# The best chart for a relationship

> Seventeen charts, and when each one fits.

**Short answer.** To see how two measures relate, use a scatter plot. To turn it into decisions, split it into a quadrant. For how every pair of measures moves together, use a correlation triangle. For links between items, use a network; for three trade-offs, a ternary.

Updated Nov 2, 2026 · E-commerce sample data

## 034 · Scatter

Spend against revenue: Partner · co-op returns the most per dollar.

Spend and revenue by campaign, Q3 2026. Twelve campaigns, spend against revenue (correlation 0.77); Partner · co-op returns the most per dollar: $184K on $40K.

| Campaign | Spend | Revenue |
| --- | --- | --- |
| Search · brand | $34K | $139.4K |
| Search · generic | $40K | $164K |
| Social · video | $40K | $112K |
| Social · carousel | $34K | $74.8K |
| Email · weekly | $8K | $25.6K |
| Email · win-back | $46K | $156.4K |
| Display · retarget | $12K | $16.8K |
| Display · prospect | $46K | $59.8K |
| Affiliate · deals | $34K | $108.8K |
| Affiliate · reviews | $18K | $61.2K |
| Podcast | $12K | $33.6K |
| Partner · co-op | $40K | $184K |

## 072 · Quadrant

The same campaigns by growth and margin, in four zones.

Campaigns by growth and margin, Q3 2026. Twelve campaigns in four zones split at 0% growth and 30% margin: grow 6, protect 2, fix margin 3, review 1.

| Campaign | Growth | Margin | Zone |
| --- | --- | --- | --- |
| Search · brand | 28% | 47% | Grow |
| Search · generic | −2% | 31% | Protect |
| Social · video | −6% | 26% | Review |
| Social · carousel | 35% | 22% | Fix margin |
| Email · weekly | 28% | 42% | Grow |
| Email · win-back | 35% | 18% | Fix margin |
| Display · retarget | 4% | 31% | Grow |
| Display · prospect | 14% | 47% | Grow |
| Affiliate · deals | 35% | 31% | Grow |
| Affiliate · reviews | −12% | 47% | Protect |
| Podcast | 21% | 35% | Grow |
| Partner · co-op | 9% | 22% | Fix margin |

## 085 · Correlation triangle

Every pair of measures at once: spend and revenue move together (0.77).

How the campaign measures move together. Correlation of every pair of spend, revenue, margin and growth: spend and revenue 0.77, spend and margin −0.36, spend and growth 0.08, revenue and margin −0.44, revenue and growth 0.01, margin and growth −0.19.

| Pair | Correlation |
| --- | --- |
| Spend · revenue | 0.77 |
| Spend · margin | −0.36 |
| Spend · growth | 0.08 |
| Revenue · margin | −0.44 |
| Revenue · growth | 0.01 |
| Margin · growth | −0.19 |

**Same 12 campaigns.** The scatter shows that more spend brings more revenue (correlation 0.77). The quadrant says what to do with each campaign, and the triangle shows which measures move together and which do not (margin and growth: -0.19).

## If you ask

| Question | Use a |
| --- | --- |
| How do two measures relate? | 034 · Scatter |
| What should we do with each item? | 072 · Quadrant |
| Which measures move together? | 085 · Correlation triangle |
| What connects to what? | 039 · Network |
| How do items trade off three ways? | 095 · Ternary |
| How do many items compare on many measures? | 049 · Parallel coordinates |

## All seventeen

| Type | Use it when | Data shape |
| --- | --- | --- |
| [027 · Cells matrix](https://usechartory.com/charts/cells-matrix) | Clusters across two sets of categories | A cluster colour and dense or sparse per cell |
| [034 · Scatter](https://usechartory.com/charts/scatter) | Two measures per item, as dots | Two values, a size, a group and a name per item |
| [039 · Network](https://usechartory.com/charts/network) | One hub and what it connects to | A hub and its nodes, each with a value |
| [040 · Tree](https://usechartory.com/charts/tree) | A release, its epics and their tasks | Nodes with a parent and a status |
| [041 · Orbit](https://usechartory.com/charts/orbit) | Tiers around a centre, with how active each is | Tiers with a count and a rate |
| [049 · Parallel coordinates](https://usechartory.com/charts/parallel-coordinates) | Many items across many measures | Several measures per item |
| [064 · Arc diagram](https://usechartory.com/charts/arc-diagram) | Flows between groups, as arcs on one line | Amounts between pairs of groups |
| [072 · Quadrant](https://usechartory.com/charts/quadrant) | Items in four zones that tell you what to do | Two values per item, two thresholds |
| [085 · Correlation triangle](https://usechartory.com/charts/correlation-triangle) | How every pair moves together | A correlation per pair |
| [091 · Cluster map](https://usechartory.com/charts/cluster-map) | Similar items grouped into named clusters | A centre, a spread and a count per cluster |
| [095 · Ternary](https://usechartory.com/charts/ternary) | Items placed by three measures that trade off | Three scores per item |
| [109 · Frontier map](https://usechartory.com/charts/frontier-map) | Models on quality against cost, with the best edge | Quality, cost and spend per model |
| [112 · Threshold sweep](https://usechartory.com/charts/threshold-sweep) | Alerts per week at every threshold | Alerts per threshold |
| [118 · Coverage map](https://usechartory.com/charts/coverage-map) | Real traffic, and which parts the tests cover | A traffic share and a case count per cluster |
| [121 · Reach tree](https://usechartory.com/charts/reach-tree) | Everything one key can reach | A root, systems and access per link |
| [127 · Permission matrix](https://usechartory.com/charts/permission-matrix) | Roles against actions, used and unused | A grant state per role and action |
| [129 · Break-even](https://usechartory.com/charts/break-even) | When the next plan becomes cheaper | Cost per plan across usage |

## Common mistakes

- **Reading a link as a cause** Two measures can move together for a third reason.
- **Overlapping dots** Hundreds of dots hide each other. Use see-through dots or a hex grid.
- **Thresholds pulled from the air** Quadrant lines should be targets people agreed on.

## Every relationship chart, and the other 115 types.

In Figma and React, in four sizes, light and dark. Pay once, every update. [See the seventeen in Charts](https://usechartory.com/charts#relationship)

## Questions

### What does a correlation of 0.9 mean?

The two measures rise and fall together almost every time. Near 0 means no link; below 0, they move apart.

### Scatter or quadrant?

A scatter to see the link, a quadrant to act on it.

### How many items fit on a scatter?

Dozens with labels, hundreds without.

### Do these charts come in every size?

Yes. All of them come in XS, S, M and L, each laid out again for its size, in light and dark.

## Other questions

- [Change over time](https://usechartory.com/questions/change-over-time)
- [Compare items](https://usechartory.com/questions/compare-items)
- [Part of a whole](https://usechartory.com/questions/part-of-a-whole)
- [Distribution](https://usechartory.com/questions/distribution)
- [Flow between steps](https://usechartory.com/questions/flow-between-steps)
- [Schedule and status](https://usechartory.com/questions/schedule-and-status)
- [Space and place](https://usechartory.com/questions/space-and-place)

Not sure it is a relationship? Start from what the chart has to show. [Open the Chart Chooser](https://usechartory.com/chooser)