Editorial Chart Quality Checklist: Make Data Visuals Clear, Fair, and Accessible
A chart should answer one honest question
A chart can make data clear in seconds. It can also hide an important gap or make a small change look huge. Before choosing colors or shapes, decide what the reader should learn. Is the chart showing a trend, a share, a rank, or a link between two measures? Write the answer in one plain sentence. If the team cannot do that, the chart may not be ready. Data visuals are editorial content, not decoration. A useful chart helps a person understand a fact and judge its limits. It should not force the reader to solve a puzzle.
Check the source before drawing
Find out where each value came from. Record the maker of the dataset, the date, the unit, and how the data was collected. A survey of app users is not a survey of the whole world. A small sample may be useful but should not be sold as a global rule. Read the source notes and any changes made to the dataset. Do not copy a chart image from a blog when the raw values are available from the first source. If a number cannot be checked, leave it out or label the uncertainty. A sharp graphic cannot rescue a weak claim.
Choose a form that fits the data
Use a line chart for change over time when points follow a clear order. A bar chart can help compare groups. A table is often best when readers need exact figures. A pie chart may work for a few parts of one whole, but it becomes hard to read with many slices. Do not use a map only because countries are involved; a ranked list may answer the question faster. Keep the scale honest. A chart type should help the reader see the true pattern, not make the result look dramatic.
Keep the axis and unit clear
Make the scale easy to judge
A chart without units leaves the reader guessing. State whether the measure is people, visits, money, rate, or time. Use a visible date range. For bar charts, a zero baseline usually helps people judge size fairly. For line charts, a narrower range can sometimes reveal small changes, but the scale must be clear. Do not crop a scale merely to make a small gain look like a huge jump. If data from two groups uses different units, avoid putting it on one axis without a clear reason. A reader should not need to inspect the source file to know what the line means.
Use color as support, not the whole message
A red line and green line may look clear to some people but not to everyone. Add labels, patterns, or direct text so the chart works without color alone. Check contrast on a phone and in grayscale. A brand palette may look stylish while hiding thin labels. Pick a few colors with a clear purpose. Do not give every category a new shade if the colors do not carry meaning. For a large chart, highlight the one series the article discusses and keep the rest easy to compare. This lets the reader focus on the point without losing context.
Write a useful description
A chart needs more than alt text that says “bar graph.” Explain the main message in nearby text. For a complex chart, give a longer text description or a data table that people can reach. The W3C guidance on complex images shows ways to make charts and graphs easier to use with assistive tools. The description should name what the chart measures and the key pattern. Avoid a long list of every value when a linked table can show those values more clearly. Make sure the chart’s lesson is available to someone who cannot see the image.
Example: a small store’s monthly orders
Imagine a shop that tracks completed orders for six months. The editor wants to show whether demand rose. A simple line chart uses months on the bottom and completed orders on the side. The article states the change in words and notes a one-week sale that may have affected the trend. The editor checks the store’s export against the chart points. A second staff member opens the image on a phone and confirms that labels are readable. The chart does not claim that a new ad campaign caused the rise. It shows what happened and gives the reader context.
Show the limits near the visual
A chart should not hide an important caveat in a footnote far below. If data covers only one region, say so near the title or caption. If a month is missing, show the gap. Do not connect it with a smooth line that suggests a known value. If the measure changed halfway through the series, mark that point. When figures are estimated, use a clear note about the method. Readers can work with uncertainty when it is shown honestly. They cannot judge a chart that hides the difference between measured and guessed values.
Use labels that readers know
A good chart can still fail when labels use internal team names or short codes. Spell out terms when space allows. If a technical term is needed, explain it in the caption. For dates, use a form that makes sense to a global audience, such as a clear month name. If money is shown, name the currency. If values are rates, explain the base. Do not call a number “growth” if it is only a count of new sign-ups without a prior period. Check every title and label against the story. The words are part of the data.
Make a review checklist
What a second editor should check
Before release, ask a second person to check the source, values, unit, scale, labels, color, and main claim. Test the chart at its real page size. If the layout crops it, make a better version rather than forcing readers to zoom. Keep the source and the code or sheet used to make the chart. A future editor may need to update it. Do not replace a chart with a new one that uses a different method without explaining the change. Record the date of each update so readers know which view they see.
Do not turn a chart into a fake quote
A visual may be shared on social media without the article around it. It needs enough context to stand on its own. Do not add an expert’s photo or name beside a number if that person did not make or endorse the claim. If a source provides a useful quote, link the name to the original statement and keep it in context. Avoid making a statistic sound universal when the source describes a narrow group. A short honest caption may travel farther than a dramatic claim, but it also does less harm when someone shares it alone.
Build a chart habit the team can keep
Start with one template for common needs: trends, comparisons, and small tables. Add a source field and a review date. Keep a place for the chart’s main point and limits. Teach writers to ask whether the visual helps the reader act or decide. If a table would be simpler, use it. If a chart repeats the text without adding insight, leave it out. Good data design is not a contest for the most shapes. It is a quiet way to help people see what the evidence really says.
