Intelligence reports live or die by their ability to communicate complex information
quickly. Visuals -- charts, maps, timelines, link diagrams, and infographics -- are
not decorations. They are analytic tools that can convey patterns, relationships, and
trends far more efficiently than text alone. This lesson covers when visuals beat text,
how to choose the right visual type, fundamental design principles, the most common
visual sins, and how to integrate graphics into your narrative without turning your
report into a picture book.
Not everything needs a chart. But many things need one desperately and do not get one,
because the analyst either did not think of it or was afraid of PowerPoint. Here is
when a visual is not just nice to have but essential:
The Visual Advantage
- Showing relationships and networks. Describing a 15-person
organizational network in text requires paragraphs of "A reports to B, who
coordinates with C." A link diagram shows the same thing in seconds.
- Revealing patterns over time. "Attacks have increased by 40% over
the past six months" is a sentence. A trend line showing the weekly attack count with
annotated events is a story.
- Geographic context. "The facility is located 12 kilometers northwest
of the capital, adjacent to the main highway and 3 kilometers from the airfield" is
a sentence that begs for a map.
- Comparing quantities. Any time you are asking the reader to compare
more than three numbers, a chart will be more effective than text.
- Timelines of complex events. Chronologies with multiple overlapping
sequences are almost impossible to follow in paragraph form.
When Text Wins
Visuals are not always the answer. Nuanced analytic judgments,
caveats, and conditional reasoning ("if X, then
likely Y, unless Z") are almost always better expressed in text. A chart can show what
happened; text explains why it matters and what might happen next.
If your chart requires a 200-word caption to explain what it means, you have not
created a visual aid -- you have created a visual obstacle.
The single most common visual mistake is using the wrong chart type for the data. A
pie chart showing trends over time, a bar chart with 47 categories, a 3D exploding
donut chart for any reason -- these are crimes against information.
A Quick Reference Guide
- Bar chart: Comparing quantities across categories. "Attacks by
province," "Funding by source," "Weapons seized by type."
- Line chart: Showing trends over time. "Monthly incident count,"
"GDP growth trajectory," "Force strength over the past decade."
- Map: Any data with a geographic dimension. Incident locations,
territorial control, infrastructure, routes of movement.
- Link/network diagram: Relationships between people, organizations,
or entities. Command structures, communication networks, financial flows.
- Timeline: Sequence of events, especially when multiple parallel
threads need to be tracked. Operations, political developments, intelligence
collection milestones.
- Table: When the reader needs to look up specific values or compare
across multiple dimensions simultaneously.
- Pie chart: Parts of a whole, but only when there are 5 or fewer
categories. If your pie has 12 slices, use a bar chart instead.
When in doubt, use a bar chart. It is the most versatile, most readable, and hardest
to misinterpret chart type. Horizontal bars are particularly useful in intelligence
reporting because they accommodate long category labels (country names, group names)
without rotating text into illegibility.
Design Principles That Matter
You do not need to be a graphic designer, but you do need to follow a few fundamental
principles that separate professional intelligence graphics from amateur hour.
- Data-to-ink ratio: Every drop of ink (or pixel) should convey
information. Remove gridlines, borders, backgrounds, and decorative elements that
do not help the reader understand the data. Edward Tufte called unnecessary visual
elements "chartjunk" -- and he was right.
- Color with purpose: Use color to encode meaning, not to decorate.
Red for threats, green for friendly forces, amber for caution -- these conventions
exist for a reason. Limit your palette to 5-7 colors maximum. If you need more,
your chart is too complex.
- Typography: Use a clean, sans-serif font. Label axes and data
points directly on the chart rather than forcing the reader to cross-reference a
legend. If the reader has to look away from the data to understand it, you have
added friction.
- Annotation: Add brief text annotations to highlight key data points
or inflection moments. "Ceasefire announced" on a trend line is worth a paragraph
in the body text.
The Seven Deadly Visual Sins
- Misleading scales: Starting a y-axis at a non-zero value to
exaggerate trends. A 2% increase looks like a 200% increase if you truncate the
axis.
- 3D charts: They add visual noise, distort proportions, and make
accurate comparison impossible. There is no analytical reason to use them. Ever.
- Dual y-axes: Two different scales on the same chart invite
misinterpretation. If you must compare two variables with different units, use two
charts side by side.
- Rainbow color schemes: Using the full spectrum makes it impossible
for colorblind readers (8% of men) to distinguish categories and creates visual
chaos for everyone else.
- Overcrowded charts: If your scatter plot has 10,000 overlapping
points, you need a density map, not more data points.
- Missing context: A chart without a title, axis labels, source
attribution, and date range is not a chart -- it is a Rorschach test.
- Decorative illustrations: Clip art, stock photos, and decorative
borders do not make data more compelling. They make your report look like a middle
school science project.
If your pie chart looks like a beach ball and your bar chart has drop shadows, you
have not enhanced your analysis -- you have auditioned for a job in marketing.
Visuals as Part of the Story
A visual should never appear in your report without being referenced in the text. The
text introduces the visual, the visual provides the evidence, and the text interprets
what the reader should take away. This three-part integration ensures that visuals
enhance rather than interrupt the analytic narrative.
- Introduce before showing: "Figure 3 shows the sharp increase in
border crossings following the policy change" tells the reader what to look for.
- Interpret after showing: "As Figure 3 illustrates, the 60% increase
occurred within two weeks of the announcement, suggesting the policy change was the
primary driver" tells the reader what to conclude.
- Place visuals near the relevant text. A chart on page 12 that is
referenced on page 4 is a chart that will not be seen.
Accessibility Matters
Intelligence reports are read by people with varying visual abilities, on screens of
varying quality, and sometimes printed in black and white. Design for the worst case:
- Do not rely on color alone to convey meaning. Use patterns, labels,
or shapes as secondary encoding. If your chart is meaningless in grayscale, fix it.
- Use sufficient contrast. Light gray text on a white background is
invisible on a projected slide. Test your visuals in the environments where they
will actually be viewed.
- Include alt text for digital reports. Not just "Figure 4" but
"Bar chart showing top five threat groups by number of attacks in 2025, with Group A
leading at 47 incidents."
- Provide data tables alongside complex charts when the report will
be distributed digitally. Screen readers cannot parse a JPEG of a chart.
Before finalizing any visual, show it to someone unfamiliar with your analysis and
ask them what they see. If their takeaway does not match your intended message, the
chart needs work -- no matter how accurate the underlying data. A chart that is
technically correct but routinely misread is a liability, not an asset.
