Start with the question, then choose the chart
Before opening the chart menu, write the question the audience needs to answer. “Which workstream needs the most funding?” calls for a comparison. “When does spending peak?” calls for a view over time. Both might use the same project workbook, but they need different views of it.
Specify the measure as well as the question. Revenue, margin, order count and average order value describe different aspects of a business. A chart of one cannot quietly stand in for another. Check that the available data can answer the question before deciding how to display it.
The Financial Times Visual Vocabulary groups charts by relationships such as ranking, change over time, distribution and correlation. That is a useful way to approach a presentation: identify the relationship you need to explain, then choose a visual that makes it visible.
Use this chart-choice table as a starting point
The table below applies common chart families to questions that arise in business presentations. These are starting points to test against your actual data and audience, rather than a rule that each question has only one correct answer.
The UK Government Analysis Function’s guidance similarly links chart selection to the message and statistical relationship.
| Reader question | Start with | Check first |
|---|---|---|
| Which category is larger? | Bar chart; horizontal bars help with long category names | Comparable units, period and category definitions |
| How has the measure changed over time? | Line chart; columns can suit individual period totals | Time intervals, missing periods and changes in measurement |
| How is a total divided? | Stacked bar; a pie or donut can suit a few clearly different shares | A defined whole, with categories that do not overlap |
| How much do results vary? | Histogram or dot plot; a box plot for an audience familiar with it | Enough observations and an explanation of grouping |
| Are two measures associated? | Scatterplot | Paired observations; association alone does not establish cause |
| What are the exact figures? | A short table, with a chart if the pattern also matters | Readable precision, units and alignment |
Source: UK Government Analysis Function — Data visualisation: charts
Work through the decision with a project-budget example
Imagine a fictional project review with a workbook containing budgets by workstream, expected monthly expenditure and revised estimates. This is an illustrative planning exercise, not a Pitchwave or client result.
If the decision is which workstream needs the largest allocation, compare workstreams on a shared scale. Horizontal bars let the team compare amounts while reading the full workstream names. Include the currency and the period the budget covers.
If the decision is when the project needs cash, show expenditure in chronological order. Monthly columns can make individual period amounts clear; a line may help readers follow the overall pattern. A workstream ranking would not answer this timing question.
If the decision is where a revised estimate differs from the approved budget, compare those two amounts for each workstream or show their difference. Keep the approved and revised status explicit. A percentage-of-total chart could show the budget mix while concealing the size of a shortfall.
The useful edit happens before styling: decide which of these questions the slide will answer. Move the remaining views to separate slides or the appendix when they are needed for review.
Finish this sentence: “The reader needs to compare…” Name the categories, periods or measures. If the chart makes a different comparison easiest, reconsider the chart.
Keep bar lengths meaningful
When bars represent amounts, use a zero baseline so their lengths show the relative sizes faithfully. Starting the value axis near the smallest amount can make a modest difference look much larger. The Government Analysis Function’s bar-chart module explains this risk and recommends starting bars at zero.
Order categories to support the comparison. Ranking by value often helps when the question is which is largest. Retain a natural sequence when the categories have one, such as stages in a process. Make the ordering deliberate.
If two charts will be compared side by side, check their scales together. The tallest bar in each panel should not appear to represent the same amount when the underlying values differ. Use a shared scale for direct size comparisons, or clearly explain why a different scale is necessary.
Check what a line connects
A line draws attention to movement between observations. Use time labels and spacing that represent the actual intervals. Identify missing observations or a change in measurement, and avoid implying that an unobserved value was measured.
The Government Analysis Function’s line-chart module discusses time series, breaks in the data and the risks of dual axes. For a presentation, start by asking whether separate, clearly labelled charts would explain two measures more reliably than overlaying them on different scales.
A line chart does not always need a zero baseline. Choose a range that lets readers assess the variation, keep the scale clear and make any truncation apparent. Check whether the chart’s proportions make the change look unusually steep or unusually flat.
When several lines become difficult to follow, try separate panels with consistent scales where appropriate. Confirm that the labels are still readable at slide size; a collection of tiny charts may simply move the problem elsewhere.
Give the reader the context needed to interpret it
A chart should identify the measure, unit, population or business scope, and relevant period. For shares, define the total. For averages, explain what was averaged when the distinction affects the decision. Keep the source and any material exclusions close enough to find.
Use the slide headline to explain the supported point, while labels and a subtitle define the data. An intended finding such as “Installation accounts for the largest budget allocation” still needs a chart that supports that statement. Check it against the actual figures before using it.
Keep observed results and forecasts distinguishable. Proposed budgets and approved budgets should not acquire the same status simply because they use the same colour. Add clear labels and an explanation of any assumptions relevant to the comparison.
If the audience must read exact values, provide a small, legible table or the relevant figures alongside the chart. If the detailed workbook is needed for diligence, supply it separately rather than shrinking it into an unreadable slide.
Review the chart in the format you will send
Check the exported presentation as well as the editable source. Open the PDF at a normal reading size and on a phone if people are likely to review it there. Read the labels, source note and annotations without relying on zoom to rescue the main message.
Ask someone unfamiliar with the workbook what they think the chart shows. If they infer a different comparison or miss an important condition, revise the chart or its explanation. Treat their answer as a useful review of this slide, not as proof that one chart type works for every audience.
When asking for presentation design support, share the underlying data and the decision each chart needs to support. Include definitions and known limitations. That gives the designer enough context to question the visual choice as well as improve its appearance.
- QuestionIs the comparison or pattern clear?
- FitDoes the chart show the relationship the reader needs?
- BasisAre units, scope, period and definitions visible?
- IntegrityAre scales, totals, gaps and assumptions represented fairly?
- ReadabilityCan the reader follow it in the final presentation format?