Everything we've published, newest first.
We have the definition of curious. We asked 33 people for the picture, with 361 green and red dots across 24 designs. A raised eyebrow took 83% green. A woman grinning at the camera took 80% red. The winners share one structure: something in the frame is still being looked at, or something has been left for you to finish.
Read More
What does “jazz it up” actually mean? 25 people sorted 33 designs with 396 green and red dots. Dripping goo letters took 93% green. A blurred gothic masthead took 13 of 13 red. The winners share one recipe: loud color on a surface you could touch, caught in the middle of the action.
Read More
What does high quality orange juice look like? 50 people sorted 24 images of photography, labels and packaging, placing 535 green and red dots. Dew on cut fruit took 97%. A trendy can took zero. Packaging only wins when real fruit surrounds it. One variable explains every score: distance from the tree.
Read More
“Health” and “wellness” get used almost interchangeably — but do people picture the same thing? We asked 26 participants to sort 36 images between the two words with no definitions provided. The boundary they drew was strikingly consistent: health is what the body does; wellness is what it receives.
Read More
Every year, brands spend fortunes on nostalgia — the polar bears, the holiday trucks, the Santa on retainer — betting that feeling is what sells the drink. So we tested it. We asked 467 people what actually makes them crave a Coke, and the results were unkind to the marketing budget: not the beloved icons, but cold, wet, in-motion liquid — condensation on glass, a bottle at the lips. People don’t crave the brand. They crave the drink.
Read More
What makes design feel elevated? A Constellations visual study reveals the heritage codes, colors, textures, and forms behind elevated aesthetics.
Read More
The Aesthetic of “Edgy” and the Problem of Distressed Fonts What do people mean when something feels “edgy”? It’s a word that shows up in creative direction all the time, but like most aesthetic language, it points to a vibe more than a clear set of decisions. So we set out to define it. Using […]
Read More
We tested six dimensions of the film’s visual direction using Constellations, from character design to villain treatment to overall style, spanning every era of the franchise. Despite Breath of the Wild being the bestselling Zelda game by a wide margin, its characters, enemies, and aesthetics consistently split the audience. What participants converged on was Ocarina of Time…
Read More
Every designer has heard it: “Can you make it pop?” It’s one of the most common pieces of creative feedback, and one of the hardest to act on, because it points to a feeling more than a specific fix. So we decided to find out what “pop” actually means. Using Constellations, we asked creative professionals to sort images by which ones pop and which ones don’t—and the results were surprisingly consistent.
Read More
What makes a robot look helpful instead of creepy? 21 people placed green and red dots on 17 robot images. Industrial arms with no face scored 90.91% to 100% helpful. Humanoids in kitchens and living rooms scored 81.82% to 100% creepy, even while caring for the elderly. The one humanoid that won had a camera for a face.
Read More
The images people embraced all shared one thing: they caught Santa in the middle of doing something. Reading his list by candlelight. Kneeling to feed a reindeer in falling snow. The images they rejected? Cute graphics, staged photos, and counterintuitively, some of the most beautifully rendered Santas in the set.
Read More
Fifty designers rated visual data, revealing that structure—not style—is the key driver of visual trust and credibility.
Read More
What is everyone’s favorite Halloween candy? 26 people sorted 24 candies on a 75% threshold. KitKat took 91.67%, M&Ms 92.86%. Jawbreakers, Laffy Taffy and York took no green dots at all. Three nougat bars were refused, Snickers split the room, candy corn landed at exactly 50%. Texture, not flavour, predicts the verdict.
Read More
“Spooky” and “scary” are not interchangeable. 22 people sorted 36 Halloween images: 16 spooky, 13 scary, 7 neither. Spooky was warm palettes, handmade texture and friendly ghosts. Scary was cold contrast, hard geometry, masks and faceless figures. Presence within safety versus absence within power.
Read More
For July 4th, we explored what “American” looks like in 2025—revealing shared civic values through visual perception mapping.
Read More
Twelve images somebody nodded at share half a dozen qualities, most of which mean nothing. The rejected images were in the same room, from the same board, seen by the same people, so whatever separates the two piles is the only variable that changed. A liked pile gives you a description. Liked against rejected gives you a reason, and a reason is what you can design from.
Read More
Point a perception test at AI agents and, early on, nothing happened: responses scattered, no pattern of preference anywhere, while every human audience clustered. Current models do produce a pattern now. It doesn't match the human one, and they write a confident, articulate explanation for it, which is the more dangerous version of the same problem.
Read More
A logo is not one thing you keep or replace. It is a bundle of decisions: letterforms, weight, shape, proportion, color, texture. Each does a wildly different amount of work in your audience's head. Some carry meaning people will notice losing. Some are just old. Testing tells you which, and turns a redesign from an aesthetics argument into a risk decision.
Read More
Red lions, bigger logos, the CEO who hates blue. We asked designers where clients cost them the most time, and it wasn’t vagueness. It was the specific, arbitrary, unarguable demand. Overstepping is a symptom of a gap in the process: the client is fully involved through discovery, then goes silent for six weeks.
Read More
Search failed rebrands and you get the same lesson twenty times: respect your equity, don't change too much. Every successful dramatic rebrand contradicts it. The real variable is whether the strategic premise about the audience was ever checked, or reasoned into existence by smart people in a room with a deck.
Read More
The usual fix for design by committee is fewer decision-makers. That treats headcount as the disease. A room is a machine for producing the appearance of consensus: two people talk, eighteen stay quiet, and the quiet is four different things, none of them agreement. Change the format and the number can go up instead of down.
Read More
A conventional brand audit collects everything and marks it on or off brand. That grades the work against a document a small group wrote at a particular moment, and a brand can be consistently wrong. The measured version adds one step: sample the spectrum, test it with the people closest to the brand, and compare what read as the company against what read as off.
Read More
Two segments respond to the same elements in different places. That split is the finding, not a flaw in the test. Overlap becomes the core of the brand. Divergence becomes a targeting decision. And the rule for what to segment is asymmetric: blending groups back together is free, and splitting them apart afterward is impossible.
Read More
Clients live in a world where A leads to B. Creative work has no such line, which is why “that doesn’t feel like us” is often the most precise note a client can give you. The gap isn’t obstruction, it’s vocabulary. Judgment and description are different skills. Ask for a reaction instead of a word.
Read More
A mood board records what the room agreed on. The market never sat in that room. Where mood boards work, where they stop, and what gets you real signal.
Read More
A concept board built to be tested holds raw design elements, not finished ideas. Line, form, texture, color and type, in enough volume to produce a pattern.
Read More
When a new leadership team stepped into Express Newark at Rutgers University, they inherited a decade of history — and a staff in wait-and-see mode. Here’s how a visual research process got past “work answers,” turned instinct into evidence, and made a strategic realignment feel like a shared project.
Read More
Vague creative feedback isn’t a sign the stakeholder is being difficult. It’s a sign they’re reacting to something real they don’t have the vocabulary to name. “Feels off,” “more premium,” “needs more energy”—each one is a precise feeling wearing an imprecise word. Here’s what those phrases actually mean, and how to translate them into something measurable.
Read More
Creative projects rarely lurch off course at the end—they start wobbling much earlier, the moment a new stakeholder, leadership change, or unresolved internal thread enters the room. Most pivots trace back to alignment assumed but never established. Here’s what actually triggers them—and how to spot the warning signs weeks before the deadline slips.
Read More
You can’t test a logo by posting it to a Facebook group, running it past a focus group, or asking your spouse. Those methods measure who’s loudest, not what your audience associates with the mark. Here’s how to test the parts that matter—and the one thing to measure above all: association.
Read More
A single round of major rebrand revisions costs $10K–$50K—and the invoice is the smallest part of it. The real cost is burnout, missed deadlines, and the people who quit after months of chasing “I just don’t love it.” Here’s where that cost actually lives, and how audience data shrinks it.
Read More
A year ago, AI couldn’t complete a Constellations perception test at all — every result was an even split. Today it handles simple prompts fine, but on the nuanced, abstract prompts branding depends on, it’s wildly off from human data. The danger isn’t that AI gets it wrong. It’s that it now sounds convincing while doing it.
Read More
What Heatmaps Actually Tell You About Your Audience (and What They Don’t)Quick answer: “Heatmap” means three different things depending on who you ask. A mouse-tracking heatmap (Hotjar) shows where cursors move on a page. An eye-tracking heatmap (Tobii) shows where eyes land in the first instant, which is mostly survival-level reaction. A perception heatmap, the […]
Read More
Creative direction shouldn’t be decided by decibel. It should be decided by audience data. See how perception mapping aligned over 20 stakeholders with zero revisions.
Read More
Focus groups were built in 1946 for conversation, not visual creative: 8-12 people, one confident voice, anchoring bias steering the room. What each alternative actually measures, from survey panels to A/B tests, and why Audience Perception Maps are the only one that captures what 300-500 people see, in under 48 hours.
Read More
Most brand launches don't go badly. They go blindly. Focus groups give you 12 people and anecdotes, surveys give you stated preference, A/B tests cost six figures and burn trust. Audience perception mapping shows what 300-500 people see, in under 48 hours, and aligned 22 stakeholders with zero revisions.
Read More
Approval means nobody blocked it. Alignment means the team believes in it. The gap costs agencies $30,000-$60,000 per mid-market project in rework, lost launch windows and eroded trust. The five-step process that got 22 stakeholders to one direction with zero revisions.
Read More
Clients react to design in seconds and explain it in vague words. Drawing on choice-blindness research, dual process theory and interviews with working designers, this paper traces the perceptual gap, tests focus groups, surveys and neuromarketing against it, and shows how visual mapping closes it.
Read More
A global capital construction software company needed a brand that matched its leadership position. Four Audience Perception Maps went to 500+ employees as a proxy for hard-to-reach executives. Innovation was expected; confident leadership won. Twenty-two stakeholders, one direction, zero reworks.
Read More
A liberal arts college tested campus imagery with prospective students. Mixed maps confirmed the obvious: energetic scenes win. Giving each Audience Perception Map one theme, academic or student life, surfaced which classrooms felt welcoming and which activities stood out. Test structure decides what you see.
Read More
Four rounds of development, alignment reached, then a senior stakeholder redirects the work in the final week and the campaign underperforms. The creative team owns the outcome without the control. Design due diligence moves audience testing upstream, so shared evidence settles the direction before production begins.
Read More