Compare and rank

Bump chart (rank changes)

Tracks changes in the rank of each item across ordered periods.

  • Chart
  • Ranking
  • Trend over time

What it answers

How did the ranking change?

Shape
Lines connecting rank positions
What it shows
Overtakes, stable leaders and rank volatility.

Example

The question

How did our top five products move in the sales ranking this year?

Sales rank of the top five products by quarter, 2025

Sales rank of the top five products by quarter, 2025Rank by quarter, 1 is highest. Atlas: #1, #1, #2, #3; Birch: #2, #3, #4, #5; Comet: #3, #2, #1, #1; Delta: #4, #5, #5, #4; Ember: #5, #4, #3, #2.Q1Q2Q3Q4AtlasAtlasBirchBirchCometCometDeltaDeltaEmberEmber
Sales rank of the top five products by quarter, 2025Rank by quarter, 1 is highest. Atlas: #1, #1, #2, #3; Birch: #2, #3, #4, #5; Comet: #3, #2, #1, #1; Delta: #4, #5, #5, #4; Ember: #5, #4, #3, #2.Q1Q2Q3Q4#1#2#3#4#5AtlasBirchCometDeltaEmber
Rank by quarter, 1 is highest. Atlas: #1, #1, #2, #3; Birch: #2, #3, #4, #5; Comet: #3, #2, #1, #1; Delta: #4, #5, #5, #4; Ember: #5, #4, #3, #2.

Illustrative data. It belongs to no company.

When to use it

Use it when

Tracking leaders across periods.

Avoid it when

When actual magnitude differences matter more than rank.

How to read it

Follow each item across rank positions; rank 1 is normally at the top.

Note

Rank 1 is normally at the top.

The data you need

Layout
Long table
Fields
Period: ordered date/number; Item: text; Value or Rank: number
One row is
One item in one period. What is the grain?
Rules
If deriving ranks, state ascending/descending order and tie rules. Explicitly distinguish rank from magnitude.
Optional
Optional precomputed Rank. Fill missing periods only under a stated rule.
Preparation
Compute ranks within each period using a declared ordering and tie rule; align item IDs across periods.
Units
Ordinal rank, normally 1=highest; not an interval or magnitude unit. State ties and eligible ranking population.
Example of the input layout: Long tableA long table: one row per observation, few columns. New data adds rows.PeriodItemValue or RaAABBC
A long table: one row per observation, few columns. New data adds rows.

Worked example

Data Jan: A30, B45; Feb: A50, B40. Derive descending ranks. A moves from rank 2 to rank 1.

PeriodItemValue
2026-01A30
2026-01B45
2026-02A50
2026-02B40