“People learn almost twice as well from words and pictures together as from words alone.” When your slide appears on stage, you notice half the room quietly reaching for their phones. Why did your visual just lose them—and what tiny changes would make it do the talking for you?
“People learn almost twice as well from words and pictures together as from words alone.” When your slide appears on stage, you notice half the room quietly reaching for their phones. Why did your visual just lose them—and what tiny changes would make it do the talking for you?
So far, we’ve focused on how your brain handles spoken language alongside visuals. Now we’ll zoom in on *what* belongs on that screen in the first place. Think of the moment your slide appears as opening a small window in your audience’s mind: too much clutter and nothing gets through; the right single image or phrase and the idea snaps into focus.
In this episode, we’ll turn vague advice like “don’t crowd your slides” into specific, testable choices: how many words, which colors, what kind of images, and when to use no slide at all.
Your slides don’t exist in isolation; they compete with the demo on screen, the Slack pings in people’s pockets, and the tiny notification light on a laptop lid. In a technical talk, this competition gets even fiercer: diagrams, logs, and code are all “visuals,” but they don’t all help the same way. Some clarify; others become visual noise that quietly sabotages your main point. The opportunity is to treat each slide like a carefully chosen scene in a film: you decide where the audience should look, how long, and what they should feel is important enough to remember once the slide disappears.
Start by deciding what *job* each visual has. Before you open your slide tool, ask: “When this appears, what decision, feeling, or realization should it trigger?” If you can’t answer in a single sentence, you don’t have a slide yet—you have a vague intention.
Different slide “jobs” demand different visuals:
- To **surface a problem**, use contrast. A simple before/after graph, a failed error budget, or two short log snippets—one normal, one catastrophic—makes the gap obvious without you narrating every detail. - To **show structure**, think maps: architecture diagrams that hide nonessential components, user flows that show only primary paths, or layered diagrams that reveal complexity step by step as you speak. - To **reveal change over time**, favor sequences: a series of snapshots of the same dashboard, a timeline of incidents, a version history of an API. Several ultra-simple slides often beat one overloaded “everything” chart. - To **make a choice feel concrete**, place options side by side with just the deciding factors visible: latency vs. cost, maintainability vs. raw performance, or on-prem vs. managed service.
In technical talks, code and data are tempting visual crutches. Treat them instead like close-up shots: brief and intentional. For code, show the smallest snippet that proves your point—often 5–10 lines with irrelevant parts faded or removed. For data, shrink the dataset, not the font. If someone in the back row can’t see the key number, the slide is doing theater, not teaching.
Color and motion should guide attention, not decorate. Use one accent color to say “look here,” and let everything else stay quiet. When highlighting, change *one* thing at a time—bold one line, circle one node, animate one transition—so the audience’s eyes don’t have to choose.
Sometimes the most powerful visual is none at all. If you’re telling a critical incident story, making a bold claim, or asking for a decision, try a blank screen. The sudden absence of visuals throws all focus onto your words and voice, creating a deliberate pause in the visual stream rather than yet another slide to ignore.
A simple way to test your visuals is to strip them down until they almost feel “too empty,” then add back only what your story truly needs. For example, take a dense architecture slide from a past talk. First, duplicate it. On the copy, remove every label that isn’t absolutely necessary to follow your spoken explanation. Then, collapse any repeated components into a single icon with a short, neutral label. Finally, introduce a single accent cue—a colored outline, arrow, or spotlight—to mark the element you’re currently discussing, and keep everything else visually quieter.
You can do the same with data. Start with a busy dashboard screenshot. Replace it with a hand-built chart that shows only one relationship you care about, at the time scale that matters. If your narrative depends on multiple relationships, break them into a small sequence of slides, advancing as your story shifts, instead of forcing the audience to decode everything at once. Treat the original, complex view as your private reference; the audience only sees the distilled version that fits the moment you’re guiding them through.
As visuals get smarter, expect them to behave less like static posters and more like collaborative teammates. Slides could adapt mid-talk: zooming into a graph when faces look confused, or swapping a bar chart for a map when questions shift to geography. Think of it like a jazz trio: you set the theme, and your visuals improvise along with the room. This will demand new skills—curating trustworthy data sources, setting ethical guardrails, and rehearsing branches in your narrative, not just one linear deck.
Treat this like ongoing prototyping, not a one‑time deck. After each talk, grab a teammate and replay a recording with the sound off: could they still trace the story arc from the screens alone? Mark every moment where the “plot” breaks. Those gaps tell you what to redesign next, turning your slides into a living system that improves with each release.
Before next week, ask yourself: Where in my next presentation am I currently planning to use a slide that’s basically a script, and how could I turn that into a single powerful image or simple diagram instead? Looking at one existing slide deck, which 2–3 cluttered slides could I simplify today by removing at least half the text while still keeping the message clear? When I imagine my audience seeing each visual for the first time, what do I actually want them to notice or feel in the first three seconds—and does each slide, chart, or image currently trigger that specific reaction?
