Home
What does saliency actually mean?
Saliency comes from visual neuroscience. When you look at anything — a room, a page, a feed — your eyes don't sweep across it evenly like a scanner. Attention snaps first to whatever the visual system flags as important: high contrast, motion, faces, distinctive shapes, certain positions. All of this happens in roughly the first two to three seconds, before you've consciously decided to look at anything. Researchers call it pre-attentive processing.
The practical upshot is humbling. What people notice is not decided by what's important. It's decided by what's salient. Those two things overlap only when someone designs them to.
Why are humans wired this way?
Because we were visual communicators long before we were readers.
Seventeen thousand years before the alphabet, people painted animals on the walls at Lascaux — and the images still work on us today, no translation required. Egyptian hieroglyphics fused picture and meaning. Medieval Europe, mostly unable to read, learned its theology from illustrated manuscripts and the Biblia Pauperum, the "poor man's Bible" told in pictures. Even now, a handful of universal symbols — the arrow, the cross, the hand — cross every border on Earth without a word.
Then Gutenberg's press made text cheap, and for five centuries the written word dominated mass communication. In Boxcar's keynote framing, that era is starting to look like the exception rather than the rule. Feeds, short-form video, image-first platforms and now AI-generated media are pulling communication back toward its visual default. Sight was always the primary channel. The technology finally caught up to the biology.
What makes something salient?
The rules are consistent enough to be modeled. The strongest drivers:
- Contrast. The eye goes to the sharpest difference in brightness or color, not the biggest element.
- Faces. Humans fixate on faces almost involuntarily — and then follow where those faces are looking.
- Motion. Anything that moves outranks nearly everything that doesn't.
- Size and position. Larger elements and certain screen zones get first consideration; cluttered corners get none.
- Distinctiveness. One element that breaks the pattern beats five that follow it.
None of this is subjective. That's the surprising part. Saliency behaves less like taste and more like physics.
Why does saliency matter more now than ever?
Because the volume of competing visual information has exploded while the human attention window hasn't budged.
Consider the audience growing up inside this environment. Pew Research Center's December 2025 survey found roughly one in five U.S. teens are on TikTok and YouTube almost constantly, and 36% are on at least one major platform almost constantly. Every feed they scroll was engineered — by some of the best-resourced attention science in history — to win the first half-second of a glance. Boxcar calls this hacked attention: the platforms hacked it first, and now every message, brand, institution and idea competes inside the environment they built.
In that environment, being technically visible means nothing. Plenty of content is displayed. Very little is seen.
Can saliency really be measured?
Yes, and this is what turned saliency from an academic idea into a working discipline.
For decades, measuring attention meant live eye-tracking: participants, lab hardware, weeks of study. Today, predictive attention models — AI trained on large datasets of real human eye-tracking — can forecast where eyes will land in the first seconds of viewing an image, a page or a video frame. The output is an attention heatmap: a picture of what will get seen, produced before anything goes live.
The predictions cover the mechanical, pre-conscious stage of viewing — which happens to be the stage where most communication succeeds or fails. What a heatmap can't tell you is whether people will like what they see, believe it or act on it. Saliency governs noticing. Everything after noticing is still craft.
What does this mean if you communicate for a living?
It means the first question about any visual — an ad, a page, a slide, a sign, a video frame — is no longer "is it good?" but "what gets seen first?" Those are different questions with different answers, and the second one now has evidence behind it.
At Boxcar, saliency sits at the center of how we think about visual communication: creative gets tested for attention before it gets debated for taste. Later in this series we'll get concrete about what that looks like for ads and landing pages. For now, the idea stands on its own — and once you have it, you can't unsee it.
What to do now
A few ways to start working with saliency this week, no tools required:
- Try the two-second test. Show a colleague any visual you've published for two seconds, then take it away and ask what they remember. What they name is what's salient. Notice how often it isn't the point.
- Find your accidental focal points. Most layouts have one — a decorative element, a background sky, a logo — quietly outcompeting the message. You only need to look once to start spotting them everywhere.
- Count your competing elements. If four things are fighting to be seen first, the honest answer is that none of them will be.
- When you're ready for evidence, run an attention heatmap on something real. Seeing your own work through a predicted gaze is usually the moment this stops being theory.
