luvv to helpDiscover the Best Free Online Tools
Topic 3 of 13

Working with Axes and Scales

Learn Working with Axes and Scales for free with explanations, exercises, and a quick test (for Data Analyst).

Published: December 20, 2025 | Updated: December 20, 2025

Why this matters

Axes and scales turn raw numbers into shapes you can compare. As a Data Analyst, you will:

  • Show trends over time (time axis + appropriate y-scale)
  • Compare categories fairly (zero baselines, consistent units)
  • Communicate growth patterns (log scale vs linear)
  • Prevent misinterpretation (avoid misleading truncation and dual-axis confusion)

Good axes make insights obvious. Bad axes create doubt or, worse, wrong decisions.

Concept explained simply

An axis is a ruler. A scale is the rule for how numbers map to positions on that ruler.

  • Linear: equal steps in data = equal spacing on chart. Best default for most numeric data.
  • Logarithmic (log): equal ratios (x2, x10) = equal spacing. Useful when values span orders of magnitude or growth is multiplicative.
  • Time/continuous: treats dates/times as a continuous line (not categories). Required for time series.
  • Ordinal/categorical: ordered categories (A, B, C) equally spaced.
  • Diverging: centered around a meaningful midpoint (e.g., 0, target), used for positive/negative deviations.
  • Percentage/Proportion: 0–100% scale or 0–1 fractions.

Mental model

Pick the scale that preserves the comparison you care about:

  • Absolute difference → linear scale with appropriate baseline
  • Relative growth or ratios → log scale or index-to-100
  • Part-to-whole → percentage scale
  • Change around a reference → diverging scale centered at reference
Quick check: linear vs log
  • If doubling matters more than absolute difference, consider log.
  • If absolute gaps matter (e.g., budget differences), use linear.

Practical rules and defaults

  • Bars must start at zero on the value axis. Otherwise heights mislead.
  • Lines can start above zero if you focus on variation; clearly mark the axis range.
  • Time series: use a continuous time axis; avoid treating months as categories unless values are sparse.
  • Log scale: only for positive values; add small offsets only if justified and explained.
  • Axis range: avoid tight cropping that exaggerates changes; avoid overly wide ranges that flatten signals.
  • Ticks: aim for 4–7 major ticks; use minor gridlines sparingly.
  • Units: show once in axis label or tick format (e.g., $k, $M, %, per 100k). Be consistent across charts.
  • Dual axes: risky. Prefer small multiples or normalize (index at 100) to compare different units.
  • Labels: rotate or wrap categorical labels; consider sorting categories to reduce clutter.
  • Rounding: use human-friendly tick marks (e.g., 0, 50, 100; 0, 1k, 2k).
When to consider diverging scales

Use when values can be above/below a meaningful reference (e.g., budget variance vs 0). Center the axis at that reference.

Worked examples

Example 1: Monthly revenue line chart

Task: Show monthly revenue for the last 24 months.

  • X-axis: continuous time (months).
  • Y-axis: linear, currency.
  • Baseline: not required to start at 0 for a line; set sensible min/max so the trend is visible without exaggeration.
  • Ticks: 4–6 on Y; monthly ticks on X with quarterly labels.
  • Format: $k or $M. Label the axis as "Revenue ($)".
Why not log?

If revenue ranges within a small factor (e.g., 1.2x–2x), log adds little benefit. Linear maintains intuitive differences.

Example 2: Exponential user growth

Task: Compare user counts across products where one grows 100x faster.

  • Y-axis: log scale (base 10). All values must be > 0.
  • Interpretation: equal slopes indicate equal percentage growth rates.
  • Tip: add tick labels at 100, 1k, 10k, 100k; annotate doubling time.
Sanity check

On log scale, the distance from 1k to 10k equals 10k to 100k. If that feels surprising, that’s expected: you’re comparing ratios, not differences.

Example 3: Team budgets comparison (bars)

Task: Compare annual budgets by team.

  • Chart: bars with Y-axis linear starting at 0.
  • Sort: descending by budget to improve readability.
  • Units: show $M on axis; no symbols in labels to reduce clutter.
  • Ticks: 0, 2, 4, 6, 8 (example).
Common pitfall avoided

Truncating the Y-axis above 0 would make small differences look huge. Keep bars grounded at zero.

Example 4: Fair comparison across different-sized regions

Task: Compare incident counts across regions with different populations.

  • Transform: convert counts to rate per 100k people.
  • Chart: bars starting at 0, or a line over time using rate.
  • Axis label: "Incidents per 100k".
Why transform?

Raw counts reward larger populations. Rates make risk comparable.

Example 5: Avoiding dual axes

Task: Show marketing spend (USD) and signups (count) together.

  • Preferred: small multiples (two aligned charts sharing the same X time axis).
  • Alternative: normalize both to index=100 at start date and plot on one axis.
  • Avoid: dual-axis with separate scales; readers may misread magnitude and correlation.

Who this is for

  • Data Analysts and aspiring analysts
  • Anyone building charts for reports, dashboards, or presentations
  • Researchers and PMs who need clear comparisons

Prerequisites

  • Basic chart types (bar, line, scatter)
  • Comfort with numeric data and percentages
  • Familiarity with a charting tool (Excel, Google Sheets, Tableau, Power BI, Python/R is optional)

Learning path

  • Start: Axes and Scales (this lesson)
  • Next: Choosing the right chart for your question
  • Then: Labeling, annotations, and color for emphasis
  • Finally: Dashboard layout and storytelling with data

Hands-on exercises

Do these in your preferred tool. Mirror the expected outputs as closely as you can.

Exercise 1 — Linear or log?

You have product A with monthly active users: 120, 180, 270, 405, 610, 915, 1370. Decide on the Y-axis scale and explain why. Create a line chart across months with appropriate ticks and labels.

  • Deliverable: a line chart with chosen scale and a one-sentence justification in a text box or caption.
Hint
  • Are ratios (x1.5) or absolute differences more meaningful?
  • Do values span an order of magnitude?

Exercise 2 — Bar chart with fair baseline

Compare marketing spend (in thousands of USD) across four channels: Search 180, Social 120, Affiliates 90, Email 35.

  • Deliverable: a bar chart with Y-axis starting at 0, tick labels in $k, sorted descending, and a clear axis label.
Hint
  • Bars must start at zero. Keep 4–7 ticks. Consider 0, 50k, 100k, 150k, 200k.

Self-check checklist

  • Ticks are readable (4–7 major ticks)
  • Units are consistent and shown once (axis label or tick format)
  • No negative or zero values on log scales
  • Bars begin at zero baseline
  • Time is continuous on time-series charts
  • Ranges are sensible (not excessively cropped or overly wide)

Common mistakes and how to self-check

  • Truncated bar axes: if using bars, confirm Y starts at 0.
  • Using log on zero/negative values: check your minimum is > 0.
  • Overcrowded ticks: if more than ~7 major ticks, reduce or format smarter.
  • Inconsistent units across charts: standardize to $k/$M, %, or per 100k.
  • Dual-axis confusion: if you used two Y-axes, try small multiples or index-to-100 instead.
  • Unlabeled transformations: if you normalized or used rates, say so in the axis label or subtitle.
Quick audit
  1. Read only the axes and ticks — would the audience still know what the chart is about?
  2. Hide the data: does the scale choice still match the story (absolute vs relative)?
  3. Show to a colleague for 10 seconds — can they summarize correctly?

Practical projects

  • Dashboard makeover: take a messy dashboard and fix axis baselines, units, and tick density. Add before/after snapshots.
  • Growth report: compare 3 products on a log-scale time series, annotate doubling times, and provide a one-paragraph interpretation.
  • Fair regional comparison: convert raw counts to per 100k for 6 regions, visualize with bars, and write a 3-bullet summary of insights.

Mini challenge

You have two charts to show churn (% of customers who leave each month) and active users (counts). What’s a clear setup?

Show answer
  • Use small multiples: top chart churn (%) with Y from 0–20%; bottom chart active users (linear or log if spanning large range).
  • Shared time axis; distinct Y labels with units.

Next steps

  • Practice setting tick marks and labels that match audience literacy.
  • Experiment with diverging scales for variance around targets.
  • Try index-to-100 to compare different series on one axis safely.
  • Combine with annotations to call out thresholds, events, or targets.

Quick test

Available to everyone. Sign in to save your progress and resume later.

Practice Exercises

2 exercises to complete

Instructions

You have product A with monthly active users: 120, 180, 270, 405, 610, 915, 1370. Decide on the Y-axis scale and explain why. Create a line chart across months with appropriate ticks and labels.

  • Deliverable: a line chart with chosen scale and a one-sentence justification.
Expected Output
A time series chart with a log Y-axis (base 10) OR a linear axis with a clear rationale. If log: ticks at 100, 200, 500, 1k; caption explains multiplicative growth. If linear: ticks in 0–1.5k with reasoning about audience familiarity.

Working with Axes and Scales — Quick Test

Test your knowledge with 8 questions. Pass with 70% or higher.

8 questions70% to pass

Have questions about Working with Axes and Scales?

AI Assistant

Ask questions about this tool