart.
- Pick a Slack channel or email destination.
- Schedule it weekly or monthly. Don't pie-chart things daily.
If you change the slice colours in Sheets, the next pulse uses the new colours. Full reference: Premade Charts docs.
If you find yourself wanting to compare proportions over time, slices changing, use a stacked bar chart instead. A pie shows one moment. A stacked bar shows how that moment changes.
