A number is a claim. Give it a comparison, one sentence of meaning, a motion that points at the change, and a visible source.
“Engagement up 40%.” Up from what? Compared with when? Measured how? A number on its own is not information; it is a claim waiting for context. The same is true of the chart under it, which most people glance at and scroll past. Data becomes a story when someone decides what the reader should notice—and makes that one thing impossible to miss.
A number is a claim. Give it context.
Every figure needs three companions: a baseline, a period and a definition. “Enquiries rose from 25 to 40 a month between March and August, counting only messages that asked about a specific service” is a sentence someone can trust and check. “Enquiries up 60%” is not.
Context also protects you. A big percentage on a small base is easy to misread; stating the base makes the claim honest and still lets the improvement show.
One chart, one sentence.
Title the chart with the finding, not the subject. “Evening classes fill twice as fast as morning ones” tells the reader what to look for; “Class bookings by time of day” leaves them to work it out. If you cannot write the sentence, the chart is not ready.
Take an illustrative gym in Al Mouj deciding where to spend its next advertising month. A chart titled with the finding settles the meeting in a minute. The same data labelled generically starts a debate.
If the chart needs a paragraph to explain it, the chart is the paragraph’s problem.
Animate the change, not the chart.
Motion in data graphics has one legitimate job: pointing at the difference. Bars that grow into place, a number that counts up to its value, a line that draws itself to the point that matters. Everything else—spinning pies, bouncing labels, decorative particles—competes with the finding.
Keep it short, play it once, and make sure the final frame works as a still image. Many people will see it paused, in a feed, or with motion reduced.
Design for the thumb.
Most reports and posts are read on a phone. Use a vertical or square frame, one chart per screen, numbers large enough to read without zooming, and labels placed on the data rather than in a legend. Three colours are plenty: one for the thing that matters, one for everything else, one for the baseline.
In Arabic versions, decide once whether numbers use Western digits or Arabic-Indic digits, and keep it consistent across the report, the website and the social post. Mirror the layout; do not mirror the chart’s meaning.
Be honest with axes and sources.
A truncated axis can make a 3% change look like a cliff. A conveniently chosen date range can hide a bad quarter. Averages hide outliers. These are the shortcuts that make data look exciting and make it untrustworthy. Start axes at zero for bars, show the whole period, and say when a figure is an estimate.
Put the source and the date on the graphic itself, not in a footnote nobody sees. If a number came from a platform’s own dashboard, say which one. Never publish a figure you could not reproduce if a client asked.
- Bar charts start at zero; line charts show the full period.
- Label estimates and projections as estimates and projections.
- Source and date on every graphic, in both languages.
- Keep the spreadsheet behind every published number.
Turn reports into stories.
A monthly report should open with three findings written as sentences, each with its chart, before any tables. The same three findings become the agenda for the review meeting and, with confidential details removed, a “what we learned this month” post that shows the thinking behind the work.
Over a year, those posts become the most credible content a business can publish: real decisions, real numbers, honestly shown.
FROM THINKING TO DOING
Let’s make your data worth reading.
We design reports and data graphics that make the finding obvious, in both languages, with sources you can stand behind.
